Anil Kumar Verma, Gaurav Sharma, Kuldeep Singh - WILEY ACING THE GATE COMPUTER SCIENCE AND INFORMATION TECHNOLOGY (2015, Wiley India) PDF

WILEY ACING THE GATE
COMPUTER SCIENCE AND

INFORMATION TECHNOLOGY
COMPUTER SCIENCE AND
INFORMATION TECHNOLOGY
Anil Kumar Verma

Associate Professor
Department of Computer Science and Engineering
Thapar University, Patiala
Gaurav Sharma
Assistant Professor
School of Computing Science and Engineering
Galgotias University, Greater Noida
Kuldeep Singh
Department of Computer Science and Engineering
Thapar University, Patiala
COMPUTER SCIENCE AND INFORMATION TECHNOLOGY
Copyright © 2015 by Wiley India Pvt. Ltd., 4435-36/7, Ansari Road, Daryaganj, New Delhi-110002.
All rights reserved. No part of this book may be reproduced, stored in a retrieval system, or transmitted in
any form or by any means, electronic, mechanical, photocopying, recording or scanning without the written
permission of the publisher.
Limits of Liability: While the publisher and the author have used their best efforts in preparing this book,
Wiley and the author make no representation or warranties with respect to the accuracy or completeness of
the contents of this book, and specifically disclaim any implied warranties of merchantability or fitness for
any particular purpose. There are no warranties which extend beyond the descriptions contained in this
paragraph. No warranty may be created or extended by sales representatives or written sales materials.
Disclaimer: The contents of this book have been checked for accuracy. Since deviations cannot be
precluded entirely, Wiley or its author cannot guarantee full agreement. As the book is intended for
educational purpose, Wiley or its author shall not be responsible for any errors, omissions or damages
arising out of the use of the information contained in the book. This publication is designed to provide
accurate and authoritative information with regard to the subject matter covered. It is sold on the
understanding that the Publisher is not engaged in rendering professional services.
Other Wiley Editorial Offices:

John Wiley & Sons, Inc. 111 River Street, Hoboken, NJ 07030, USA
Wiley-VCH Verlag GmbH, Pappellaee 3, D-69469 Weinheim, Germany
John Wiley & Sons Australia Ltd, 42 McDougall Street, Milton, Queensland 4064, Australia
John Wiley & Sons (Asia) Pte Ltd, 1 Fusionpolis Walk #07-01 Solaris, South Tower Singapore 138628
John Wiley & Sons Canada Ltd, 22 Worcester Road, Etobicoke, Ontario, Canada, M9W 1L1
First Edition: 2015

ISBN: 978-81-265-5086-9
ISBN: 978-81-265-8199-3 (ebk)
www.wileyindia.com
PREFACE
The objective of writing this book is to help the GATE aspirants in understanding the various subjects of Computer
Science and Information Technology. Being a national level exam, it measures the candidate’s ability on different
parameters, namely, basic fundamentals, concepts, interpretation and analytical ability. This examination assumes
more importance as the GATE score has been used by public sector undertakings (PSUs) for recruiting graduate-level
engineers at the entry level, JRF at CSIR Labs, etc. Even certain foreign universities (like Singapore, Germany) have
also started recognizing the GATE score, for using it as a benchmark for providing scholarships for students interested
in pursuing PG and PhD programs.
The major strength of this book lies in the demystifying complex concepts by unpeeling levels of abstraction down
to the basic level. The book follows a balanced approach on systematic representation of the text. Each chapter begins
with the basics and then in-depth concepts are described with the help of various example(s)/numerical(s) and sche-
matic diagrams. Different ways of interpretation are also highlighted while providing the solutions to previous year
GATE question papers and then the chapter is culminated with a summary of important terms/formulas used in that
chapter. Therefore, the book is self-contained.
The organization of this book reflects our belief that significant amount of practice is required for achieving the
c omplete mastery. Therefore, each chapter contains a set of multiple questions of different levels (approximately
100 questions in each chapter) for self-assessment/mock test preparation, designed by keeping in mind the latest
pattern.
While writing this book, we have drawn upon the long experience of teaching the subjects as well as our own under-
standing of the subjects. The contents have been carefully compiled and have been presented in a lucid manner to
the reader. Utmost care has also been taken to relieve the candidate from the unnecessary burden of recollecting the
concepts during the preparation, so that the student qualifies with a high GATE score.
We sincerely hope that students looking for an up-to-date, excellent self-study, compact and comprehensive book
will be more than happy to have this book.
It is innate that some error(s) might have crept in, we shall be thankful to the reader for highlighting the same as
well as the suggestions in the improvement of the text presented.
FM.indd 5 4/9/2015 10:14:41 AM

vi PREFACE
ACKNOWLEDGEMENTS
Writing a book, is like a major project and is rarely the work of a single individual.
We thank our teachers and seniors for providing the wealth of knowledge, sharing the depth and breadth of their
experience and answering a barrage of our questions. Our numerous students, whose valuable discussions are a key
constituent of this book.
During the compilation and writing of this work, we have been fortunate enough to receive useful suggestions, timely
assistance and humble support from our colleagues and friends.
Special thanks to our family members who were extremely generous with their time being consumed for writing this
book and tolerated us when we were trying tooth-and-nail to meet the deadlines.
We are grateful to the Wiley India, particularly, Siddhartha Deb, Publishing Editor, for initiating all of us into this
venture and Sruthi Guru, Developmental Editor, for the continuous cooperation and assistance for the timely feedback.
Without their conscientious effort and personal involvement, this piece of work would not have been possible. We
would also like to thank the production team, Rakesh Poddar and Samir Das, for their efforts.
We wish to acknowledge the efforts of our reviewer(s) for providing countless suggestions for improving the content
and presentation.
Thank you!
Anil Kumar Verma (akvtu.1001@gmail.com)
Gaurav Sharma (sharmagaurav86317@gmail.com)
Kuldeep Singh (kuldeepisp@gmail.com)
FM.indd 6 4/9/2015 10:14:42 AM

ABOUT THE AUTHORS
Anil Kumar Verma is currently working as an Associate Professor in the Department of CSE,
Thapar University, Patiala. He is a member of MISCI (Turkey), LMCSI (Mumbai), GMAIMA
(New Delhi) and certified software quality auditor by MoCIT, Govt. of India. He received his
B.S., M.S. and PhD, majoring in Computer Science and Engineering. He has a rich teaching expe-
rience of over 23 years, which include the post of Lecturer at M.M.M. Engineering College (now,
MMM University of Technology), Gorakhpur from 1991 onwards, apart from being a visiting
faculty to many institutions. He has visited USA, South Korea and Japan for academic purposes,
and published over 150 papers in refereed journals and conferences (India and abroad). He is
member of various program committees for different international/national conferences and is on the review board of
various international journals. His research interests include wireless networks, routing algorithms and securing Ad hoc
networks. He can be reached at akvtu.1001@gmail.com.
Gaurav Sharma is currently working as an Assistant Professor in the School of Computing

Science and Engineering at Galgotias University, Greater Noida. He received his M.E.(CSE)
and PhD degrees from Thapar University, Patiala, and has published various papers in refereed
journals and conferences. He is a member of IEEE and ACM, and of various program commit-
tees for different international/national conferences, besides being on the review board of various
internationally renowned journals. His research interests includes routing and security issues in
Ad hoc networks. He can be reached at sharmagaurav86317@gmail.com.
Kuldeep Singh is currently a PhD research scholar and freelance consultant at Thapar
University, Patiala, working in the area of Ad hoc Network Security. He has done his M.E. (IS)
from Thapar University. Prior to that he had done his MCA and qualified GATE and CSIR-
UGC NET. He has commendable and growing list of publications, and is a member of reputed
computer societies like ACM. He can be reached at kuldeepisp@gmail.com.
FM.indd 7 4/9/2015 10:14:42 AM

FM.indd 8 4/9/2015 10:14:42 AM
ACING THE GATE
ABOUT GATE
Purpose of Examination
The Graduate Aptitude Test in Engineering (GATE) essentially examines the aptitude of an engineering graduate.
It differs from other examinations as it calls for an aptitude based learning and approach on engineering topics. Being
a national level exam, GATE measures the candidate’s ability on different parameters, namely, basic fundamentals,
concepts, interpretation and analytical ability. This examination assumes more importance as the GATE score has
been used by public sector undertakings (PSUs) for recruiting graduate-level engineers at the entry level, JRF at CSIR
Labs, etc. Even certain foreign universities (like Singapore, Germany) have also started recognizing the GATE score,
for using it as a benchmark for providing scholarships for students interested in pursuing PG and PhD programs.
1. Financial assistance: A GATE qualified student pursuing PG is liable to receive a stipend of H12,000/- pm
for 24 months from MHRD, India and other Government agencies.
2. Jobs in Public Sector Undertakings: Major PSUs such as – Indian Oil Corporation Ltd. (IOC), Power Grid
Corporation of India Ltd.(PGCIL), Oil and Natural Gas Corporation(ONGC), Punjab State Power Corporation
Ltd. (PSPCL), Bharat Broadband Network Limited (BBNL), etc. are recruting candidates based on their GATE
score.
3. Foreign Universities: National University, Singapore,; Nanyang University, Singapore; and some universities
in Germany do consider the GATE score.
Applying for the Examination
1. Eligibility: The candidate fulfilling the following eligibility criteria can apply for the examination:
Bachelor’s degree in Engineering/Technology/Architecture/ Pharmacy (Post-Diploma/Post B.Sc./4 years after 10+2)
OR
Master’s degree in any branch of Science/ Mathematics/ Statistics/ Computer Applications or equivalent
Note: Students currently in the final year of the above courses can also apply.
FM.indd 9 4/9/2015 10:14:42 AM

x Acing the GATE
2. Application Procedure: GATE Online Application Processing System (GOAPS) Website opens for enrol-
ment, application filling, application submission starts in month of September. For details, refer to the website of
the current organizing institute.
3. Examination Schedule: GATE is conducted in online (Computer Based Test (CBT), mode only and is
spread over two weeks. It starts in the last week of January to first week of February. The GATE (CS/IT) paper
is held in three sessions.
4. Announcement of Results: The results are announced on the GATE Online Application Processing System
(GOAPS) Website managed by IITs/IISc in the month of March.
Structure of Examination
The question paper comprises 65 questions carrying 100 marks to be attempted in 3 hours. The paper is distributed
over General Ability, Engineering Mathematics and Technical Subject.
1. General Ability accounts for 15% of the question paper containing 10 questions carrying 15 marks (5 questions
of 1 mark each and 5 questions of 2 marks each).
2. Engineering Mathematics accounts for 15% of the question paper.
3. Computer Science and Information Technology (CS & IT) accounts for 70% of the question paper.
The emphasis is on the following type of questions:
1. Recall: Based on facts, principles or laws involved. So, the candidate has to answer them from his/her memory
or simple computation.
2. Comprehension: Based on candidate’s understanding of basic concepts that help him to draw conclusions.
3. Application: Based on the ability of the candidate to apply his/her knowledge through computation or logical
reasoning.
4. Analysis and Synthesis: Based upon diagrams/data/tables/images that help in analysis. It also involves sepa-
rating useful information from the irrelevant information.
The types of questions asked are Multiple Choice Questions in which one correct option out of the four given
options needs to be chosen; and Numerical Answer Questions in which the correct answer needs to be keyed in
using an online keyboard.
The marking scheme is as follows:
1. Multiple Choice Questions: These questions carry 1 or 2 marks. There is negative marking for every
incorrect answer. For every wrong answer, 1/3 is deducted for a 1-mark question and 2/3 marks are deducted for
a 2-mark question.
2. Numerical Answer Questions: These questions carry 1 or 2 marks. No choices are given for these questions.
There is no negative marking for these questions.
Computerized Mode of Examination
From 2014 onwards, GATE is conducted mandatorily in online mode for all subject papers. Prior to this, it was
conducted online for 4 papers in 2011 and 6 papers in 2012 and 2013.
In the Online Computer Based Test (CBT) mode, the candidate is required to select the correct option for MCQ
type and enter the answer (numerical value) for numerical answer type question using a mouse on a virtual keyboard
(keyboard of the computer will be disabled). Blank paper pads for rough work are provided for rough work, and these
have to be returned back after the examination with roll numbers mentioned in them.
At the end of the 3-hour window, the computer will automatically close the screen from further actions. Another
screen with submit button will appear which needs to be clicked to end the examination.
FM.indd 10 4/9/2015 10:14:42 AM

Acing the GATE xi
ATTRIBUTES FOR SUCCESS IN EXAMINATION
Preparing for the Examination
Knowing Your Subject (Knowledge of Material)
The exhaustive syllabi could be overwhelming as it covers the entire course studied over a period of 6-7 semesters.
Thus, the emphasis must be on “knowing the material”, and not just being familiar with the material.
1. A systematic organization of the topics can be understood by going through the examination syllabus of the
subject and aligning it with the course studied.
2. Practicing questions from topics having more weightage to the ones having lower weightage helps understand the
fundamentals of concepts and their application.
3. Familiarize with the use of proper syntactical expressions, assumptions, laws, theorems, standard symbols and
notations for solving various problems based on programming codes and logic.
Preparation of a Study Plan
A well-chartered study plan taking into account the focus areas and learning objectives based on analysis of previous
year papers is important to build a connect between knowing the subject and its quantitative applications. The most
relevant topics for each unit and unit-wise weightage is provided in the part opener preceding each chapter opener.
These can be helpful in deriving focus areas.
1. The study plan may include a schedule of chapter or sub-topic level study with equal focus on revising fundamen-
tals and practicing sufficient questions based on each topic.
2. The focus areas can be evaluated based on individual assessment into strong and weak areas. The time devoted
to revise fundamentals in weak areas should be compensated with the time spent in solving questions in strong
focus areas.
3. The next step would be to spend sufficient time in practicing questions in weak areas as well as assess the
progress made.
4. After having enough confidence on both strong and weak areas, mock tests covering all the topics can be
attempted. Alternately, one-fourth of MCQs from each topic can be attempted.
Refining Problem-Solving Skills
The examination tests the ability to analyze information provided, make judgments, apply the understanding of topic
to arrive at the best-suited answer.
1. A quick revision of all the important formulas and unit summary would be useful before attempting practice tests
or mock tests.
2. Through the solved examples, the most effective way to solve a problem and the possible mistakes that led to the
wrong options may be identified.
3. Attempt as many previous years’ GATE questions based on the topic and follow the above approach for
self-evaluation.
4. Practice more and more questions based on the problem areas and rectify mistakes committed in the earlier
practice rounds.
FM.indd 11 4/9/2015 10:14:42 AM

xii Acing the GATE
Taking the Examination
Effective Preparation and Time Management
Towards the end of study plan, it is important to focus on exam preparedness and time management skills.
1. It is important to enhance problem-solving skills by continuously practicing questions. While doing so, it is
important to give equal importance to both 1-mark and 2-mark type questions. Although the time required for
solving 1-mark questions might be a little less than the 2-mark questions. It is important to practice both types
to enhance the speed and proficiency to attempt all questions types.
2. The time taken to solved a complete paper replicating the examination paper, that is, a mock test or practice test
can be evaluated, and the number of incorrectly answered questions can be identified.
3. The above exercise is helpful in determining the questions which took longer time to solve, and those for which
preparation was lacking in some form. More and more of these questions can be practiced upon to increase
efficiency and accuracy for solving.
Maximizing your Exam-Taking Efficiency
As the examination is conducted in online (computer based test) mode, it is helpful to familiarize oneself with the
examination procedure and instructions by going through the screen layout, button functions, etc., by accessing the
mock test available on GAOPS website.
1. It is important to take proper rest and sleep well in the night before the examination, to avoid having a cluttered
memory on the day of the examination.
2. Read the instructions carefully within the allotted time, and raise any equipment/software concerns immediately
to the invigilator.
3. Once the examination begins, be calm and adopt a positive attitude. Read the question carefully and then the
options. Carefully note the negatives in the question, such as “Which of the statements in incorrect?”, etc., and
work accordingly.
4. Set different time limits for the different sections, giving the maximum time for the technical section.
5. If any question appears unfamiliar or it is difficult to recollect the topic on which the question is based upon, then
mark it for review and move to the next question. Reserve some time to revisit such questions and try to narrow
down the possible responses and make an educated guess.
6. Constantly monitor the progress against the time remaining, which is displayed on the top right corner of the
screen.
BEST OF LUCK!!!
FM.indd 12 4/9/2015 10:14:42 AM

CONTENTS
Preface v
Acing the GATE ix
About GATE ix
Attributes for Success in Examination xi
1 Digital Electronics 3
1.1 Introduction 3
1.2 Number System 3
1.2.1 Conversions of Number System 4
1.2.2 Complement of a Number 4
1.2.3 Representation of Negative Numbers 5
1.2.4 The IEEE Standard for Floating Point Numbers 5
1.3 Boolean Logic 6
1.3.1 Boolean Algebra Laws (Huntington’s Postulates) 7
1.3.2 Duality Theorem 7
1.3.3 Consensus Theorem 7
1.3.4 Positive Logic and Negative Logic 8
1.4 Digital Circuits 11
1.4.1 Combinational Circuits 11
1.4.2 Sequential Circuits 16
1.5 Digital Logic Families 26
FM.indd 13 4/9/2015 10:14:42 AM

xiv CONTENTS
Important Formulas 28
Solved Examples 28
Gate Previous Years’ Questions 33
Practice Exercises 46
Answers to Practice Exercises 51
2 Computer Organization and Architecture 57

2.1 Introduction 57
2.2 Computer Architecture 57
2.2.1 Register Set 57
2.2.2 Quantitative Principles to Design High-Performance Processor 58
2.3 Machine Instructions and Addressing Modes 58
2.3.1 Machine Instructions 58
2.3.2 Addressing Modes 59
2.4 Arithmetic Logic Unit 61
2.4.1 Arithmetic Micro-Operations 61
2.4.2 Logic Micro-Operations 61
2.5 CPU Control Design 62
2.5.1 Instruction Execution 63
2.5.2 CPU Data Path 64
2.5.3 Control Unit Design 64
2.5.4 RISC versus CISC Processors 65
2.6 I/O Interface (Interrupt and DMA Mode) 65
2.7 Instruction Pipelining 67
2.8 Memory Hierarchy 72
2.8.1 Main Memory 72
2.8.2 Secondary Memory 73
2.8.3 Cache Memory 73
2.8.4 Cache Mapping Techniques 74
Solved Examples 77
GATE Previous Years’ Questions 80
3 Programming and Data Structures 105

3.1 Introduction 105
3.2 Basic Terminology 105
3.2.1 Programming Languages 105
3.2.2 Classification of High-Level Languages 106
3.2.3 Procedural Programming and Object-Oriented Programming 106
3.2.4 Data Types 107
3.2.5 Control Flow Statements 108
3.2.6 Array 110
3.2.7 Function and Recursion 111
3.2.8 Pointers 113
FM.indd 14 4/9/2015 10:14:42 AM

CONTENTS xv
3.2.9 Parameter Passing 115

3.2.10 Structures and Unions 117
3.2.11 Enumerated Data Types 117
3.2.12 Scoping 117
3.2.13 Binding 118
3.2.14 Abstract Data Types 119
3.3 Stack 119
3.3.1 PUSH Operation on Stack 119
3.3.2 POP Operation on Stack 119
3.3.3 Application of Stack 120
3.4 Queue 122
3.4.1 Basic Operations on Queue 122
3.4.2 Types of Queue 123
3.4.3 Applications of Queues 123
3.5 Linked List 123
3.5.1 Basic Operations 123
3.5.2 Linked List Implementation 124
3.6 Trees 125
3.6.1 Binary Tree 125
3.6.2 Types of Binary Trees 125
3.6.3 Array Representation of Binary Trees 126
3.6.4 Tree Applications 126
3.6.5 Binary Search Tree 126
3.6.6 Graphs 127
Solved Examples 138
4 Algorithms 177
4.2 Algorithm 177
4.2.1 Properties of Algorithm 177
4.2.2 Steps to Solve a Problem 177
4.2.3 Algorithm Analysis 178
4.2.4 Asymptotic Notation 178
4.2.5 Recurrence Relations 180
4.3 Hashing 182
4.3.1 Hash Table 182
4.3.2 Hashing Functions 182
4.3.3 Collisions 182
4.4 Binary Heap 184
4.4.1 Insertion in Min-Heap Tree 185
4.4.2 Deletion in Min-Heap Tree 185
4.4.3 Time Complexity 186
FM.indd 15 4/9/2015 10:14:42 AM

xvi CONTENTS
4.5 Searching and Sorting 186

4.5.1 Linear Search 186
4.5.2 Binary Search 187
4.5.3 Bubble Sort 187
4.5.4 Selection Sort 187
4.5.5 Insertion Sort 188
4.5.6 Heap Sort 188
4.5.7 Merge Sort 190
4.5.8 Quick Sort 192
4.5.9 Randomized Quick Sort 193
4.5.10 Counting Sort 193
4.5.11 Comparison of Sorting Techniques 194
4.6 Graph 194
4.6.1 Types of Graph 194
4.6.2 Types of Simple Graph 194
4.6.3 Graph Representation 195
4.7 Greedy Approach 195
4.7.1 Applications of Greedy Approach 196
4.7.2 Fractional Knapsack 196
4.7.3 Huffman Coding 197
4.7.4 Minimum Spanning Tree 198
4.7.5 Single-Source Shortest Path 200
4.8 Graph Traversal 201
4.8.1 Breadth-First Traversal 201
4.8.2 Depth-First Traversal 202
4.9 Dynamic Programming 203
4.9.1 Applications of Dynamic Programming 204
4.9.2 Fibonacci Series 204
4.9.3 0/1 Knapsack 204
4.9.4 Longest Common Subsequence 205
4.10 All-Pair Shortest Path 205
4.10.1 Floyd’s Algorithm 205
4.11 Concepts of Complexity Classes 205
4.11.1 P-Complexity Class 205
4.11.2 NP Complexity Class 206
4.11.3 NP-Complete 206
4.11.4 NP-Hard 206
Solved Examples 207
5 Theory of Computation 237

5.2 Finite Automata 237
5.2.1 Finite Automaton Model Characteristics 238
5.2.2 Technical Terms 238
5.2.3 Grammar 239
FM.indd 16 4/9/2015 10:14:42 AM

CONTENTS xvii
5.2.4 Noam Chomsky Grammar Classification 239

5.2.5 Chomsky Hierarchy 240
5.2.6 Different Grammar Types 240
5.3 Finite Automata and Regular Language 241
5.3.1 Deterministic Finite Automata 241
5.3.2 Non-deterministic Finite Automata 241
5.3.3 Comparison of DFA and NFA 241
5.3.4 DFA and NFA Design 241
5.3.5 Applications of DFA/NFA 243
5.3.6 Regular Expression and Grammar 243
5.3.7 Pumping Lemma 244
5.3.8 Myhill—Nerode Theorem 245
5.3.9 Transducer or Finite State Machine 245
5.3.10 Conversion in Finite Automata 246
5.3.11 Properties of Regular Sets/Languages 249
5.3.12 Some Important DFA 250
5.4 Pushdown Automata 250
5.4.1 Model of PDA 250
5.4.2 Transition Function 251
5.4.3 Non-deterministic PDA 251
5.4.4 Deterministic PDA 251
5.4.5 Parsing 252
5.5 Context-Free Grammar and Languages 252
5.5.1 Context-Free Grammar 252
5.5.2 Standard Context-Free Language 252
5.5.3 Derivation Tree or Parse Tree 253
5.5.4 Ambiguous Grammar 254
5.5.5 Removal of Ambiguity 254
5.5.6 Context-Free Grammar Simplification 254
5.5.7 Chomsky Normal Form 256
5.5.8 Greibach Normal Form 256
5.5.9 Identification of Language 256
5.6 Turing Machine 256
5.6.1 Model of a Turing Machine 257
5.6.2 Designing a Turing Machine 258
5.6.3 Recursive and Recursive Enumerable Languages 258
5.6.4 Variation of Turing Machine 258
5.7 Closure and Decidability 258
Solved Examples 260
6 Compiler Design 287

6.2 Compilers and Interpreters 287
6.2.1 Compiler 287
FM.indd 17 4/9/2015 10:14:42 AM

xviii CONTENTS
6.2.2 Interpreter 288

6.2.3 Phases of a Compiler 288
6.2.4 Grouping of Phases 290
6.3 Lexical Analyzer 290
6.3.1 Functions of a Lexical Analyzer 290
6.3.2 Implementation of a Lexical Analyzer 290
6.4 Parser 291
6.4.1 Context-Free Grammar 292
6.4.2 Derivation Tree or Parse Tree 292
6.4.3 Ambiguous Grammar 293
6.4.4 Top-Down Parser 293
6.4.5 Bottom-Up Parser 297
6.5 Syntax-Directed Translation 302
6.5.1 Attributes of Syntax-Directed Translation 302
6.5.2 Types of Syntax-Directed Translation 302
6.5.3 Applications of SDT 302
6.6 Runtime Environment 303
6.6.1 Storage Organization 303
6.6.2 Activation Record and Activation Trees 304
6.6.3 Procedure Call Return Model 304
6.6.4 Lexical Versus Dynamic Scoping 305
6.6.5 Symbol Table 305
6.7 Intermediate Code Generation 305
6.7.1 Intermediate Representations 305
6.8 Code Optimization 307
6.8.1 Types and Levels of Optimization 307
6.8.2 Primary Source of Optimization 307
Solved Examples 309
7 Operating System 327

7.2 Types of Operating System 328
7.2.1 Batch Operating System 328
7.2.2 Multiprogramming Operating System 328
7.2.3 Multitasking Operating System 328
7.2.4 Multiprocessor Operating System 328
7.2.5 Real-Time Operating System 329
7.3 Process Management 329
7.3.1 Process 329
7.3.2 Schedulers 330
7.4 CPU Scheduling 331
7.4.1 Scheduling Algorithms 331
FM.indd 18 4/9/2015 10:14:42 AM

CONTENTS xix
7.5 Process Synchronization 335

7.5.1 Types of Processes 335
7.5.2 Interprocess Communication 336
7.5.3 Classical Synchronization Problems 336
7.5.4 Race Condition 336
7.5.5 Critical Section 337
7.5.6 Peterson’s Algorithm for Two Processes 337
7.5.7 Semaphores 338
7.5.8 Deadlock and Starvation 339
7.6 Deadlocks 339
7.6.1 Resource Types 340
7.6.2 Characteristics of a Deadlock 340
7.6.3 Resource Allocation Graph 340
7.6.4 Deadlock Prevention 341
7.6.5 Deadlock Avoidance 342
7.6.6 Deadlock Detection and Recovery 346
7.7 Threads 347
7.7.1 Benefits of Threads 347
7.7.2 Types of Threads 347
7.8 Memory Management 348
7.8.1 Partitioning 348
7.8.2 Partition Allocation Policies 348
7.8.3 Loading and Linking 349
7.8.4 Paging 350
7.8.5 Multi-Level Paging 352
7.8.6 Segmentation 354
7.8.7 Virtual Memory 355
7.8.8 Page Replacement Algorithms 355
7.9 File System 359
7.9.1 Directory Structures 359
7.9.2 Allocation Methods 359
7.10 I/O Systems 363
7.10.1 Disk Structure 363
7.10.2 Disk Scheduling Algorithms 366
7.11 Protection and Security 368
7.11.1 Security 368
7.11.2 Security Environment 369
7.11.3 Cryptography 369
7.11.4 Attacks from within the System 370
7.11.5 Attacks from Outside the System 370
7.11.6 Anti Virus Approaches 371
Solved Examples 372
FM.indd 19 4/9/2015 10:14:42 AM

xx CONTENTS
8 Databases 403
8.1.1 Traditional File Processing Approach 403
8.1.2 Database Management System 404
8.2 Components of Database Systems 404
8.3 DBMS Architecture 405
8.3.1 3-Tier Architecture 405
8.3.2 Data Independence 405
8.4 Data Models 405
8.4.1 Relational Model 406
8.4.2 ER Model 407
8.5 Database Design 410
8.5.1 Integrity Constraints 410
8.5.2 Normal Forms 411
8.5.3 Attribute Closure 411
8.5.4 Key 411
8.5.5 Decomposition 412
8.6 Query Languages (SQL) 414
8.6.1 SQL Commands 414
8.7 File Structures 416
8.7.1 Sequential Files 416
8.7.2 Indexing 416
8.7.3 B Tree 416
8.7.4 B+ Trees 416
8.8 Transactions and Concurrency Control 417
8.8.1 Transactions 417
8.8.2 Schedule 417
8.8.3 Classification of Schedule Based on Recoverability 419
8.8.4 Classification of Schedule Based on Serializability 420
8.8.5 Concurrency Control Protocol 423
Solved Examples 425
9 Information Systems and Software Engineering 453

9.2 Information Systems 453
9.2.1 Components of Information Systems 454
9.2.2 Types of Information Systems 454
9.3 Software 456
9.3.1 Characteristics of Software 456
9.3.2 Software Engineering 457
9.4 Process Models 458
9.4.1 Conventional Process Model 458
9.4.2 Evolutionary Process Model 459
FM.indd 20 4/9/2015 10:14:43 AM

CONTENTS xxi
9.5 Measurement of Metrics 461

9.5.1 Metrics 461
9.5.2 Size-Oriented Metrics 462
9.5.3 Effort and Schedule (Duration) Estimation 463
9.6 Risk Analysis 466
9.6.1 Risk Identification 466
9.6.2 Risk Projection or Risk Estimation 466
9.7 Software Development Life Cycle 467
9.7.1 Requirement 467
9.7.2 Design 470
9.7.3 Coding 472
9.7.4 Testing 472
Solved Examples 477
10 Computer Networks 491

10.2 Network 491
10.2.1 Network Topology 492
10.3 LAN Technologies 492
10.3.1 Ethernet 492
10.3.2 Token Bus 493
10.3.3 Token Ring 493
10.4 ISO/OSI Stack 494
10.4.1 ISO/OSI Model 494
10.5 Routing Algorithms 498
10.6 Network Layer Protocols 501
10.6.1 Internet Protocol (IPv4) 501
10.6.2 ICMP 505
10.7 Layer 4: Transport Layer 505
10.8 Congestion 505
10.8.1 Congestion Versus Flow Control 506
10.9 User Datagram Protocol and Transmission Control Protocol 506
10.9.1 User Datagram Protocol 506
10.9.2 Transmission Control Protocol 506
10.10 Sockets 507
10.11 Layer 5: Session Layer 507
10.12 Layer 6: Presentation Layer 508
10.13 Layer 7: Application Layer 508
10.14 Devices 509
10.15 Network Security 510
10.15.1 Basic Concepts in Cryptography 510
10.15.2 Digital Signature 511
10.15.3 Firewall 511
FM.indd 21 4/9/2015 10:14:43 AM

xxii CONTENTS

Solved Examples 513
11 Web Technologies 537

11.2 HTML 538
11.2.1 HTML Tags 538
11.2.2 HTML Attributes 538
11.2.3 HTML Elements 539
11.2.4 HTML Character Entities 539
11.2.5 HTML Formatting 539
11.2.6 HTML Phrase Tags 540
11.2.7 HTML Background 540
11.2.8 HTML Lists 540
11.2.9 HTML Links 540
11.2.10 HTML Images 540
11.2.11 HTML Tables 540
11.2.12 HTML Frames 541
11.2.13 HTML Forms 541
11.2.14 HTML Marquees 541
11.3 Cascading Style Sheets 541
11.4 XML 541
11.4.1 Advantages of XML 542
11.4.2 Document-Type Definition 542
11.4.3 Tag Rules for XML Documents 542
11.4.4 Types of XML Documents 543
11.4.5 XHTML 543
11.4.6 Document Object Model (DOM) 543
11.4.7 XUL 543
11.4.8 Flash and Silverlight 543
11.4.9 User-Interface Language 543
11.5 Basic Concepts of Client— Server Computing 544
11.5.1 Server Types 544
11.5.2 Stateless and Stateful Servers 544
11.5.3 Thin Client and Fat Servers 544
11.5.4 Functions of a Client 544
11.5.5 Functions of a Server 544
11.5.6 Topologies for Client-Server 545
11.5.7 Types of Client-Server Model 545
11.5.8 Merits and Demerits of Client-Server 546
11.6 J2EE platform 546
11.6.1 J2EE Services 546
11.6.2 J2EE Architecture 547
11.6.3 EJB Container 547
11.6.4 J2EE Application Server 547
FM.indd 22 4/9/2015 10:14:43 AM

CONTENTS xxiii

Solved Examples 548
Solved GATE (CS) 2014 Set 1 555

Index 631
FM.indd 23 4/9/2015 10:14:43 AM

FM.indd 24 4/9/2015 10:14:43 AM
UNIT I: DIGITAL ELECTRONICS
LAST SIX YEARS' GATE ANALYSIS
5
Number of questions
4
Marks 1
3 Marks 2
Total number of questions
2
0
2015 2014 2013 2012 2011 2010
Concepts on which questions were asked in the previous six years
Year Concepts
2015 K-maps, Half and full adder, JK flip-flop, Boolean operators, Prime implicates, Base
values, Johnsen counter, Binary operator, D flip-flop
2014 Minterm expressions, Combinational functional blocks, Boolean expressions, flip-flops
2013 2’s complement, Truth tables, Logic gates, RAM chips
2012 K-maps, Truth tables, IEEE single precision
2011 Flip-flops, Boolean expressions, Logic gates
2010 K-maps, 2’s complement, Multiplexer, Boolean expression, Sequential circuits
Chapter 1.indd 1 4/9/2015 9:40:08 AM

Chapter 1.indd 2 4/9/2015 9:40:08 AM
CHAPTER 1
DIGITAL ELECTRONICS
Syllabus: Digital electronics: logic functions, minimization, design and synthesis of combinational and sequential
circuits, number representation and computer arithmetic (fixed and floating point).
1.1 INTRODUCTION Number system
Digital electronics represents discrete signals instead of Radix/Base representation Codes representation
signals in a continuous range. It uses two binary levels
0’s (corresponding to false) and 1’s (corresponding to
true). The main reason for advancement in digital elec-
Numeric codes Alphanumeric codes
tronics is integrated circuits (ICs).
EBCDIC
Weighted Non-weighted ASCII
1.2 NUMBER SYSTEM
Binary Gray code
Number system is an age old method to represent BCD XS3 code

numerals (Fig. 1.1). Decimal number system is the most 2421 code
common number system and binary number system is
used by computers. Figure 1.1 | Number system.
Chapter 1.indd 3 4/9/2015 9:40:08 AM

4 Chapter 1: Digital Electronics
Table 1.1 | Types of number systems

Base/Radix Unique Numbers Terminology Example
10 0 to 9 Decimal number system 10.47
8 0 to 7 Octal number system 65.32
2 0 to 1 Binary number system 110.11
16 0 to 15 Hexadecimal number system BAD.1A
0 to 9 are numerals and 10
to 15 are represented with
alphabets 10 = A, 11 = B,
12 = C, … 15 = F (111001.0100)2 = 1 × 25 + 1 × 24 + 1 × 23 + 0 × 22
+ 0 × 21 + 1 × 20 + 0 × 2-1
Radix or base is the number of unique digits, so the + 1 × 2-2 + 0 × 2-3 + 0 × 2-4
different base or radix can be expressed as represented
in Table 1.1. For example, (57)10 − here 10 is the base
= 32 + 16 + 8 + 0 + 0 + 1 + 0
(or radix) and 57 is the number. + 0.25 + 0 + 0 = (57.25)10
Another example is of (101.11)2 = (5.75)10 = (5.6)8 =

Example 1.1 (5.C)16.
Let us understand the logic of representation by consid-
ering (57.1)10. We know 57 means 50 + 7 + (1/10). Now, 1.2.2 Complement of a Number
representing the same by making use of radix (which
is 10), we get Any given number will have two complements:
1. Radix minus one complement (R-1’s
5 × (10)1 + 7 × (10)0 + 1 × (10)-1 complement): It is obtained by subtracting the
given number from the highest possible number. For
example, if there is a two-digit decimal number, it
1.2.1 Conversions of Number System will be subtracted from 99; whereas a 3-bit number
in a binary system will be subtracted from 111.
A number of a particular number system can be con- Note: The 1’s complement of a binary number
verted into another number system, as follows: is obtained by interchanging the 1’s and 0’s. For
example, 1’s complement of (1011)2 is (0100)2.
1. Decimal number system to any number
system: Consider the example of (57.3)10. To find 2. Radix complement (R’s complement): This
the binary equivalent, we have is also known as true complement. It is obtained
by adding one to R-1’s complement. For example,
2 57 (57)10 − its 9’s complement is 42 and its 10’s com-
2 28 1 plement is 43.
2 14 0 2 × 0.3 = 0.6 0 1.2.2.1 Subtraction Using R-1’s Complement

27 0 2 × 0.6 = 1.2 1
23 1 2 × 0.2 = 0.4 0 For the following subtractions,
21 1 2 × 0.4 = 0.8 0 A
(111001)2 (0.0100)2 −B
----------
Therefore, (57.3)10 = (111001.0100)2 ----------
2. Any number system to decimal number the steps involved are as follows:
system: Consider the example of (111001.0100)2.
Step 1. Find R-1’s complement of B.
To find the decimal equivalent we have
Step 2. Add A and B.
(111001.0100)2 = 1 × 25 + 1 × 24 + 1 × 23 + 0 × 22
Step 3. If there is carry, then add this carry to the
+ 0 × 21 + 1 × 20 + 0 × 2-1 answer.
+ 1 × 2-2 + 0 × 2-3 + 0 × 2-4
= 32 + 16 + 8 + 0 + 0 + 1 + 0
+ 0.25 + 0 + 0 = (57.25)10
Chapter 1.indd 4 4/9/2015 9:40:09 AM

1.2 NUMBER SYSTEM 5
If there is no carry, then find the R-1’s complement

of the answer and place a negative sign in front of the Problem 1.4: (011)2 − (101)2
answer. Solution:
Step 1. 2’s complement of 101 is 011.
Problem 1.1: (101)2 − (011)2
Step 2. The addition of these two numbers is (011
Solution: + 011 = 0 110).
Step 1. 1’s Complement of 011 is 100. Step 3. There is a no carry, so now we will find 2’s
Step 2. The addition of these two numbers is (101 + complement of the answer obtained (which is 010)
100 = 1 001). and place a negative sign in front of it. So, the answer
is (−010)2.
Step 3. There is a carry of 1, so now we will add this
carry to the answer (1 + 001), which is (010)2.
1.2.3 Representation of Negative Numbers
Problem 1.2: (011)2 − (101)2
In−(57)10 = (1111001), the most significant bit is 1,
Solution: showing that the number is negative binary number.
Step 1. 1’s Complement of 101 is 010.
Step 2. The addition of these two numbers is 1.2.4 The IEEE Standard for Floating Point
(011 + 010 = 0 101). Numbers
Step 3. There is a no carry, so now we will find 1’s The IEEE (Institute of Electrical and Electronics
complement of the answer obtained (which is 010) Engineers) has provided a standard for floating point
and place a negative sign in front of it. So, the answer numbers and their arithmetic. This standard speci-
is (−010)2. fies how single precision (32-bit) and double precision
(64-bit) floating point numbers are to be represented.
1.2.2.2 Subtraction Using R’s Complement
1. Single Precision: The IEEE single precision
f loating point standard representation requires a
In this method, we find R’s complement of the subtra-
32-bit word. The first bit to the left is the sign bit,
hend and we do not add the carry, as in step 2. Rest of
the next eight bits are the exponent bits, and the
the steps are the same.
final 23 bits form the fraction part (Fig. 1.2).
of a normalized number = (−1)s ×
Problem 1.3: (101)2 − (011)2
The value
e-127
1.m × 2
Solution: where s = 1 for negative number, s = 0 for positive
Step 1. 2’s complement of 011 is 101. number, m = mantissa and e = exponent.
Step 2. The addition of these two numbers is (101 + 2. Double Precision: The IEEE double precision
101 = 1 010). f loating point standard representation requires a
Step 3. There is a carry of 1, so now we will not add 64-bit word. The first bit is the sign bit, the next
this carry to the answer (001), which is (010)2. eleven bits are the exponent bits, and the final
52 bits form the fraction part (Fig. 1.3).
31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16 15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0
s exponent significand/mantissa
1-bit 8-bits 23-bits
Figure 1.2 | Single Precision (32 bits).
31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16 15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0
s exponent significand/mantissa
1-bit 11-bits 20-bits
significand (continued)
32-bits
Figure 1.3 | Double Precision (64 bits).
Chapter 1.indd 5 4/9/2015 9:40:10 AM

its output can be represented by Boolean expression.

Problem 1.5: Consider the decimal number: +116.25.
The symbols used to represent logic gates are shown
Represent it using single precision IEEE floating
in Fig. 1.4.
point standard.
Solution: The equivalent binary representation is
+1110100.01
Normalizing number = +1.11010001 × 26

= +1.11010001 × 2133−127 AND gate NAND gate

=+
1.11010001
× 210000101−01111111
In IEEE 754 format:
Sign Exponent Fraction OR gate NOR gate
0 10000101 11010001000000000000000
1.3 BOOLEAN LOGIC

XOR gate XNOR gate
NOT gate
The concept of Boolean logic was proposed by George
Boole, in which any information or data is represented
in the form of two states, which are complement to each
Figure 1.4 | Logic gates symbols.

other (such as 0 and 1, hot and cold, black and white,
north and south, etc.).
Logic gates form the basic building block of the digi-
tal system. A logic gate is an electronic device that takes Table 1.2 represents 15 basic operations that can be
logical decisions based on given input combinations and performed.
Table 1.2 | Basic operations in Boolean algebra

Inputs Operations
X Y Null Identity Inhibit Transfer
X Inhibit Y Y Inhibit X
0 0 0 1 0 0 0
0 1 0 1 0 1 1
1 0 0 1 1 0 0
1 1 0 1 0 0 1
Inputs Operations
X Y AND OR NOT XOR XNOR
NOT X NOT Y
0 0 0 0 1 1 0 1
0 1 0 1 1 0 1 0
1 0 0 1 0 1 1 0
1 1 1 1 0 0 0 1
(Continued)
Chapter 1.indd 6 4/9/2015 9:40:10 AM

1.3 BOOLEAN LOGIC 7
Table 1.2 | Continued
Inputs Operations
X Y NAND NOR Implication
X Implication Y Y Implication X
0 0 1 1 1 1
0 1 1 0 1 0
1 0 1 0 0 1
1 1 0 0 1 1
Table 1.2 is interpreted as follows: 3. X + X = X

1. Null − All the outputs are zero. 4. X + 1 = 1
2. Identity − All the outputs are one. 5. X + Y = Y + X
3. X inhibit Y − Represented as (x ⋅ ∼y)
4. Y inhibit X − Represented as (y ⋅ ∼x)
6. X + (Y + Z ) = (X + Y ) + Z

5. Transfer − Same as y 7. X ⋅ (Y + Z ) = X ⋅ Y + X ⋅ Z
6. AND − x ⋅ y (Similar to intersection in set theory) 8. ∼(X + Y ) = ∼X ⋅ ∼Y (Also called DeMorgan’s law)
7. OR − x + y (Similar to union in set theory)
9. X + (X ⋅ Y ) = X
8. NOT X − ∼x (Also known as complement of x)
9. NOT Y − ∼y (Also known as complement of y) 10. ∼(∼X ) = X
10. XOR − (∼x ⋅ y) + (x ⋅ ∼y) (Also known as parity
checker) 1.3.2 Duality Theorem
11. XNOR − (x ⋅ y) + (∼x ⋅ ∼y)
12. NAND − ∼(x ⋅ y) Every expression remains valid if we interchange the
13. NOR − ∼(x + y) operator and identity elements. Following laws are the
14. X implication Y (∼x + y) dual of the respective laws written above.
15. Y implication X (∼y + x)
1. X ⋅ 1 = X
2. (X ⋅ ∼X ) = X ⋅ X ′ = 0
Problem 1.6: It is defined that x * y = x ′y ′ + xy and

z = x * y . Among the following P, Q and R, which 3. X ⋅ X = X
statement is correct? 4. X ⋅ 0 = 0
P: x = y * z 5. X ⋅ Y = Y ⋅ X
Q: y = x * z 6. X ⋅ (Y ⋅ Z ) = (X ⋅ Y ) ⋅ Z
R: x * y * z = 1 7. X + (Y ⋅ Z ) = (X + Y ) ⋅ (X + Z )
8. (X ⋅ Y )′ = X ′ + Y ′
Solution:
9. X ⋅ (X + Y ) = X
x Y z P Q R
0 0 1 0 0 1 1.3.3 Consensus Theorem
0 1 0 0 1 1 It is applicable only if a Boolean function has
1 0 0 1 0 1 1. three variables
1 1 1 1 1 1 2. each variable is used two times
3. only one variable in complemented or non-
complemented form
1.3.1 Boolean Algebra Laws (Huntington’s
Then, the term related to the complemented and non-
Postulates)
complemented variable is the answer.
The laws of Boolean Algebra are given by Huntington’s
postulates: Example 1.2
1. X + 0 = X AB + A ′C + BC = AB + A ′C
2. X + (∼X ) = 1
Chapter 1.indd 7 4/9/2015 9:40:15 AM

1.3.4 Positive Logic and Negative Logic 1.3.4.2 Minimization of Expression
Usually, the two (binary) levels are represented in two A Boolean expression which cannot be further reduced
different ways: is said to be in minimal form. Each term in minimal
1 (signifying high) and 0 (signifying low) − Known as Boolean expression is called prime implicant. The follow-
positive logic ing are the two methods for minimization:
or 1. Boolean Theorems (postulates): Using the
1 (signifying low) and 0 (signifying high) − Known as Boolean postulates, we can simplify an expression
negative logic using algebraic method. This method is relatively
difficult as we need to remember all the postulates.
1.3.4.1 Canonical and Standard Forms Problem 1.7: Simplify the Boolean function X = (A + B)(A + C )
X = (A + B)(A + C ).
The minterms and the maxterms are given in Table 1.3.
Solution:
Product-of-sums (POS) form: X = AA + AC + AB + BC
F = (x ′y ′z ′) + (xy ′z ′) + (xyz)
= A + AC + AB + BC
= A(1 + C ) + AB + BC
= m0 + m4 + m7 = A(1) + AB + BC
= ∑ m(0, 4, 7) = A + AB + BC
= A(1 + B) + BC
Sum-of-products (SOP) form: = A(1) + BC
F = (X + Y + Z ) ⋅ (X ′ + Y + Z ) ⋅ (X + Y + Z ) = A + BC
= M 0 ⋅ M 4 ⋅ M7
Problem 1.8: Simplify the Boolean function X = A + A¢B
= ∏ M (0, 4, 7) X = A + A¢B.
In order to convert minterm to maxterm, put all those Solution:

numbers in maxterm which are not present in minterm, X = A(1 + B) + A ′B
and vice versa. = A ⋅ B + A + A ′B = A ⋅ B + A ′ B + A
= (A + A ′) ⋅ B + A = 1 ⋅ B + A = B + A
∑ m(0, 1, 5) = ∏ M (2, 3, 4, 6, 7)
Table 1.3 | Minterms and Maxterms

Minterms (SOP) Maxterms (POS)
Decimal Binary Term Designation Term Designation
0 000 x′y′z′ m0 X+Y+Z M0
1 001 x′y′z m1 X + Y + Z′ M1
2 010 x′yz′ m2 X + Y′ + Z M2
3 011 x′yz m3 X + Y′ + Z′ M3
4 100 xy′z′ m4 X′ + Y + Z M4
5 101 xy′z m5 X′ + Y + Z′ M5
6 110 xyz′ m6 X′ + Y′ + Z M6
7 111 xyz m7 X′ + Y′ + Z′ M7
Chapter 1.indd 8 4/9/2015 9:40:16 AM

1.3 BOOLEAN LOGIC 9
2. Karnaugh-Map Method (K-Map method): Grouping of two adjacent cells removes one vari-
Karnaugh-map (K-map) involves graphical repre- able and output will contain common variables in
sentation for simplifying Boolean expressions. It two terms.
consists of group of adjacent cells in which each 2. Quads: We can group four adjacent cells in K-map
cell corresponds to one of the combinations of max- forming a quad. Grouping a quad eliminates the
terms or minterms of n variables and is represented two variables that appear in both normal and com-
by group of literals. For n-variable K-map, there plemented form. Resultant terms contain (n − 2)
are 2n cells required (Figs. 1.5−1.7). From one variables, where n is the number of variables.
cell to the neighbouring cell there must be only 3. Octet: A group of eight adjacent cells is called
one literal change. While solving the K-map from octet. Grouping of 1’s eliminates three variables.
higher-order group to lower-order group only, the Resultant terms contain (n − 3) variables, where
simplification must be done. In K-map, leftmost n is the number of variables.
column and rightmost column of the same row are
considered to be adjacent. 1.3.4.4 SOP Simplification Using K-Map
If K = 1, it represents that all the cells are 1 so
there is no need of circuit. A higher priority will be given to more possible covering,
that is, octet then to quad and then to pairs. The follow-
1.3.4.3 Grouping of Cells for Simplification ing procedure is followed:
1. Check the SOP expression and convert to standard
The grouping of cells for simplification is done in the and canonical form if required.
following ways: 2. Draw the K-map based on the number of variables.
1. Pairs: If two adjacent cells either in row or in 3. On the basis of SOP canonical form, 1’s are entered
column contain 1’s, they are said to be in pairs. to corresponding cells. Group the adjacent cells
and take the common variable from each grouping.
4. Combine each group variable to obtain simplified
B 0 1 B 0 1 expression.
A A
0 00 01 0 0 1
1 10 11 1 2 3 1.3.4.5 POS Simplification Using K-Map
Two-variable K-map Cell designation
1. Check the POS expression and convert it to stan-
dard and canonical form if required.
B B B´ 2. Draw the K-map based on the number of variables.
A
A A + A+ 3. Enter 0’s in corresponding cells and group the
B B´ B adjacent cells.
A B B´
A´ A´B´ A´B A´ A´ + A´ + 4. Take the common variable from each grouping and
A AB´ AB B B´ write terms in OR form. The uncomplemented vari-
able is assigned 0 and complemented is assigned 1.
Cells with minterm Cells with maxterm
5. Combine each OR term from the different groups
Figure 1.5 | Two-variable K-map. to obtain simplified expression.
BC 00 01 11 10 BC 00 01 11 10
A A
0 000 001 011 010 0 0 1 3 2
1 100 101 111 110 1 4 5 7 6
Three-variable K-map Cell designation
BC BĆ BC BC´ BC B + C B + C´ B´ + C´ B´ + C
BĆ´
A A
A´ A´BĆ´ A´BĆ A´B C A´B C´ A A + B + C A + B + C´ A + B´ + C´ A + B´ + C
A ABĆ´ ABĆ AB C AB C´ A A´ + B + C A´ + B + C´ A´ + B´ + C´ A´ + B´ + C
Cells with minterm Cells with maxterm
Figure 1.6 | Three-variable K-map.
Chapter 1.indd 9 4/9/2015 9:40:17 AM

CD 00 01 11 10 CD 00 01 11 10
AB AB
00 0000 0001 0011 0010 00 0 1 3 2
01 0100 0101 0111 0110 01 4 5 7 6
11 1100 1101 1111 1110 11 12 13 15 14
10 1000 1001 1011 1010 10 8 9 11 10
Four-variable K-map Cell designation
CD C´D´ C´D CD CD´

AB
A´B´ A´BĆ´D´ A´BĆ´D A´BĆD A´BĆD´
A´B A´BC´D´ A´BC´D A´BCD A´BCD´
AB ABC´D´ ABC´D ABCD ABCD´
AB´ ABĆ´D´ ABĆ´D ABĆD ABĆD´
Cells with minterm
CD
AB C+D C + D´ C´ + D´ C´ + D
A+B A+B+C+D A + B + C + D´ A + B + C´ + D´ A + B + C´ + D
A + B´ A + B´+ C + D A + B´+ C + D´ A + B´+ C´ + D´ A + B´ + C´+ D
A´+ B´ A´+ B´+ C + D A´+ B´+ C + D´ A´+ B´+ C´+ D´ A´+ B´+ C´+ D
A´+ B A´+ B + C + D A´+ B + C + D A´+ B + C´+ D´ A´+ B + C´+ D
Cells with maxterm
Figure 1.7 | Four-variable K-map.
Problem 1.9: Simplify F (A, B, C , D) = ∑ m(0, 2, 5, 7, 13, 15Problem

) 1.10: Simplify F (A, B, C , D) = ACD + A ′B + D ′
D) = ∑ m(0, 2, 5, 7, 13, 15). F (A, B, C , D ) = ACD + A ′ B + D ′
Solution: Converting to standard SOP form. The first
Solution:
term ACD has one missing variable B.
CD
00 01 11 10 ACD = ACD(B + B ′) = ABCD + AB ′CD
AB
00 1 1 ABD Similarly,
0 1 3 2
A ′B = A ′B(C + C ′)(D + D ′)
01 1 1
4 5 7 6 = A ′BCD + A ′BCD ′ + A ′BC ′D + A ′BC ′D ′
BD
11 1 1 D ′ = D ′(B + B ′)(C + C ′) + (A + A ′)
12 13 15 14
Combining all terms, we get
10
8 9 11 10 F = ABCD + AB ¢CD + A¢BCD + A¢BCD ¢
+ A¢BC ¢D + A¢BC ¢D ¢ + ABC ¢D ¢
The simplified Boolean expression is
F = A¢B ¢D ¢ + BD. +AB ¢CD ¢ + AB ¢C ¢D ¢ + A¢B ¢CD ¢ + A¢B ¢C ¢D ¢
Chapter 1.indd 10 4/9/2015 9:40:19 AM

1.4 DIGITAL CIRCUITS 11
CD Problem 1.11: Simplify S = PM (0, 1, 2, 3, 4, 7)

00 01 11 10
AB
1
Solution:
00
D´
1
0 1 3 2 BC
A 00 01 11 10
01 1 1 1 1 A´B
4 5 7 6 0 0 0 0 0 Group 1 (A)
0 1 3 2
11 1 1 1 0 0
12 13 15 14 (B + C)1
4 5 7 6
10 1 1 11 1 10 AC Group 2
8 9 Group 3(B´ + C´)
The simplified Boolean expression is F = D ′ + A ′B + AC
The simplified Boolean expression is S = A ⋅ (B + C ) ⋅ (B ′ + C
F = D ′ + A ′B + AC . S = A ⋅ (B + C ) ⋅ (B ′ + C ′)
Problem 1.12: Simplify F (A, B, C , D) = ΠM (4, 5, 6, 7, 8, 12) ⋅ d(0, 1, 2, 3, 9, 11, 14)

Solution: A logic circuit output may be either 0 or 1 but some circuit has output that will never occur for certain
combinations of inputs. These conditions are don’t care conditions and represented by X in K-map.
CD
00 01 11 10
AB
00 X X X X A
0 1 3 2
01 O O O O
4 5 7 6
11 Need not worry about

O X
12 13 15 14 this don’t care case
C +D
10 O X X
8 9 11 10
Invalid grouping because it does

not have atleast one O
The simplified Boolean expression is F = A(C + D).
1.4 DIGITAL CIRCUITS 1.4.1 Combinational Circuits
A logic circuit whose output at any instant of time

The classification of digital circuitsis shown in Fig. 1.8. depends only on the present input is called combina-
tional circuit. It contains no memory element.
Digital circuits
1.4.1.1 Half Adder
Combinational circuits Sequential circuits
A half adder is a combinational circuit for the addi-
tion of two 1-bit numbers. X and Y are inputs and
outputs are sum (S) and carry (C) (Fig. 1.9 and
Synchronous Asynchronous Table 1.4).
Figure 1.8 | Types of digital circuits.
Chapter 1.indd 11 4/9/2015 9:40:21 AM

X′
Y′
X
Sum (S)
Y Cin
X´
Carry (C) Y
C′in
Sum (S)
Figure 1.9 | Logic diagram of a half adder. X
Y′
C′in
Table 1.4 | Truth table X
Inputs Outputs Y
Cin
X Y Carry (C ) Sum (S )
0 0 0 0 X
0 1 0 1 Y
1 0 0 1 Y
Cin Carry (C)
1 1 1 0
X
The sum and carry are as follows: Cin
S = X ′ Y + XY ′ Figure 1.10 | Logic diagram of a full adder.

C = XY
1.4.1.3 Half Subtractor
1.4.1.2 Full Adder
A half subtractor is a combinational circuit that subtracts
A full adder is a combinational circuit that performs 2 bits (minuend and subtrahend) and produces their dif-
the addition of 3 bits (2 significant bits and previous ferences and borrow as output (Fig. 1.11 and Table 1.6).
carry). It consists of three inputs and two outputs,
two input bits that are to be added, the third input X
is the carry from the previous position (Fig. 1.10 and Difference (D)
Y
Table 1.5).
Table 1.5 | Truth table

Borrow (B)
Inputs Outputs Figure 1.11 | Logic diagram of a half subtractor.

X Y Cin Sum (S) Cout
0 0 0 0 0 Table 1.6 | Truth table
0 0 1 1 0 Inputs Outputs
0 1 0 1 0 X Y Difference (D) Borrow (B)
0 1 1 0 1
0 0 0 0
1 0 0 1 0
0 1 1 1
1 0 1 0 1
1 0 1 0
1 1 0 0 1
1 1 1 1 1 1 1 0 0
The difference and borrow are as follows:

The sum and carry are as follows:
S = X ′Y ′Cin + X ′YCin
′ + XYCin D = XY ′ + X ′Y
Cout = ACin + XY + YCin B = XY
Chapter 1.indd 12 4/9/2015 9:40:22 AM

X´
Y´
Bin X´
X´ Y
Y
Bín Y
Difference (D)
Bin Borrow
X
Y´ (Bout)
Bín X´
Bin
X
Y
Bin
Figure 1.12 | Logic diagram of a full subtractor.
1.4.1.4 Full Subtractor adders and 1 OR gate, so n-bit parallel adder can be made
from (2n − 1) half adders and (n − 1) OR gates.
A full subtractor is a combinational circuit that per-
forms subtraction involving 3 bits (minuend bit, sub- Y3 X3 Y2 X2 Y1 X 1 Y0 X0
trahend bit and borrow from previous stage) (Table 1.7
and Fig. 1.12).
Table 1.7 | Truth table Full adder Full adder Full adder Half adder
Inputs Outputs
X Y Bin Difference (D) Bout
Cout S3 S2 S1 S0
0 0 0 0 0
0 0 1 1 1 Figure 1.13 | Block diagram of a parallel adder.
0 1 0 1 1
1.4.1.6 Look-Ahead Carry Header
0 1 1 0 1
1 0 0 1 0 In a parallel adder, carry propagation delay is present.
1 0 1 0 0 To overcome this difficulty, look-ahead carry adder is
1 1 0 0 0 used. It uses logic gates to look at lower-order bits of
augend and addend to check if higher-order carry is gen-
1 1 1 1 1 erated or not. It uses two functions, carry generate and
carry propagate, for the same (Fig. 1.14).
The difference and borrow are as follows:
D = X ′Y ′Bin + X ′YBin
′ + XY ′Bin
′ + XYBin Problem 1.13: How many AND and OR gates are
required to design 7-bit look-ahead carry adder.
B = X ′Y + X ′YBin + YBin
Solution: To design a 7-bit look-ahead carry adder,
we need
1.4.1.5 Parallel Adder (Ripple Carry Adder)
n(n + 1) 7(7 + 1)
A parallel adder consists of full adders connected in AND gate = = = 28
2 2
cascade, that is, carry output of each adder is connected to
OR gate = n = 7
the carry input of the next higher-order adder (Fig. 1.13).
An n-bit parallel adder is constructed using (n − 1) full
where n is the number of bits.
adders and 1 half adder. As full adder consists of 2 half
Chapter 1.indd 13 4/9/2015 9:40:23 AM

Ai Table 1.8 | Truth table of 2X4 decoder

Bi Pi
Ci Si Inputs Outputs
A B Y3 Y2 Y1 Y0
Gi
0 0 0 0 0 1
Ci + 1 0 1 0 0 1 0
1 0 0 1 0 0
Full Adder circuit with Pi and Gi
1 1 1 0 0 0
Carry propagation Pi = Ai + Bi
1.4.1.8 Encoder
Carry generation Gi = AiBi
Output sum Si = Pi + Ci An encoder is a combinational circuit that has 2n (or
fewer) inputs and n output lines (Fig. 1.17). It accepts
Output carry Ci+1 = Gi + PiCi
an active level on one of its inputs representing a decimal
Figure 1.14 | Logic diagram of a look-ahead adder. or octal digit and converts it to a coded output.
.
1.4.1.7 Decoder . .
2n input . 2n : n . n output
. .
lines . Encoder . lines
A decoder is a combinational circuit that converts
binary information from n input lines to a maximum of
2n unique output lines (Fig. 1.15). The size of a decoder
is n to m where m ≤ 2n. It is used to route input data to Figure 1.17 | Block diagram of an encoder.
a specified output line for memory addressing (Fig. 1.16
and Table 1.8). 1.4.1.9 Multiplexer
A multiplexer is a combinational circuit that provides

. inputs one by one at single output; and which input will
.
n input . Decoder . be out at any instant depends upon the combination at
. (n : 2n) . 2n output select lines (Fig. 1.18). It converts parallel data to serial
lines
. .
.
lines
. data (Fig. 1.19 and Table 1.9).
..
2n inputs . MUX Output lines
Enable inputs .
Figure 1.15 | Block diagram of a decoder.
n select lines
A B Enable Figure 1.18 | Block diagram of a multiplexer.
Table 1.9 | Truth table of 4 to 1 MUX

Y0
Select Lines Outputs
Y1 S1 S0 Y
Y2 0 0 I0
0 1 I1
Y3
1 0 I2
Figure 1.16 | Logic diagram of a 2X4 decoder.

1 1 I3
Chapter 1.indd 14 4/9/2015 9:40:25 AM

The Boolean function is 1.4.1.10 Boolean Function Implementation

Using Multiplexers
Y = I 0S0′ S1′ + I1S0S1′ + I 0S0′ S1 + I 0S0S1
1. Take one variable for input lines and others for
selection lines.
S1 S0 2. Write the minterms with the variable in comple-
mented form in the first row and with variable
selected in uncomplemented form in the second row.
3. Then encircle the minterms which are present in
I0 the function.
••If there is circled variable in the column, then we
I1 put 0 on the corresponding line.
Y ••If there are circled variables, then we put 1 on
I2 the line.
••If bottom variable is circled and top is not cir-
I3 cled, A is used in input line.
••If bottom variable is not circled and top is cir-
Figure 1.19 | Logic diagram of 4 to 1 MUX.
cled, A′ is used as input to that line.
Problem 1.14: Implement the function F(A, B, C, D) = Σm(1, 2, 5, 7, 9, 14) using MUX.
Solution: Select variable A for input as B, C and D for selection lines.
As there are four variables so size of MUX is 2N − 1 = 23 = 8 to 1 MUX.
1. Minterms with A′ are 0 - 7.

2. Minterms with A are 8 - 15.
I0 I1 I2 I3 I4 I5 I6 I7
A′ 0 1 2 3 4 5 6 7
A 8 9 10 11 12 13 14 25
0 1 A′ 0 0 A′ A A′
0 0
1 1
A´
2
0 3 8 to 1 Y
0 4 MUX
A´
5
A 6
A´ 7
B C D
Logic diagram
Chapter 1.indd 15 4/9/2015 9:40:26 AM

Problem 1.15: The combinatorial circuit below is Example 1.4

equivalent to ______. Design 8 × 256 decoders using 4 × 64 decoders
256/64 = 4
A I0 4/64 = 0
f(A,B,C) Total = 4
1 G: 1
1 2 MUX
0 3 Example 1.5
Design 6 × 64 decoders using 2 × 4 decoders
S1 S0 64/4 = 16
16/4 = 4
Solution:
4/4 = 1
I0 I1 I2 I3 Total = 21
A′ 0 1 2 3
A 4 5 6 7
1.4.2 Sequential Circuits
A A′ 1 0
A logic circuit whose output at any instant of time
It is equivalent to ∑m(1, 2, 4, 6). depends on the present input as well as past output is
known as sequential circuit. It contains memory element.
1.4.1.11 Demultiplexer 1.4.2.1 Flip−Flops
A demultiplexer is a combinational circuit that receives A flip−flop or bistable multivibrator is a synchronous

information on a single line and transmits this on one of 2n bistable circuit made up of logic gates that can exist in
of possible output line (Fig. 1.20). The selection of possi- either of two stable states in presence of a falling or rais-
ble output line is controlled by bit values of n select lines. ing edge of a clock signal. Flip−flops can change its state
by means of some external signal. It can “store” binary
information. There are different types of flip−flops based
on the manner the inputs affect their output state:
One
....
input
DEMUX 2n output lines 1. RS flip−flop
2. D flip−flop
3. JK flip−flop
....
4. T flip−flop
n select lines
Figure 1.20 | Block diagram of a demultiplexer. RS Flip−Flop (Direct Coupled)
This sequential logic circuit has two stable states
Implementation of higher order decoders and multiplex- (Q = 1 set state and Q = 0 reset state) achieved by
ers by using lower order. Whenever odd number imple- giving proper inputs to R and S inputs (Figs. 1.21 and
mentation is there, one NOT gate is required. 1.22 and Table 1.10). If the circuit is in one particular
state it continues to remain in that state and can store
1-bit data.
Example 1.3
Design 4 ×16 decoders using 2 × 4 decoders
S Q
16/4 = 4 RS
flip—flop
4/4 = 1 R Q´
Total = 5
Figure 1.21 | Block diagram of an RS flip−flop.
Chapter 1.indd 16 4/9/2015 9:40:27 AM

Clocked RS Flip−Flop
S
G3 G1 S
(Set) This circuit sets or resets the memory cell in synchro-
nization with the clock pulse (Figs. 1.23 and 1.24 and
Table 1.12). If clock pulse is not present, the gates G3
and G4 are inhibited; their outputs are 1.
R G4 G2 Q´
(Reset) S Q
SR
Figure 1.22 | Logic diagram of an RS flip−flop.
CLK
flip−flop
R Q´
Table 1.10 | Truth table of RS flip−flop
Figure 1.23 | Block diagram of a clocked SR flip−flop.
Inputs Outputs
Sn Rn Qn+1 Table 1.12 | Truth table of SR flip−flop
0 0 Qn (No change)
Clock Flip−Flop Inputs Outputs
0 1 0 (Reset)
CLK Sn Rn Qn+1
1 0 1 (Set)
1 0 0 Qn
1 1 Prohibited
1 0 1 0
1 1 0 1
The characteristic equation which gives the algebraic 1 1 1 Prohibited
description of the next state of a flip−flop obtained by 0 X X Qn
K-map simplification is as follows (see Table 1.11):
Qn +1 = S + R ′Qn
S G3
Table 1.11 | Truth table
G1 Q
Present Next CLK

Flip−Flop Inputs
State State
R G4 G2 Q´
Rn Sn Qn Qn+1
0 0 0 0 Figure 1.24 | Logic diagram of a clocked SR flip−flop.
0 0 1 1
0 1 0 1
0 1 1 1 D Flip−Flop
1 0 0 0
1 0 1 0 This circuit has only one input terminal, and output is
followed by input. It has delay of exactly one clock cycle
1 1 0 X
(Figs. 1.25 and 1.26 and Table 1.13). The input to R is
1 1 1 X
through an inverter from S.
S Qn
J or S
R 00 01 11 10 D Q
D
1 1 1 CLK
flip−flop
0 1 3 2
Q´
X X K or R
4 5 7 6 JK or SR Flip−Flop converted to D
Figure 1.25 | Block diagram of a D flip−flop.
Qn+1 = S + R¢Qn
Chapter 1.indd 17 4/9/2015 9:40:29 AM

Table 1.13 | Truth table of D flip−flop Table 1.14 | Truth table of JK flip−flop
Flip−F lop Present Next Flip−Flop Present Next
Inputs State State Inputs State State
Dn Qn Qn+1 Jn Kn Qn Qn+1
0 0 0 0 0 0 0
0 1 0 0 0 1 1
1 0 1 0 1 0 0
1 1 1 0 1 1 0
1 0 0 1
D S 1 0 1 1
Q
1 1 0 1
CLK 1 1 1 0
J G3 G1 Q
Q´
R Q´
Figure 1.26 | Logic diagram of a D flip−flop.

CLK
Q
G4 G2 Q´
The characteristic equation is Qn+1 = D (Fig. 1.27) K
Qn 0 Figure 1.29 | Logic diagram of a JK flip−flop.

1
The characteristic equation is Qn +1 = JQn′ + KQn
D
0 0 1
(Fig. 1.30).
1 1 D
1 2 3 K Qn
Qn+1 = D J 00 01 11 10
Figure 1.27 | K-Map 1 1

0 0 1 3 2
JK Flip−Flop
The uncertainty in SR flip−flop when Sn = Rn = 1 can 1
1 1
be avoided by using JK flip−flop (Figs. 1.28 and 1.29 4 5 7 6
and Table 1.14). The output of JK flip−flop is con- T
nected back to the inputs of gate. The inputs J and K Qn+1 = K´Qn + JQń
are ANDed with Q and Q′ to obtain S and R inputs:
Figure 1.30 | K-Map
S = JQ′ Repetition of toggle for a single-clock pulse is known
R = J ⋅Q as race-around problem. The output of JK flip−flop
will oscillate between 0 and 1 during a single-clock pulse
The flip−flop sets when input J is high and resets when so that at the end of clock pulse the value of Q is uncer-
input K is high. If J = K = 1, the flip−flop will toggle on tain. This problem can be avoided by decreasing the
the next positive clock. pulse width, by edge triggering or increasing propaga-
tion delay.
J Q
Master−Slave JK Flip−flop
CLK JK It is a cascade of two RS flip−flops (master and slave)
flip−flop with feedback from outputs of slave to input of master
K Q´ to avoid race-around problem. Positive clock pulses are
applied to first flip−flop and pulses are inverted before
Figure 1.28 | Block diagram of a JK flip−flop. applying to slave flip−flop (Fig. 1.31).
Chapter 1.indd 18 4/9/2015 9:40:31 AM

Qm SS
J G3m G1m G3s G1s Q
CLK
G4m G2m G4s G2s Q′

K RS
Qm
Master FF
Slave FF
CLK
Figure 1.31 | Block diagram of a master−slave JK flip−flop.
T Flip−flop
It is also known as toggle flip−flop, which is obtained Qn 0 1
by connecting both inputs of JK flip−flop (Fig. 1.32 and T
Table 1.15). 1
0 0 1
J
T Q
1
JK (T) 1 2 3
CLK
flip−flop
Q´ Qn+1 = T´Qn + T Qń
K
Figure 1.32 | Block diagram of a T flip−flop. Figure 1.33 | K-Map
Table 1.15 | Truth table of T flip−flop

Flip−Flop Inputs Present State Next State Excitation Table of Flip−Flop
The excitation table of flip−flop states is given in
Tn Qn Qn+1
Table 1.16.
0 0 0
0 1 1
1 0 1 Conversion Equation
1 1 0 The flip−flop conversion equations are given in Table 1.17.
Table 1.16 | Excitation Table of Flip-flop States
Flip-Flop States RS Flip-Flop Inputs JK Flip-Flop Inputs D Flip-Flop T Flip-Flop

Inputs Inputs
Present Next R S J K
State State
(Qn) (Qn +1)
0 0 X 0 0 X 0 0
0 1 0 1 1 X 1 1
1 0 1 0 X 1 0 1
1 1 0 X X 0 1 0
Chapter 1.indd 19 4/9/2015 9:40:32 AM

Shift
registers
Based Based on
on direction input−output
of data
movement mode
Shift left Shift right Bidirectional Serial-in Serial-in Parallel-in Parallel-in Universal
register register shift serial-out parallel-out serial-out parallel-out register
Figure 1.34 | Types of registers
Table 1.17 | Flip−flop conversion equations computation. There are different types of registers as
shown in Fig. 1.34.
S = JQ′
1 SR to JK
R = K ⋅Q 1. Serial-In/Serial-Out Shift Register: It accepts
data serially on single input line and produces output
S =D in serial form. Data can be shifted right or left.
2 SR to D •• Shift right register: In shift right register,
R = D′ serial data is entered from the left side of a regis-
S = TQ′ ter and leaves from the right side. For example,
3 SR to T a 4-bit number 1101 (QDQCQBQA) is entered
R = TQ serially into first flip−flop D from the left side
(Fig. 1.35).
J =S
4 JK to SR
K =R
J =D FF1 FF2 FF3 FF4
5 JK to D Din D QD D QC D QB D QA Dout
K = D′
D C B A
J =T
6 JK to T
K=T
7 D to JK D = TQ′ + K ′Q CLK
Figure 1.35 | 4-Bit shift right register.
8 D to SR D = S + R ′Q
9 D to T D = T ⊕Q The content of register after each clock pulse is

shown in Table 1.18.
10 T to SR T = SQ′ + RQ
Table 1.18 | Content of register after each clock pulse
11 T to D T =D⊕Q
Clock QD QC QB QA
12 T to JK T = JQ′ + KQ Pulse
0 0 0 0 0
Applications of Flip−Flops 1 1 0 0 0
Registers 2 0 1 0 0
Registers are collection of flip−flops, and each flip−flop 3 1 0 1 0
is capable of storing single bit of information. They make 4 1 1 0 1
stored information available to the logic elements for the
Chapter 1.indd 20 4/9/2015 9:40:35 AM

•• Shift left register: In shift left register, serial Note: A universal register can perform all the
data is entered from the right side of the register operations of shift register.
and leaves from the left side. For example, a
2. Serial-In to Parallel-Out (SIPO): The regis-
4-bit number 1101 (QDQCQBQA) is entered seri-
ter is loaded with data serially, one bit at a time,
ally into input of shift left register (Fig. 1.36).
with the stored data being available at the output
FF4 FF3 FF2 FF1 in parallel form (Fig. 1.37).
Dout Din
QD D QC D QB D The content of register after each clock pulse is
given in Table 1.20.
D C B A
Table 1.20| Content of register after each clock pulse
CLK Clock Pulse QA QB QC QD
Figure 1.36 | 4-Bit shift left register. 0 0 0 0 0
The content of register after each clock pulse is 1 1 0 0 0
given in Table 1.19.
2 0 1 0 0
Table 1.19 | Content of register after each clock pulse
3 0 0 1 0
Clock Pulse QD QC QB QA
4 0 0 0 1
0 0 0 0 0
5 0 0 0 0
1 0 0 0 1
2 0 0 1 1 3. Parallel-In to Serial-Out (PISO): The parallel
3 0 1 1 0 data is loaded simultaneously into the register and
is shifted out of the register serially, one bit at a
4 1 1 0 1
time (Fig. 1.38).
4-bit Parallel Data Output
QA QB QC QD
D Q D Q D Q D Q
Serial FFA FFB FFC FFD
Data In
CLK CLK CLK CLK
CLR CLR CLR CLR
Clear (CLR)
Clock (CLK)
Figure 1.37 | 4-Bit SIPO register.
1 Q Q
mux D Q mux D Q mux D Q mux D
FFA FFB FFC FFD Serial
Data Out
CLK CLK CLK CLK
Clock
PD PC PB PA
4-bit Parallel Data Input
Figure 1.38 | 4-Bit PISO register.
Chapter 1.indd 21 4/9/2015 9:40:36 AM

4-bit Parallel Data Output
QD QC QB QA
D Q D Q D Q D Q
FFA FFB FFC FFD
CLK CLK CLK CLK
Clock
PD PC PB PA
4-bit Parallel Data Input

Figure 1.39 | 4-Bit PIPO register.
4. Parallel-In to Parallel-Out (PIPO): The par- The counter with n flip−flops has a maximum modulus
allel data is loaded simultaneously into the register, 2n so it can count from 0 to 2n−1. It can also produce
and transferred together to their respective outputs MOD numbers less than 2n by skipping the states.
by the same clock pulse (Fig. 1.39).
1. 4-Bit Asynchronous Up Counter: Figures
1.40−1.42 show a 4-bit asynchronous counter con-
Counters
structed using JK flip−flops. The clock input of first
A counter is a sequential logic circuit that is capable
flip−flop (the one with the Q0 output) is connected
of counting the number of clock pulses arriving at its
to the external clock. The clock of second flip−flop
input in a cyclic sequence. It is just a register that shifts
drives from output Q0 of the first flip−flop and so
through a predetermined sequence of states upon the
on. Each flip-flop input (J = K = 1) is connected
application of input clock pulses. There are two differ-
to HIGH voltage so it toggles when negative edge is
ent types of counters − synchronous counter (parallel
arrived at clock input and counts from 0000 to 1111.
counter) and asynchronous counter (serial or ripple
counter) (Table 1.21). 0000
1111 0001
Table 1.21 | Differences between asynchronous and
0 1
1110 15
synchronous counter. 2 0010
14
Asynchronous Synchronous 0011
1101 13 3
Counter Counter
The output of first There is no connection
flip−flop drives the clock between output of first 1100 12 4 0100
for next flip−flop. So all flip−flop and clock input
flip−flops are not clocked of next flip−flop. All the 5 0101
simultaneously. flip−flops are clocked 1011 11
simultaneously.
10 6
Speed is low. Speed is high. 0110
1010
9 7
Simple logic circuit as Complex design as number 8
number of states increases. of states increases. 1001 0111
1000
Figure 1.40 | State diagram of a 4-bit asynchronous
Example: Ring counter, Example: Binary ripple
Johnson counter. counter, Up-down counter.
counter.
Chapter 1.indd 22 4/9/2015 9:40:37 AM

Vcc = 5V
J0 Q0 J1 Q1 J2 Q2 J3 Q3
CLK
K0 Q´0 K1 Q´1 K2 Q´2 K3 Q´3
Figure 1.41 | Logic diagram for a 4-bit asynchronous counter.
Count
QA 0 1 0 1 0 1 0 1 0 1 0 1 0 1 0 1 0
QB 0 0 1 1 0 0 1 1 0 0 1 1 0 0 1 1 0
QC 0 0 0 0 1 1 1 1 0 0 0 0 1 1 1 1 0
QD 0 0 0 0 0 0 0 0 1 1 1 1 1 1 1 1 0
Figure 1.42 | Timing diagram for a 4-bit asynchronous counter.
2. MOD-10 Asynchronous Counter: A MOD-10 otherwise it maintains the previous state. The flip−
asynchronous counter has 10 states and number of flop 2 toggles when both Q1 and Q0 are 1. The flip−
flip−flops required are 24 ≥ 10 (Figs. 1.43 and 1.44). flop 3 toggles when Q2, Q1 and Q0 are 1 (Table 1.22).
It counts the sequence from 0000 to 1001 and after
that it returns to the initial state. Table 1.22 | Truth table for 4-bit synchronous counter
0000 Count Q3 Q2 Q1 Q0
0001
15 0 1 0 0 0 0 0
2 0010
14 1 0 0 0 1
3 0011 2 0 0 1 0
13
3 0 0 1 1
12 4 0100 4 0 1 0 0
5 0101 5 0 1 0 1
11
6 0 1 1 0
10 6
0110 7 0 1 1 1
1010
9 7 8 1 0 0 0
8
1001 0111
1000 9 1 0 0 1
Figure 1.43 | State diagram of a 4-bit MOD-10 counter. 10 1 0 1 0
3. 4-Bit Synchronous Counter: In a 4-bit synchro- 11 1 0 1 1
nous counter, the least significant bit of flip−flop is 12 1 1 0 0
connected at high level and other inputs of JK flip−
13 1 1 0 1
flop are driven by some combination of flip−flop out-
puts. Figure 1.45 shows the logic diagram of a 4-bit 14 1 1 0 0
synchronous counter. The output of flip−flop0 always 15 1 1 0 1
toggle, and that of flip−flop1 toggles when Q0 is 1,
Chapter 1.indd 23 4/9/2015 9:40:38 AM

Vcc
0 1 0 1
J0 Q0 J1 Q1 J2 Q2 J3 Q3
CLK
K0 Q´0 K1 Q´1 K2 Q´2 K3 Q´3

CLR CLR CLR CLR
1
Figure 1.44 | Logic diagram for a MOD-10 asynchronous counter.
Q0 Q1 Q2 Q3
Vdd
J Q J Q J Q J Q
FF 0 FF 1 FF 2 FF 3
C C C C
K Q´ K Q´ K Q´ K Q´
This flip−flop This flip−flop This flip−flop This flip−flop

toggles on every toggles only if toggles only if toggles only if
clock pulse Q0 is “high” Q0 AND Q1 Q0 AND Q1 AND
are “high” Q2 are “high”
Figure 1.45 | Logic diagram of a 4-bit synchronous counter.
4. MOD-5 Synchronous Counter: Table 1.23 | Excitation table for JK flip−flop
Step 1: Number of states is five, so counting sequence
0 to 4. Flip−Flop States JK Flip−Flop Inputs
Flip−flops required are 2n ≥ 5, so n = 3. Present Next J K
Let us label these flip−flops as A, B and C. State State
Step 2: Obtain the excitation table for JK flip−flop (Qn) (Qn+1)
(Table 1.23). 0 0 0 X
Step 3: Develop state table using the excitation table
(Table 1.24). 0 1 1 X
Step 4: K-map simplification for finding expressions 1 0 X 1
corresponding to each input (Fig. 1.46). 1 1 X 0
Chapter 1.indd 24 4/9/2015 9:40:39 AM

Table 1.24 | State table

Present State Next State Flip−Flop Inputs
QC QB QA QC+1 QB+1 QA+1 JC KC JB KB JA KA
0 0 0 0 0 1 0 X 0 X 1 X
0 0 1 0 1 0 0 X 1 X X 1
0 1 0 0 1 1 0 X X 0 1 X
0 1 1 1 0 0 1 X X 1 X 1
1 0 0 0 0 0 X 1 0 X 0 X
1 0 1 X X X X X X X X X
1 1 0 X X X X X X X X X
1 1 1 X X X X X X X X X
Expression for JC Expression for KC

QB Q A QBQA
QC 00 01 11 10 QC 00 01 11 10
0 1 0 X X X X
0 1 3 2 0 1 3 2
1 X X X X 1 1 X X X
4 5 7 6 4 5 7 6
J C = Q BQ A KC = 1
Expression for JB Expression for KB

QB QA QBQA
QC 00 01 11 10 QC 00 01 11 10
0 1 X X 0 X X 1
0 1 3 2 0 1 3 2
1 X X X 1 X X X X
4 5 7 6 4 5 7 6
J B = QB KB = QA
Expression for JA Expression for KA

QB QA QBQA
QC 00 01 11 10 QC 00 01 11 10
0 1 X X 1 0 X 1 1 X
0 1 3 2 0 1 3 2
1 X X X 1 X X X X
4 5 7 6 4 5 7 6
JA = QĆ KA = 1
Figure 1.46 | K-map simplification.
Chapter 1.indd 25 4/9/2015 9:40:39 AM

Step 5: Logic diagram of MOD-5 counter (Fig. 1.47). Table 1.25 | Continued
CLK CLR′ Q3(MSB) Q2 Q1 Q0

1 1 0 0 0 1
1 1 0 0 1 0
J A QA J B QB JC QC
1 1 0 1 0 0
1 1 1 0 0 0
KA QÁ KB Q´B KC QĆ
7. Johnson counter: A Johnson counter (switch-
High High

CLK tail ring counter, twisted-ring counter or walking-
Figure 1.47 | Logic diagram of a MOD-5 counter. ring counter) is a variation of shift register counter,
where the complement of the output of the last
flip−flop is fed back as input to the first flip−flop
5. Shift register counter: It is basically a shift reg-
(Fig. 1.49 and Table 1.26). The counter has count-
ister whose serial output is connected back to serial
ing sequence length of twice the length of the shift
input in order to produce special sequences.
register, that is, number of flip−flops.
6. Ring counter: A ring counter is a circular shift
register in which output of last flip−flop is con- Q0 Q1 Q2 Q3
nected to the input of the first in a ring (Fig.
1.48 and Table 1.25). In this, only one flip−flop
is in state one while others are in their zero
states. Each state repeats after every n clock D3 Q3
D 0 Q0 D 1 Q1 D2 Q2
cycles, where n is the number of flip−flops used. FF4
Initially the counter is reset by applying low FF1 FF2 FF3 Q´3
signal to active low CLR input. All flip−flops
reset except the first flip−flop. CLK
Figure 1.49 | 4-Bit Johnson counter.
Q0 Q1 Q2 Q3
Table 1.26 | Truth table of 4-bit Johnson counter
LSB MSB CLK Q0 Q1 Q2 Q3 Q3′
PRE
D 0 Q0 D 1 Q1 D 2 Q2 D 3 Q3 0 0 0 0 0 1
0FF 1FF 2FF 3FF 1 1 0 0 0 1
2 1 1 0 0 1
CLK
3 1 1 1 0 1
CLR´
4 1 1 1 1 0
Figure 1.48 | Logic diagram of a 4-bit ring counter.
5 0 1 1 1 0
Table 1.25 | Truth table of 4-bit ring counter

6 0 0 1 1 0
7 0 0 0 1 0
CLK CLR′ Q3(MSB) Q2 Q1 Q0
8 0 0 0 0 1
0 0 0 0 0 1
1 1 0 0 1 0
1.5 DIGITAL LOGIC FAMILIES
1 1 0 1 0 0
1 1 1 0 0 0
Refer Fig. 1.50 and Table 1.27 for classification and pros
(Continued) and cons of logic families.
Chapter 1.indd 26 4/9/2015 9:40:40 AM

1.5 DIGITAL LOGIC FAMILIES 27
Logic families
Bipolar Unipolar
Saturated Non- saturated PMOS
RTL (resistor ECL (emitter coupled

transistor logic) logic) NMOS
DTL (diode transistor

logic) Schottky TTL CMOS
HTL (high threshold

logic)
TTL (transistor
transistor logic)
I2L (integrated injection

logic)
Figure 1.50 | Digital logic families.
Table 1.27 | Pros and cons of logic families

Logic Families Pros Cons
RTL Economical and simple to design, can Low threshold and fan out, poor noise
easily interface with other logic families immunity, low speed and low power dissipation
DTL High speed with 30-ns propagation delay, Poor noise immunity and fan out
low power dissipation
TTL Fastest saturating logic family, high- Power dissipation is more than MOS gates,
speed Schottky, low-power Schottky, noise immunity is not very high
noise margin about 0.4 V, more fan out,
can drive up to 10 gates
ECL Highest speed because transistors operate Very low noise margin, highest power
in active region, propagation delay of dissipation, high cost
2-ns, parameters resistant to temperature
change
NMOS Very low power dissipation, large fan out High power dissipation than CMOS, low speed
up to 20, very high noise margin of operation, large propagation delay per gate
I2L Composed of BJT so high speed of Lower packing density than NMOS, lower noise
operation, low power dissipation, low margin, external resistance required for proper
power supply requirement functioning
CMOS Large fan out (>50), lowest power Increased cost, less packing density than NMOS
dissipation, very high noise immunity,
temperature resistant
Chapter 1.indd 27 4/9/2015 9:40:40 AM

IMPORTANT FORMULAS
1. Null − All the outputs are zero. 28. X⋅X = X

2. Identity − All the outputs are one. 29. X⋅0 = 0
3. X inhibit Y - Represented as (x ⋅ ∼y) 30. X⋅Y = Y⋅X
4. Y inhibit X - Represented as (y ⋅ ∼x) 31. X⋅(Y⋅Z) = (X⋅Y)⋅Z
5. Transfer - Same as y 32. X + (Y⋅Z) = (X + Y)⋅(X + Z)
6. AND - x ⋅ y (Similar to intersection in set theory) 33. (X⋅Y)′ = X′ + Y′
7. OR - x + y (Similar to union in set theory) 34. X⋅(X + Y) = X
8. NOT X− ∼x (Also known as complement of x)
9. NOT Y− ∼y (Also known as complement of y) Half Adder:
10. XOR − (∼x ⋅ y) + (x ⋅ ∼y) (Also known as parity

35. S = X′Y + XY′
checker) 36. C = XY
11. XNOR − (x ⋅ y) + (∼x ⋅ ∼y)
Full Adder:
12. NAND − ∼(x ⋅ y)
′ + XY ′Cin
37. S = X ′Y ′Cin + X ′YCin ′ + XYCin
13. NOR − ∼(x + y)
38. Cout = ACin+ XY + YCin
14. X implication Y (∼x + y)
15. Y implication X (∼y + x) Half Subtractor:
16. X + 0 = X 39. D = XY′ + X′Y
17. X + (∼X) = 1 40. B = X′Y
18. X + X = X
Full Subtractor:
19. X + 1 = 1
′ + XY ′Bin
41. D = X ′Y ′Bin + X ′YBin ′ + XYBin
20. X + Y = Y + X
42. B = X′Y + X′YBin + YBin
21. X + (Y + Z) = (X + Y) + Z
22. X ⋅ (Y + Z) = X ⋅Y + X⋅Z Characteristic Equations:
23. ∼(X + Y) = ∼X⋅∼Y (Also called DeMorgan’s law) 43. SR flip-flop: Qn+1= S + R′Qn
24. X + (X ⋅Y) = X 44. D flip-flop: Qn+1= D
25. ∼(∼X) = X
45. JK flip-flop: Qn +1 = JQn′ + KQn
46. T flip-flop: Qn +1 = TQn′ + T ′Qn
Duality Theorem:
26. X ⋅ 1 = X
27. (X⋅∼X) = X⋅X′ = 0
SOLVED EXAMPLES
1. A correct output is achieved from a master-slave Solution: The clock is high, only the flip-flops
JK flip-flop only if its inputs are stable while the are stable when the inputs are stable.
Ans. (d)
(a) Clock is LOW
(b) lave is ready to receive the input from master 2. An inverter circuit can be realized with how many
(c) Master gives a reset signal to slave NAND gates?
(d) Clock is HIGH
Chapter 1.indd 28 4/9/2015 9:40:41 AM

SOLVED EXAMPLES 29
(a) 1 (b) 2 The SR (Set-Reset) flip-flop is a simple sequential

(c) 3 (d) Not possible circuit which consists of two gates connected as
shown above. The output of each gate is connected
Solution: An inverter or logic NOT gate can be to one of the inputs of the other gate, thus giving
made using standard NAND gate by connecting positive feedback or cross-coupling. The circuit has
together ALL their inputs to a common input two active low inputs NOT S and NOT R and two
signal, for example, outputs Q and NOT Q. When there is 0 on the S
input and a 1 on the R input, then NOT S is 0 and
NOT R is 1, so Q is 0 and NOT Q is 1. Thus, Q is
A B reset to 0 by 0 on NOT R.
Ans. (b)
Two-input NAND gate equivalent 6. The NAND or NOR gates are referred to as
“universal” gates because
Ans. (a)
(a) they are easier to implement.
3. A modulus-12 ring counter requires a minimum of
________.
(b) they can be used to build all the other types
of gates.
(a) 10 flip-flops (b) 12 flip-flops (c) they are available in different IC families.
(c) 8 flip-flops (d) 6 flip-flops (d) they are available everywhere in the world.
Solution: In the ring counter, n flip-flops are used Solution: NAND and NOR gates are univer-
for the mod-n. 12 flip-flops because ring counter sal because all Boolean functions that can be
uses states n, so here modulus is 12. implemented by AND, OR and NOT gates can be
Ans. (b) implemented by NAND and NOR gates.
4. How is a strobe signal used when serially loading a Ans. (b)
shift register? 7. Match the following:
(a) To turn the register set and reset
(b) To control the number of clocks IC Family GATE
(c) To determine which output has a HIGH value TTL NOR
(d) To cascade it with other shift register CMOS NOT
Solution: In digital circuits, a shift register is a ECL NAND
cascade of flip-flops, sharing the same clock each
time. This is called clocking or strobing to the (a) TTL:NAND; ECL:NOR; CMOS:NOT
register to set the clocks. (b) TTL:NAND; ECL:NOT; CMOS:NOR
Ans. (b) (c) TTL:NOT; ECL:NAND; CMOS:NOR
5. An active-HIGH SR latch has a 0 on the S input (d) TTL:NOR; ECL:NOT; CMOS:NAND
and a 1 on the R input. Now when the R input goes Solution: The NAND function is actually the
to 0, the latch will be ________. simplest, most natural mode of operation for this
(a) SET (b) RESET TTL design. The basic gates are OR/NOR gates
(c) CLR (d) STROBE for ECL Construction. Inverters can be con-
structed using two complementary transistors in
Solution: a CMOS configuration.
Ans. (a)
S′
Q 8. Which of the following satisfies the commutative
law but not the associative law?
(a) NOR (b) OR
(c) XOR (d) NAND
Solution
Let ↑ signify the NAND operation. Then there
Q′ exist propositions p, q, r such that
R′ p↑ (q↑r) (p↑q) ↑r
Chapter 1.indd 29 4/9/2015 9:40:42 AM

This implies NAND is not associative. 12. A sequential circuit gives an output depending upon
p↑q = q↑p (a) its present state inputs only.
This implies NAND is commutative. (b) its present-1 state input only.
Ans. (d) (c) both present- and past-state input.
9. Match the following: (d) the size of the memory.
IC Family Characteristics Solution: Sequential logic is a logic circuit in

which the output depends not only on the present
X. TTL High fan-out value of its input signals but on their past history.
Y. ECL Low propagation delay This is in contrast to combinational logic, where
Z. CMOS High power dissipation the output depends only on the present input.
Ans. (c)
Code:
XY Z 13. Synchronization is achieved by a timing device
(a) 3 2 1 called a ________.
(b) 3 1 2
(c) 2 1 3 (a) Strobe
(d) 1 2 3 (b) Reset
(c) CLK
Solution: The power dissipation is dependent on (d) Master clock generator
the power supply voltage, frequency, output load,
and input rise time. Power dissipation for TTL Solution: Synchronization is achieved by a timing
gates is usually 10 mW per gate. ECL is the fastest device called a clock generator that produces a
logic gate and has least propagation delay of 2 ns. periodic train of clock pulses.
CMOS gates have high fan-outs. Ans. (d)
Ans. (a)
14. A complete microcomputer system contains
10. Which of the following statements is false?
(a) Microprocessor
A. The 2’s complement of 0 is 0. (b) Memory
B. In 2’s complement, the leftmost bit cannot be (c) I/O device
used to express a quantity. (d) All of the above
C. For an n-bit word (2’s complement) which
includes the sign bit, there are 2n - 1 positive Solution: A microcomputer contains a micro-
integers, 2n + 1 negative integers and one 0 for processor, ROM and RAM, I/O ports and bus
a total of 2n unique states. or system of interconnecting wires, housed in a
D. In 2’s complement, the significant information is motherboard.
contained in the 1’s of positive numbers and 0’s Ans. (d)
of the negative numbers.
15. How many different Boolean functions of degree 4
(a) A (b) B (c) C (d) D are there?
Solution: For an n-bit word which includes the sign (a) 24 (b) 28
bit, there are 2n-1 - 1 positive integers, 2n-1 nega- (c) 212 (d) 216
tive integers and one 0, for a total of 2n unique states.
Ans. (c) Solution: Different Boolean functions of degree n
n
is given by 22 .
11. The dual of any Boolean expression is obtained by Ans. (d)
(a) Interchanging POS and SOP terms
(b) Interchanging 0’s and 1’s 16. A Boolean operator ⊕ is defined as follows:
(c) Interchanging Boolean sums and Boolean 1 ⊕ 1 = 1, 1 ⊕ 0 = 0, 0 ⊕ 1 = 0 and 0 ⊕ 0 = 1
products and also interchanging 0’s and 1’s What will be the truth value of the expression
(d) Interchanging Boolean sums and Boolean products (x ⊕ y) ⊕ z = x ⊕ (y ⊕ z)?
Solution: The dual of a Boolean expression is (a) Cannot be defined, as ⊕ is defined for two
obtained by interchanging Boolean sums and prod- inputs.
ucts and interchanging 0’s and 1’s. For example, (b) Always true.
The dual of x(y + 0) = x + (y × 1). (c) Always false.
Ans. (c) (d) True when any one of the inputs is true.
Chapter 1.indd 30 4/9/2015 9:40:43 AM

SOLVED EXAMPLES 31
Solution: f1 f
f2
x y z x ⊕ y (x ⊕ y)⊕ z y ⊕ z x ⊕(y ⊕ z)
f3 = ?
0 0 0 1 0 1 0
f1(w, x, y, z) = ∑ (8, 9, 10)
0 0 1 1 1 0 1 f2(w, x, y, z) = ∑ (7, 8, 12, 13, 14, 15)
0 1 0 0 1 0 1 f(w, x, y, z) = ∑ (8, 9)
0 1 1 0 0 1 0 The function f3 is
1 0 0 0 1 1 1 (a) ∑ 9, 10 (b) ∑ 9
1 0 1 0 0 0 0 (c) ∑ 1, 8, 9 (d) ∑ 8, 10, 18
1 1 0 1 0 0 0 Solution: By the logic circuit, function f is given as
1 1 1 1 1 1 1 f = f1f2 + f3
f1 = ∑ (8, 9, 10)
f2 = ∑ (7, 8, 12, 13, 14, 15)
Ans. (b)
17. Let * be defined as x * y = x + y . Let z = x*y, As according to logic, common minterm in f1 and
then value of z*x is f2 is 8, so
(a) x + y (b) x (c) 0 (d) 1 f1f2 = ∑ (8)
Solution: Substituting value of z = x*y in the to get output as f = ∑ (8, 9), f3 can be either
expression: ∑ (8, 9) or f3 = ∑ 9
Ans. (b)
z ∗ x = (x ∗ y) ∗ x = (x + y) * x = (x + y) + x
20. Consider the circuit shown in the following figure;
= x⋅y +x f implements
Now using absorption law: a + ab = a
0
x⋅y +x = x
C
4 to 1
C 1
So, result is: z*x = x Multiplexer f
Ans. (b) C 2
18. Let f (x, y, z) = x + yz + xz be a switching func- C 3 A B

tion. Which one of the following is valid? S1 S2
(a) yx is a prime implicant of f.

(b) xz is a minterm of f.
(c) xz is an implicant of f.
(a) ABC + ABC + ABC (b) A + B + C
(d) y is a prime implicant of f.
(c) A ⊕ B ⊕ C (d) C
Solution: Solution: The circuit for function f(A, B, C) is
yz given as follows
00 01 11 10
x
0 1 1 1 1 C I0 4 × 1
C I1 MUX
1 1 1
C I2 F (A,B,C)
Number of possible joins are 2, so there are two C I3

S1 S0
prime implicants, that is, x′ and z.
Both the prime implicants are essential. xz is an
implicant of f. A B
Ans. (c) The output of function will be
19. Consider the logic circuit shown in the following ABC++ABC

ABC ABC++ABC
ABC++ABC
ABC ==AC
AC((BB++BB))++AC
AC((BB++BB))
figure. The functions f1, f2 and f (in canonical sum- ==AC
AC++AC
AC ==CC
of-products form in decimal notation) are Ans. (d)
Chapter 1.indd 31 4/9/2015 9:40:45 AM

21. Which of the following functions implements the 23. Consider the circuit given below with initial state
Karnaugh map shown below? Q0 = 1, Q1 = Q2 = 0. The state of the circuit is
given by the value 4Q2 + 2Q1 + Q0.
CD
AB 00 01 11 10
00 0 0 1 0 D 0 Q0 D 1 Q1 D 2 Q2
01 X X 1 X
11 0 1 1 0 LSB MSB
CLK
10 0 1 1 0 Which one of the following is the correct state
sequence of the circuit?
(a) AB + CD (b) D(C + A)
(a) 1, 3, 4, 6, 7, 5, 2 (b) 1, 2, 5, 3, 7, 6, 4
(c) AD + AB (d) (C + D)(C + D)(A + B) (c) 1, 2, 7, 3, 5, 6, 4 (d) 1, 6, 5, 7, 2, 3, 4
Solution: Karnaugh map of the function is given Solution:
as follows:
D2(Q1) D1(Q0) D0 Q2 Q1 Q0 4Q2 +
CD 00 01 11 10 (Q1⊕Q2) 2Q1 + Q0
AB
00 0 0 1 0 Initially 0 0 1 1
0 1 0 0 1 0 2
01 X X 1 X
1 0 1 1 0 1 5
11 0 1 1 0 0 1 1 0 1 1 3
1 1 1 1 1 1 7
10 0 1 1 0
1 1 0 1 1 0 6
SOP form for the given function is f(A, B, C, D) = 1 0 0 1 0 0 4
CD + AD = D(C + A) 0 0 1 0 0 1 1
Ans. (b)
22. Refer to the following arrangement of master-slave Therefore, the correct state sequence of the circuit
flip-flops. is 1, 2, 5, 3, 7, 6, 4. Ans. (b)
P Q 24. A 32-bit floating-point number is represented by a
7-bit signed exponent, and a 24-bit fractional man-
J K D
tissa. The base of the scale factor is 16.
`1’
(a) The range of the exponent is .
CLK (b) The range of the exponent is if the scale
It has the initial state of P, Q as 0, 1 (respectively). factor is represented in excess −64 format.
After the clock cycles, the output state P, Q is Solution: The given floating point number format is
(respectively)
1 7 24
(a) 1, 0 (b) 1, 1 (c) 0, 0 (d) 0, 1
S EXP Fraction (Mantissa)
Solution: 31 30 24 23 0
J K D Bias + 64
1 1 (P) P Q The above floating-point architecture is IBM float-
Initially 0 1 ing-point architecture.
1 1 0 1 0 The number is represented as the following formula
1 1 1 0 1 (−1)S× 0. Mantissa × 10EXP - 64
(a) The range of the exponent is the range of the
For J = 1 and K = 1, the output is complement to 7-bit signed numbers, which is −64 to +63.
previous state. (b) The range of the exponent if the scale factor is
Ans. (a) represented in excess −64 format is 0 to 127.
Chapter 1.indd 32 4/9/2015 9:40:48 AM

GATE PREVIOUS YEARS’ QUESTIONS 33
25. Booth’s coding in 8-bits for the decimal number −57 is 26. The 2’ complement representation of the decimal
value −15 is
(a) 0 - 100 + 1000 (b) 0 − 100 + 100 − 1
(c) 0 − 1 + 100 - 10 + 1 (d) 00 − 10 + 100 − 1 (a) 1111 (b) 11111
(c) 111111 (d) 10001
Solution: −57 can be represented in 2’s comple-
ment form: 1 1 0 0 01 1 1 Solution: + 15 = 1 1 1 1
Booth’s coding for decimal number −57 is −15 in 2’s complement = 10001
1 1 0 0 0 1 1 1 0 Ans. (d)
⇓
0 −1 0 0 +1 0 0 −1 −1
Ans. (b)
GATE PREVIOUS YEARS’ QUESTIONS
1. Assuming all numbers are in 2’s complement Solution:

representation, which of the following numbers is Largest positive number, m = 11111111
divisible by 11111011? exponent, e = 111110
Second largest positive number = 11111110
(a) 11100111 (b) 11100100
exponent, e = 111110
Difference = 231 (2 − 2−9− 2 + 2−8) = 222
(c) 11010111 (d) 11011011
(GATE 2003: 1 Mark)
Ans. (c)
Solution: 3. Consider an array multiplier for multiplying two

n bit numbers. If each gate in the circuit has a unit
2’s Complement Binary Number Decimal Value delay, the total delay of the multiplier is
11100111 00011001 25
11100100 00011100 28 (a) Θ(1) (b) Θ(log n) (c) Θ(n) (d) Θ(n2)
11010111 00101001 41 (GATE 2003: 1 Mark)
11011011 00100101 37
Solution:
11111011 is 5 in decimal. Only 25 is divisible n(n + 1)
by 5. Delay = 1 + 2 + + n = = Q(n2 )
2
Ans. (a)
Ans.(d)
2. The following is a scheme for floating point number
representation using 16 bits. 4. A 1-input, 2-output synchronous sequential circuit
behaves as follows:
Bit Position 15 14.....9 8.....0
Let zk, nk denote the number of 0’s and 1’s, respec-
s e m tively, in initial k bits of the input (zk + nk = k).
sign exponent mantissa The circuit outputs 00 until one of the following
conditions holds.
Let s, e, and m be the numbers represented in
binary in the sign, exponent, and mantissa fields, (a) zk − nk = 2. In this case, the output at the kth
respectively. Then the floating point number rep- and all subsequent clock ticks is 10.
resented is (b) nk − zk= 2. In this case, the output at the kth
(-1)s (1 + m × 2-9 )2e-31
and all subsequent clock ticks is 01.
if the exponent ≠ 111111

 0 otherwise What is the minimum number of states required
in the state transition graph of the above circuit?
What is the maximum difference between two suc-
cessive real numbers representable in this system? (a) 5 (b) 6 (c) 7 (d) 8
(a) 2−40 (b) 2−9

(GATE 2003: 2 Marks)
(c) 222 (d) 231 Solution: Number of transition states are 7.
(GATE 2003: 2 Marks) Ans. (c)
Chapter 1.indd 33 4/9/2015 9:40:48 AM

5. The literal count of a Boolean expression is the input A, how many transition(s) (change of logic
sum of the number of times each literal appears levels) occur(s) at B during the interval from 0 to
in the expression. For example, the literal count 10 ns?
of (xy + xz′) is 4. What are the minimum possible
Logic 1
literal counts of the product-of-sum and sum-of- A Logic 0
product representations, respectively, of the func- Time 0 1 2 3 4 5 6 7 8 9 10 11
tion given by the following Karnaugh map? Here, (ns)
X denotes “don’t care”.
ZW (a) 1 (b) 2
00 01 11 10 (c) 3 (d) 4
XY
00 X 1 0 1
01 0 1 X 0
Solution: Total number of transitions is 3.
1: 0 to 0
11 1 X X 0 2: 0 to 1
3: 1 to 1
10 X 0 0 X Ans. (c)
7. A Boolean function x′y′ + xy + x′y is equivalent to
(a) (11, 9) (b) (9, 13) (a) x′ + y′ (b) x + y
(c) (9, 10) (d) (11, 11) (c) x + y′ (d) x′ + y′
(GATE 2003: 2 Marks) (GATE 2004: 1 Mark)
Solution:
Solution:
ZW x′y′ + xy + x′y
00 01 11 10
XY = x′(y′ + y) + xy (as y′ + y = 1)
00 X 1 0 1 = x′ + xy (as x + x′y = x + y)
= x′ + y
01 0 1 X 0 Ans. (d)
11 1 X X 0 8. In an SR latch made by cross-coupling two NAND

gates, if both S and R inputs are set to 0, then it
10 X 0 0 X will result in
(a) Q = 0, Q′ = 1 (b) Q = 1, Q′ = 0
SOP: wy + w′y′ + z′wx ′ + xyz′ (c) Q = 1, Q′ = 1 (d) Indeterminate states
Literal count: 10 (GATE 2004: 1 Mark)
POS: (y′ + z′)(w′ + z′)(z′ + y)(x + z + w)
Literal count: 9 Solution: If both S and R are 0, using SR latch
Ans. (c) with NAND gate output state will be invalid state.
Ans. (d)
6. Consider the following circuit composed of XOR
gates and non-inverting buffers. 9. If 73x (in base-x number system) is equal to 54y
(in base-y number system), the possible values of x
and y are
A B
(a) 8, 16 (b) 10, 12
(c) 9, 13 (d) 8, 11
d1 = 2 d2 = 4 (GATE 2004: 1 Mark)
The non-inverting buffers have delays d1= 2 ns and Solution: By doing hit and trial with the options,
d2 = 4 ns as shown in the figure. Both XOR gates we arrive at
and all wires have zero delay. Assume that all gate If x = 8, (73)8 = 7 × 81 + 3 × 80 = 59
inputs, outputs and wires are stable at logic level 0 If y = 11, (54)11 = 5 × 111 + 4 × 110 = 59
at time 0. If the following waveform is applied at Ans. (d)
Chapter 1.indd 34 4/9/2015 9:40:49 AM

10. What is the result of evaluating the following two Continued

expressions using three-digit floating point arith-
metic with rounding? A B c d Output
(113 + −111) + 7.51

0 1 1 0 1
113 + (−111 + 7.51) 0 1 0 1 1
(a) 9.51 and 10.0, respectively 1 0 0 0 1

(b) 10.0 and 9.51, respectively 1 0 1 1 1
(c) 9.51 and 9.51, respectively
(d) 10.0 and 10.0, respectively Now using K-map, the expression becomes a + bd
(GATE 2004: 1 Mark) + bc→a + b(d+c).
Hence, three gates will be required.
Solution: Ans. (b)
(113 + −111) + 7.51 = 9.51 13. Which are the essential prime implicants of the fol-
113 + (−111 + 7.51) = 10.0 lowing Boolean function?
Ans. (a)
f(a, b, c) = a′c + ac′ + b′c
11. Consider a multiplexer with X and Y as data inputs
(a) a′c and ac′ (b) a′c and b′c
and Z as control input. Z = 0 selects input X, and
(c) a′c only (d) ac′ and bc′
Z = 1 selects input Y. What are the connections
required to realize the 2-variable Boolean function (GATE 2004: 2 Marks)
f = T + R, without using any additional hardware?
Solution:
(a) R to X, 1 to Y, T to Z f(a, b, c) = a′c + ac′ + b′c
(b) T to X, R to Y, T to Z = a′(b + b′)c + a(b + b′)c′ + (a + a′)b′c
(c) T to X, R to Y, 0 to Z = a′bc + a′b′c + abc′ + ab′c′ + a′b′c
(d) R to X, 0 to Y, T to Z
Using K-map:
c ab 00 01 11 10
Solution: Connect R to X, 1 to Y and T to Z.
Ans. (a) 0 1 1
1 1 1
12. A circuit outputs a digit in the form of 4 bits,
where 0 is represented by 0000, 1 by 0001, …, 9
by 1001. A combinational circuit is to be designed here ca′ and c′a are essential prime implicants, we
which takes these 4 bits as input and output 1 if cannot remove them.
the digit ≥5, and 0 otherwise. If only AND, OR and Ans. (a)
NOT gates are used, what is the minimum number
14. Consider the partial implementation of a 2-bit
of gates required?
counter using T flip-flops following the sequence
(a) 2 (b) 3 0-2-3-1-0, as shown in the following figure.
(c) 4 (d) 5
Solution: X
T2 Q2 T1 Q1
a B c d Output
0 0 0 0 0 MSB LSB
0 0 0 1 0
CLK
0 0 1 0 0
0 0 0 1 0 To complete the circuit, the input X should be
0 1 0 0 0 (a) Q2′ (b) Q2 + Q1
0 1 1 1 1 (c) (Q1 ⊕ Q2 )′ (d) Q1 ⊕ Q2
(Continued) (GATE 2004: 2 Marks)
Chapter 1.indd 35 4/9/2015 9:40:51 AM

Solution: Sequence 0 - 2 - 3 - 1 - 0 means (c) f is independent of Z.

00 - 11 - 01 - 00 (d) None of X, Y, Z is redundant.
If x = Q1 ⊕ Q2 , then the given output will be (GATE 2005: 1 Mark)
produced.
Ans. (d) Solution:
15. A 4-bit carry look-ahead adder, which adds two f = XY ′ + (YZ )′
4-bit numbers, is designed using AND, OR, NOT, = Y ′ + Z′
NAND, NOR gates only. Assuming that all the
inputs are available in both complemented and So, f is independent of X.
uncomplemented forms and the delay of each gate is Ans. (a)
one time unit, what is the overall propagation delay 18. The range of integers that can be represented by
of the adder? Assume that the carry network has an n-bit 2’s complement number system is
been implemented using two-level AND-OR logic.
(a) −2n−1 to (2n−1−1)
(a) 4 time units (b) 6 time units (b) −2(2n−1−1) to (2n−1−1)
(c) 10 time units (d) 12 time units (c) −2n−1 to 2n−1
(GATE 2004: 2 Marks) (d) −2(2n−1+ 1) to (2n−1−1)
(GATE 2005: 1 Mark)
Solution: Equations for carry look-ahead adder
are as follows: Solution: Negative numbers are one more than
C1 = G0 + P0 ⋅ C0
positive number. So, the range of 2’s complement
is 2n−1 to (2n−1 − 1).
C2 = G1 + P1 ⋅ C1 Ans. (a)
C3 = G2 + P2 ⋅ C2 19. The hexadecimal representation of 6578 is
C4 = G3 + P3 ⋅ C3
(a) 1AF (b) D78
The total number of gate delay will be 6. (c) D71 (d) 32F
Ans. (b) (GATE 2005: 1 Mark)
16. Let A = 1111 1010 and B = 0000 1010 be two
Solution:
8-bit 2’s complement numbers. Their product in 2’s
(657)8 = (110 101 111)8
complement is
(0001 1010 1111)16 = (1AF)16
(a) 1100 0100 (b) 1001 1100 Ans. (a)
(c) 1010 0101 (d) 1101 0101 20. The switching expression corresponding to f (A, B,
(GATE 2004: 2 Marks) C, D) = ∑(1,4,5,9,11,12)is
Solution: (a) BC ′D ′+ A′C ′D + AB ′D

A = 1111 1010 = 0000 0110 = −6 (b) ABC ′+ ACD + B ′C ′D
B = 0000 1010 = 10 (c) ACD ′+ A′BC ′+ AC ′D ′
A×B = −60 (d) A′BD + ACD ′+ BCD ′
−60 = 11000100 (GATE 2005: 1 Mark)
Ans. (a)
Solution:
17. Consider the following circuit.
X AB CD 00 01 11 10
Y f 00 1
01 1 1
Z
Which one of the following is TRUE? 11 1
(a) f is independent of X. 10 1 1
(b) f is independent of Y.
Chapter 1.indd 36 4/9/2015 9:40:52 AM

f = BC ′D ′ + A′C ′D + AB ′D Solution:

Ans. (a)
D0 D1 Q0 Q1
21. Consider the following circuit involving a positive-
1 1 1 1
edge triggered D FF.
0 1 0 1
A 1 0 1 0
X D Q 0 0 0 0
Y
Ans. (d)
CLK
Q´ Linked Answer Questions 23 and 24: Consider the
following floating-point format.
Consider the following timing diagram. Let Ai rep-
15 14 8 7 0
resent the logic level on the line A in the ith clock
period.
Sign bit Excess-64 Mantissa

CLK Exponent
0 1 2 3 4 5
X Mantissa is a pure fraction in sign-magnitude form.
23. The decimal number 0.239 × 213 has the following

hexadecimal representation (without normaliza-
tion and rounding off):
Let A′ represent the complement of A. The cor-
rect output sequence on Y over the clock periods 1 (a) 0D 24 (b) 0D 4D
through 5 is (c) 4D 0D (d) 4D 3D
(a) A0 A1A1′ A3 A4 (b) A0 A1A2′ A3 A4
(c) A1A2 A2′ A3 A4 (d) A1A2′ A3 A4 A5′ Solution:

0.239 × 213 = 0.00111101 × 213
(GATE 2005: 2 Marks) 64 + 13 = 77
Solution: The correct output sequence on Y is 0 100 1101 0011 1101
A0 A1A1′ A3 A4 . 4D 3D
Ans. (a)
Ans. (d)
22. Consider the following circuit.
24. The normalized representation for the above
Q0 D1 format is specified as follows. The mantissa has
D0 Q1 an implicit 1 preceding the binary (radix) point.
Assume that only 0’s are padded in while shifting
Q 0′ Q 1′ a field. The normalized representation of the above
number (0.239 × 213) is
(a) 0A 20 (b) 11 34
(c) 49 D0 (d) 4A E8
CLK
The flip-flops are positive-edge triggered D FFs.
Each state is designated as a 2-bit string Q0Q1. Solution: Using normalized form
Let the initial state be 00. The state transition 1.11101 × 210
sequence is 64 + 10 = 74
(a) 00-11-01 (b) 00-11 0 100 1010 11101000
(c) 00-10-01-11 (d) 00-11-01-10
4A E8
Ans. (d)
Chapter 1.indd 37 4/9/2015 9:40:54 AM

25. You are given a free running clock with a duty Solution:
cycle of 50% and a digital waveform of which f1 = zx + zy
changes only at the negative edge of the clock.
Which one of the following circuits (using clocked f = y (zx + zy ) + yx
D flip-flops) will delay the phase off by 180°? = yzx + yzy + yx = yx + yzx + zy
= yx + y (z + zx) = yx + y (z + x)
(a) f D D Q = yx + yz + yx
Q
Now use K-map, we have
CLK YZ
X 00 01 11 10
(b) f D Q 0 1
D Q
CLK
1 1 1 1 1
(c) f D Q D Q Hence, xy + yz + xz is the correct option.

Ans. (a)
CLK
27. Given two 3-bit numbers a2a1a0 and b2b1b0 and c,
the carry in the function that represents the carry
generate function when these two numbers are
(d) f D D Q added is
Q (a) a2b2 + a2a1b1 + a2a1a0 b0 + a2a0 b1b0
CLK + a1b2b1 + a1a0 b2b0 + a0 b2b1b0
(b) a2b2 + a2b1b0 + a2a1b1b0 + a1a0 b2b1
(GATE 2006: 1 Mark) + a1a0 b2 + a1a0 b2b0 + a2a0 b1b0
(c) a2 + b2 + (a2 ⊕ b2 ) (a1 + b1 + (a1 ⊕ b1 )(a0 + b0 ))
Solution: Left flip-flop is activated, the clock for
right will be 0. There will be delay of 180°. (d) a2b2 + a2a1b1 + a2a1a0 b0 + a2a0 b1b0
Ans. (c)
+ a1b2b1 + a1a0 b2b0 + a0 b2b1b0
26. Consider the circuit shown. Which one of the fol-
(GATE 2006: 1 Mark)
lowing options correctly represents f(x, y, z)?
Solution: Option (d) generates carry c always.
Ans. (d)
X
28. Consider the circuit in the diagram. The ⊕ opera-
0 MUX tor represents XOR. The D flip-flops are initial-
1 ized to zeroes (cleared).
Y
Data
0 MUX q2 Q D q1 Q D q0 Q D
Z f
1 clk clk clk
Clock
X
The following data 100110000 is supplied to the
Y “data” terminal in nine clock cycles. After that the
values of q2q1q0 are
(a) xz + xy + yz (b) xz + xy + yz (a) 000 (b) 001
(c) xz + xy + yz (d) xz + xy + yz (c) 010 (d) 101
(GATE 2006: 1 Mark) (GATE 2006, 1 Mark)
Chapter 1.indd 38 4/9/2015 9:40:57 AM

Solution: (c) cout ⊕ cn−1

Clock Input q2 q1 q0 (d) an−1 ⊕ bn−1 ⊕ cn−1
(GATE 2006: 1 Mark)
0 1 0 0 1
Solution: The overflow will be identified cout and
1 0 0 1 0 cn−1 are XORed.
2 0 1 0 0 Ans. (c)
3 1 0 1 1 31. Consider numbers represented in 4-bit gray code.
4 1 1 1 1 Let h3h2h1h0 be the gray code representation of a
number n and let g3g2g1g0 be the gray code of (n + 1)
5 0 1 0 0
(modulo 16) value of the number. Which one of the
6 0 0 1 0 following functions is correct?
7 0 1 0 0 (a) g0(h3h2h1h0) = ∑(1, 2, 3, 6, 10, 13, 14, 15)
8 0 0 1 0 (b) g1(h3h2h1h0) = ∑(4, 9, 10, 11, 12, 13, 14, 15)
(c) g2(h3h2h1h0) = ∑ (2, 4, 5, 6, 7, 12, 13, 15)
Ans. (c)
(d) g3(h3h2h1h0) = ∑(0, 1, 6, 7, 10, 11, 12, 13)
29. Consider a Boolean function f(w, x, y, z). Suppose (GATE 2006: 1 Mark)
that exactly one of its inputs is allowed to change
Solution: (n + 1) mod 16 is a 5-bit number; range
at a time. If the function happens to be true for two
is from 0 to 31.
input vectors i1 = (w1, x1, y1, z1) and i2 = (w2, x2,
y2, z2), we would like the function to remain true g2(2,4,5,6,7,12,13,15)
as the input changes from i1 to i2 (i1 and i2 differ Ans. (c)
in exactly 1 bit position), without becoming false 32. What is the maximum number of different Boolean
momentarily. Let f(w, x, y, z) = (5, 7, 11, 12, 13, 15). functions involving n Boolean variables?
Which of the following cube covers of f will ensure
that the required property is satisfied? (a) n2
n
(b) 2n 2
(c) 22 (d) 2n
(a) w′xz, wxy′, xy′z, xyz, wyz
(GATE 2007: 1 Mark)
(b) wxy,w′xz, wyz
(c) wxyz, xz, wx′yz
(d) wzy, wyz, wxz,w′xz, xy′z, xyz Solution: Maximum n
Boolean functions with n
(GATE 2006: 1 Mark) elements can be 22 .
Ans. (c)
Solution:
33. How many 3-to-8 line decoders with an enable
yz wx 00 01 11 10
input are needed to construct a 6-to-64 line decoder
00 without using any other logic gates?
01 1 1 (a) 7 (b) 8
(c) 9 (d) 10
11 1 1 1 (GATE 2007: 1 Mark)
10 1 Solution: We need to construct 6 to 64 decoder so
apply the following:
∑ i=1 8i = 8 + 64 = 9
From the K-map, we have w′xz, wxy′, xy′z, xyz, wyz. log 64 64 64 64
Ans. (a)
Ans. (c)
30. We consider the addition of two 2’s complement 34. Consider the following Boolean function of four
numbers bn-1bn-2 … b0 and an-1an-2 … a0. A variables:
binary adder for adding unsigned binary numbers is
used to add the two numbers. The sum is denoted f(w,x,y,z) = ∑ (1,3, 4, 6, 9, 11,12,14)
by cn-1cn-2 … c0 and the carry-out by cout. The function is
Which one of the following options correctly identi- (a) independent of one variable
fies the overflow condition? (b) independent of two variables
(a) cout (an -1 ⊕ bn -1 ) (c) independent of three variables
(d) dependent on all the variables
(b) an -1bn -1 cn -1 + an -1bn -1cn -1 (GATE 2007: 1 Mark)
Chapter 1.indd 39 4/9/2015 9:40:59 AM

Solution: f(w,x,y,z) = ∑(1, 3, 4, 6, 9, 11, 12, 14) Solution:

can be represented in K-map as P: Y*Z = YZ + Y′Z′ = Y(XY + X′Y′) + Y′(X′Y
+ XY′) = X
00 01 11 10
So, P is valid.
wx yz
Q: Y = X*Z
X*Z = XZ + X′Z′ = X(XY + X′Y′) + X′(X′Y +
00 1 1 XY′) = Y
So, Q is valid.
01 1 1 R: X*Y*Z = 1
X*Y*Z = Z*Z = ZZ + Z′Z′ = 1
11 1 1 So, R is also valid.
Ans. (d)
10 1 1 37. Suppose only one multiplexer and one inverter are
allowed to be used to implement any Boolean func-
Quadruple is formed, indicating the function is tion of n variables. What is the minimum size of
independent of two variables. the multiplexer needed?
Ans. (b) (a) 2n line to 1 line (b) 2n + 1 line to 1 line
(c) 2n − 1 line to 1 line (d) 2n − 2 line to 1 line
35. Let f(w, x, y, z) = ∑ (0, 4, 5, 7, 8, 9, 13, 15). Which
of the following expressions are NOT equivalent to f ? (GATE 2007: 2 Marks)
(P) x′y′z′+ w′xy′+ wy′z + xz Solution: Since only one inverter is allowed, so
(Q) w′y′z′+ wx′y′+ xz n−1 variable will be used for selection and one
(R) w′y′z′+ wx′y′+ xyz + xy′z as input. 2n−1 line to 1 line multiplexer will be
(S) x′y′z′+ wx′y′+ w′y required.
Ans. (c)
(a) P only (b) Q and S
(c) R and S (d) S only 38. In a look-ahead carry generator, the carry generate
(GATE 2007: 2 Marks) function Gi and the carry propagate function Pi for
inputs Ai and Bi are given by
Solution:
Pi = Ai ⊕ Bi and Gi = AiBi
wx yz 00 01 11 10
The expressions for the sum bit Si and the carry
00 1 bit Ci+1 of the look-ahead carry adder are given by
Si = Pi ⊕ Ci and Ci +1 = Gi + PiCi, where C0 is
01 1 1 1 the input carry.
11 1 1 Consider a two-level logic implementation of the
look-ahead carry generator. Assume that all Pi and
10 1 1 Gi are available for the carry generator circuit and
that the AND and OR gates can have any number
Option (d) does not contain essential prime of inputs. The number of AND gates and OR gates
implicants. needed to implement the look-ahead carry genera-
Ans. (d) tor for a 4-bit adder with S3, S2S1, S0, and C4 as its
outputs are, respectively,
36. Define the connective * for the Boolean variables
X and Y as X*Y = XY + X′Y′. Let Z = X*Y. (a) 6, 3 (b) 10, 4
Consider the following expressions P, Q and R. (c) 6, 4 (d) 10, 5
P: X = Y*Z Q: Y = X*Z R: X*Y*Z = 1
Which of the following is TRUE? Solution: Two 1-bit adders are required, where
each 1-bit adder required 5 AND and 2 OR gates.
(a) Only P and Q are valid. So, 10 AND gates and 4 OR gates are required.
(b) Only Q and R are valid.
Ans. (b)
(c) Only P and R are valid.
(d) All P, Q, R are valid. 39. The control signal functions of a 4-bit binary coun-
(GATE 2007: 2 Marks) ter are given below (where X is “don’t care”):
Chapter 1.indd 40 4/9/2015 9:41:00 AM

Solution:
Clear Clock Load Count Function
ab
1 X X X Clear to 0 00 01 11 10
cd
0 X 0 0 No change 00 1 1 1
0 1 X Load input
0 0 1 Count next 01
X
The counter is connected as follows: 11

X
A4 A3 A2 A1
10 X
1 1
So, b ⋅ d + a ⋅ d
Ans. (a)
Count = 1
41. In the IEEE floating-point representation, the
Clear 4-bit counter Load = 0 hexadecimal value 0x00000000 corresponds to
Clock (a) The normalized value 2−127
(b) The normalized value 2−126
(c) The normalized value +0
Inputs
(d) The special value +0
(GATE 2008: 1 Mark)
0 0 1 1 Solution: IEEE floating point standard represent

number in normalized form. This is the smallest
Assume that the counter and gate delays are positive number, that is, +0.
negligible. If the counter starts at 0, then it cycles Ans. (d)
through the following sequence:
42. Let r denote number system radix. The only
(a) 0, 3, 4 (b) 0, 3, 4, 5 value(s) of r that satisfy the equation 121r = 11r
(c) 0, 1, 2, 3, 4 (d) 0, 1, 2, 3, 4, 5 is/are
(a) Decimal 10 (b) Decimal 11
Solution: The counter will count if count = 1 and (c) Decimal 10 and 11 (d) Any value > 2
load = 0, the sequence will be 0, 1, 2, 3, 4. (GATE 2008: 1 Mark)
Ans. (c)
Solution:
40. In the Karnaugh map shown below, X denotes a
don’t care term. What is the minimal form of the (1 + 2r 2 + r 3 )1/2 = 1 + r
⇒ 1 + 2r 2 + r 3 = (1 + r)2 = 1 + r 2 + 2r
function represented by the Karnaugh map?
ab
00 01 11 10
r 3 + r 2 - 2r = 0 ⇒r > 2
cd
00 1 1 1 Ans. (d)
43. Given f1, f3 and f in canonical sum of products form
01 X (in decimal) for the circuit
11 X f1
f2 f
10 1 1 X
f3
(a) b ⋅ d + a ⋅ d (b) a ⋅ b + b ⋅ d + ab ⋅ d f1 = ∑m(4, 5, 6, 7, 8)
f3 = ∑m(1, 6, 15)
(c) b ⋅ d + a ⋅ b ⋅ d (d) a ⋅ b + b ⋅ d + a ⋅ d
f = ∑m(1, 6, 8, 15)
(GATE 2008: 1 Mark) then f2 is
Chapter 1.indd 41 4/9/2015 9:41:02 AM

(a) ∑m(4, 6) (b) ∑m(4, 8) Which one of the following is equivalent to P ∨ Q?

(c) ∑m(6, 8) (d) ∑m(4, 6, 8)
(a) ¬Q ¬P (b) P ¬Q
(GATE 2008: 1 Mark) (c) ¬P Q (d) ¬P ¬Q
Solution: To get f = ∑m(1, 6, 8, 15), f2 should (GATE 2009: 1 Mark)
set 6 and 8.
Ans. (c) Solution: Try using different values of P and Q.
44. If P, Q, R are Boolean variables, then For example, P = True and Q = False.
Ans. (b)
(P + Q)(P ⋅ Q + P ⋅ R)(P ⋅ R + Q)
simplifies to 48. How many 32 K × 1 RAM chips are needed to
provide a memory capacity of 256 K-bytes?
(a) P ⋅ Q (b) P ⋅ R
(c) P ⋅ Q + R (d) P ⋅ R + Q (a) 8 (b) 32
(c) 64 (d) 128
(GATE 2009: 1 Mark)
Solution:
(P + Q)(P ⋅ Q + P ⋅ R)(P ⋅ R + Q) Solution:
= (P ⋅ Q + P ⋅ Q + P ⋅ Q ⋅ R + PR)(P ⋅ R + Q) 256 × 8
Number of chips required = = 64
32 × 1
= P ⋅Q+P ⋅Q+P ⋅Q⋅R +P ⋅Q⋅R
Ans. (c)
= P ⋅Q (as 1 + R = 1)
Ans. (a)
49. The minterm expansion of f (P , Q, R) = PQ + QR + PR
45. (1217)8 is equivalent to f (P , Q, R) = PQ + QR + PR is
(a) (1217)16 (b) (028F)16 (a) m2+ m4+ m6+ m7 (b) m0+ m1+ m3+ m5
(c) (2297)10 (d) (0B17)16 (c) m0+ m1+ m6+ m7 (d) m2+ m3+ m4+ m5
(GATE 2009: 1 Mark) (GATE 2010: 1 Mark)
Solution:
Solution:
(1217)8 = (001 010 001 111)8
(0010 1000 1111) = (28F)16 = (028F)16 ff((PP,,Q
Q,,R
R)) =
= PQ
PQ+ QR′′ +
+QR +PRPR′′
Ans. (b) =
= PQ
PQ((R
R++RR′′)) +
+((P
P+ +PP ′′))QR
QR′′ +
+PP((QQ+
+QQ′′))R
R′′
46. What is the minimum number of gates required to == PQR
PQR + PQR′′ +
+ PQR +PP′QR
′QR′′ + PQ′′R
+ PQ R′′
implement the Boolean function (AB + C ) if we =
=mm77 +
+mm66 +
+mm22 +
+mm44
have to use only 2-input NOR gates?
(a) 2 (b) 3 Ans. (a)
(c) 4 (d) 5
(GATE 2009: 1 Mark) 50. P is a 16-bit signed integer. The 2’s complement
representation of P is (F87B)16. The 2’s comple-
Solution: ment representation of 8×P is
(AB + C ) = (A + C)(B + C)
(a) (C3D8)16 (b) (187B)16
= ((A + C )′ + (B + C)′)′
(c) (F878)16 (d) (987B)16
Thus, it requires three NOR gates.
47. The binary operation is defined as follows
Solution:
P Q P Q (F87B)16 = (1925)2
1925 × 8 = (15400)2 = 2’s complement of (15400)2
T T T = (C3D8)16
T F T Ans. (a)
F T F
51. The Boolean expression for the output f of the mul-
F F T tiplexer shown below is
Chapter 1.indd 42 4/9/2015 9:41:04 AM

Solution:
R P + Q + Q + R + P +P
R++Q
Q++Q
R+ R + P + R + Q + R
R = (P + Q + Q + R) ⋅ (P=+(PR +
+QQ+
+QR)+ R) ⋅ (P + R + Q +
f
R = )(P Q + QR)(P R + QR)
= (P Q + QR)(P R + QR
R = P QP
= P QP R + P QQR + QRP R+R QRQR
+ P QQR + QRP R + Q
=+
= P QR + P QR + P QR P QR + P QR + P QR + QR
= P QR + QR = P QR + QR
= QR(P + 1) = QR = =
Q QR
+ R(P + 1) = QR = Q + R
P Q
Ans. (a)
(a) P ⊕ Q ⊕ R (b) P ⊕ Q ⊕ R 53. In the sequential circuit shown below, if the initial
(c) P ⊕ Q ⊕ R (d) P + Q + R value of the output Q1Q0 is 00, what are the next
four values of Q1Q0?
(GATE 2010: 1 Mark)
Solution:
1 T Q T Q
P Q Input Output Clock
0 0 R P′Q′R
0 1 R′ P′QR′
Q0 Q1
1 0 R′ PQ′R′ (a) 11,10,01,00 (b) 10,11,01,00
1 1 R PQR (c) 10,00,01,11 (d) 11,10,00,01
P′Q′R + P′QR′ + PQ′R′ + PQR = P ⊕ Q ⊕ R Solution:
Ans. (b)
Clock Q1 Q0
52. What is the Boolean expression for the output f of 1 0 0
the combinational logic circuit of NOR gates given 2 1 1
below?
3 1 0
P 4 0 1
Q 5 0 0
Ans. (a)
Q 54. Which one of the following circuits is NOT equiva-
R lent to a 2-input XNOR (exclusive NOR) gate?
f (a) (b)
P
R
(c) (d)
Q
R
(a) Q + R (b) P + Q
(GATE 2011: 1 Mark)
(c) P + R (d) P + Q + R
Solution: Option (d) does not show XNOR gate
(GATE 2010: 2 Marks) functionality.
Ans. (d)
Chapter 1.indd 43 4/9/2015 9:41:06 AM

55. The simplified SOP (Sum of Product) form of the 57. If at some instance prior to the occurrence of the
Boolean expression clock edge, P, Q and R have a value 0, 1 and 0,
(P + Q + R) ⋅ (P + Q + R) ⋅ (P + Q + R) is
respectively, what shall be the value of PQR after
the clock edge?
(a) (P ⋅ Q + R) (b) (P + Q ⋅ R) (a) 000 (b) 001
(c) (P ⋅ Q + R) (d) (P ⋅ Q + R) (c) 010 (d) 011
(GATE 2011: 1 Mark)
Solution: Solution:
P QR 00 01 11 10 Clock D1 = R D2 = (P + R)′ D3 = QR′ P Q R
0 0 1 1 1 1 0 1 0 0 1 0
2 0 1 1 0 1 1
1 0 0 0 0
3 1 0 0 1 0 0
which implies (P + R)(P + Q) 4 0 0 0 0 0 0

Ans. (b) After 010, value of PQR is 011.
56. The minimum number of D flip-flops needed to Ans. (d)
design a Mod-258 counter is 58. If all the flip-flops were reset to 0 at power on,
what is the total number of distinct outputs (states)
(a) 9 (b) 8
represented by PQR generated by the counter?
(c) 512 (d) 258
(a) 3 (b) 4
(GATE 2011: 1 Mark)
(c) 5 (d) 6
Solution: 2n ≥ 258 = 9 K flip-flops (GATE 2011: 2 Marks)
Ans. (a)
Solution: Clock cycles can be seen from the
Common Data Questions 57 and 58: Consider table given in Question 57. The number of district
the following circuit involving three D-type outputs is 4.
flip-flops used in a certain type of counter Ans. (b)
configuration.
59. The truth table
X Y f(X,Y)
0 0 0
P 0 1 0
D Q
1 0 1
clock 1 1 1
Q
represents the Boolean function
(a) X (b) X + Y
(c) X ⊕ Y (d) Y
D Q Q
(GATE 2012: 1 Mark)
clock Solution: From given truth table, equation

Q formed is
XY′ + XY⇒X(Y′ + Y) ⇒X
Ans. (a)
R 60. The decimal value 0.5 in IEEE single-precisionfloating-

D Q point representation has
(a) F
raction bits of 000…000 and exponent value
clock
Q of 0
(b) Fraction bits of 000…000 and exponent value
of −1
Chapter 1.indd 44 4/9/2015 9:41:08 AM

(c) Fraction bits of 100…000 and exponent value Solution: Smallest integer for 2’s complement is
of 0 −2n− 1
(d) No exact representation So here n = 8, −28 − 1 = −128
Ans. (b)
(GATE 2012: 1 Mark)
63. In the following truth table, V = 1 if and only if
Solution: Binary representation of 0.5 is 0.1000,
which can be written as 1.000 × 2−1
the input is valid.
Here, exponent = −1 and fraction bits are 000…000.
Ans. (b) Inputs Outputs
61. What is the minimal form of the Karnaugh map D0 D1 D2 D3 X0 X1 V
shown below? Assume that X denotes a don’t care 0 0 0 0 x x 0
term.
1 0 0 0 0 0 1
ab x 1 0 0 0 1 1
cd 00 01 11 10 x x 1 0 1 0 1
x x x 1 1 1 1
00 1 X X 1
What function does the truth table represent?

01
X 1 (a) Priority encoder (b) Decoder
(c) Multiplexer (d) Demultiplexer
11 (GATE 2013: 1 Mark)
Solution: From the given truth table it is clear

10 that it is 4:2 priority encoder.
1 X
Ans. (a)
64. Which one of the following expressions does NOT

(a) b d (b) b d + bc
represent exclusive NOR of x and y?
(c) b d + abcd (d) b d + bc + cd
(GATE 2012: 2 Marks) (a) xy + x′y′ (b) x ⊕ y′
(c) x′ ⊕ y (d) x′ ⊕ y′
Solution: (GATE 2013: 1 Mark)
1 X X 1 Solution: x′ ⊕ y′ = (x′y + y′x) = x ⊕ y, this

represents XOR. So, option (d) is not exclusive
NOR.
X 1 Ans. (d)
65. A RAM chip has a capacity of 1024 words of 8 bits

each (1K × 8). The number of 2 × 4 decoders with
enable line needed to construct a 16K × 16 RAM
from 1K × 8 RAM is
1 X
(a) 4 (b) 5 (c) 6 (d) 7
From above grouping, the equation obtained is:
cb + d b
Ans. (b)
Solution: Number of chips required =
62. The smallest integer that can be represented by an 16 × 16
= 16 × 2
8-bit number in 2’s complement form is 1× 8
(a) −256 (b) −128 So, it requires 16 vertical and 2 horizontal chips.
(c) −127 (d) 0 For this five 2 × 4 decoders are required.
(GATE 2013: 1 Mark) Ans. (b)
Chapter 1.indd 45 4/9/2015 9:41:10 AM

PRACTICE EXERCISES
Set 1
1. Find the value of x in (257) base 10 = (41) base x? (a) I-A, II-C, III-B (b) I-B, II-A, III-B
(c) I-A, II-B, III-A (d) I-A, II-C, III-B
(a) 61 (b) 64
(c) 127 (d) 256 10. Consider the following circuit:
2. What is the result of 01A2 + 001A = ? f1
f2
(a) 01BC (b) 01CB f5
f3
(c) 01CC (d) 01DC
3. The simplified function of Boolean function Given: f1(x, y, z) = ∑m(1, 2, 6, 7), f3(x, y, z) =
F(W, X, Y, Z) = Σ(0, 1, 2, 5, 8, 9, 10) is ∑m(0, 3), f5(x, y, z) = ∑m(0, 1, 3, 6)
(a) (W ′ + X′) (Y ′ + Z′) (X′ + Z). How many different minterms combinations are
(b) (W + X) (Y ′ + Z′) (X′ + Z′). possible for f2?
(c) (W ′ + X′) (Y + Z) (X + Z). (a) 15 (b) 12
(d) (W ′ + X) (Y + Z′) (X′ + Z′). (c) 8 (d) 6
4. What is the hexadecimal representation of (1356)
base 8? 11. The two numbers represented in signed 2’s comple-
ment form is A = 11101101 and B = 11100110.
(a) 2FE (b) 2DE
If B is subtracted from A, the value obtained in
(c) 2EE (d) EEE
signed 2’s complement form is
5. How many 3 × 8 decoders are needed to construct (a) 10000111 (b) 00000111
4 × 16 decoder? (c) 10001001 (d) 11001001
(a) 2 (b) 4
12. Simplify the Boolean function
(c) 3 (d) 24/64
F(w, x, y, z) = ∑(1,3,10) + ∑d(0,2,8,12)
6. If 73 (in base x number system) is equal to 54 (in
base y number system), the possible values of x and (a) F = w′x′+ x′z (b) F = wx′ + x′z′
y are (c) F = w′x + x′z′ (d) F = w′x′ + x′z′
(a) 11,8 (b) 8,13 13. Which of the following is the complement of func-
(c) 11,11 (d) 8,11 tion F = x(y′z′ + yz)
7. Total number of 8 × 1 MUX required for imple- (a) x + yz′ + y′z′ (b) x′ + yz′ + y′z
mentation of 256 × 1 MUX is (c) x′ + y′z′ + y′z (d) x′ + y′z′ + y′z′
(a) 37. (b) 35. 14. Match the following:

(c) 36. (d) 38.
IC Family Characteristics
8. Number of 8 × 1 MUX needed at second last level
P. TTL (i) High component density
is (assume level having only 1(8 × 1) MUX to be
Q. ECL (ii) Low power consumption
last level)
R. MOS (iii) Evolution of “diodetransistor-logic”
(a) 1 (b) 2 S. CMOS (iv) High-speed digital circuits
(c) 8 (d) 4
P Q R S
9. Match the columns (a) (i) (ii) (iii) (iv)
(b) (i) (iii) (iv) (ii)
Column I Column II
(c) (iii) (iv) (i) (ii)
I. xy + z(x XOR y) A. ∑m(1,2,4,7) (d) (iv) (iii) (ii) (i)
II. x (XOR) y (XOR) z B. ∑m(3,5,6,7) 15. What is the base value of binary number system?
C. ∑m(1,2,5,6)
III. xy + yz + zx (a) 0 (b) 2
(c) 1 (d) 01
Chapter 1.indd 46 4/9/2015 9:41:11 AM

PRACTICE EXERCISES 47
16. What is the maximum value that we can represent 24. Which FF has its output clearly defined on all the
using N bits? combinations inputs?
(a) 2n (b) 2n− 1 (a) SR FF (b) D FF
(c) 2n − 2 (d) 2n − 1 (c) JK FF (d) T FF
17. The number of bits required to represent 128 dif- 25. The main difference between JK and RS flip-flop
ferent numbers is is that
(a) 5. (b) 7. (a) JK flip-flop do not need a clock pulse whereas
(c) 6. (d) 8. RS flip-flop requires the same.
(b) The output Q is connected to J and K is
18. Consider the following truth table, where A, B, C
grounded whereas it is opposite in R and S.
represents the inputs and F the output function
(c) JK flip-flop accepts both inputs as 1.
(d) JK flip-flop is acronym of Johnson-Keith’s
A B C F multivibrator and RS of reset-set multivibrator.
0 0 0 0
26. AND gate provides a LOW output .
0 0 1 0
0 1 0 0 (a) when any one input is LOW
0 1 1 1 (b) always
1 0 0 0 (c) when both the inputs are HIGH
1 0 1 0 (d) when both the inputs are LOW
1 1 0 1 27. How many pins does the 4049 IC have?
1 1 1 0
(a) 22 (b) 16
(a) B′(A + C ′)(A′ + C ′) (b) B(A + C ′)(A′ + C ) (c) 20 (d) 40
(c) B′(A′ + C ′)(A′ + C ) (d) B(A + C )(A′ + C ′)
28. Which of the following expressions is in the sum-of-
19. Which of the following expressions is equivalent to products (SOP) form?
A ⊕ B ⊕ C? (a) A′B + C ′D ′ (b) AB ′(CD)
(a) (A + B + C)(A′ + B′ + C ′) (c) (A + B ′)(C + D) (d) (A)B ′(C + D ′)
(b) (A + B + C)(A′ + B′ + C) 29. The significance of MSD in a decimal number
(c) ABC + A′(B ⊕C) + B′(A ⊕ C) system is
(d) (A′ + B + C)(A′ + B′ + C ′)
(a) the middle digit of a stream of numbers.
20. A half adder is also known as (b) the digit to the right of the decimal point.
(a) NOR circuit. (b) XNOR circuit. (c) the last digit on the right.
(c) NAND circuit. (d) XOR circuit. (d) the digit with the most weight.
21. Which of the following is minimum error code? 30. The Generic Array Logic (GAL) device is .
(a) ANSI code (b) Grey code (a) complement of PAL.

(c) XS3 code (d) ASCII code (b) reprogrammable device.
(c) CMOS device.
22. Register is a (d) CMOS device which is also reprogrammable.
(a) set of capacitors and inductors used to provide 31. Derive the Boolean expression for the logic circuit
input instructions in a digital computer shown below:
(b) set memory cells used to store the I/O data.
(c) temporary storage unit within the CPU. A
(d) permanent storage unit within the CPU. B
23. A T FF can be realized by SR FF by? C
(a) Connecting S to Q ′.

D
(b) Connecting R is connected to Q ′.
E
(c) Connecting S and R to Q. (a) C(A + B)DE (b) C(A + B)D + E
(d) S and R are connected to Q and Q ′, respectively. (c) ABCDE (d) (DE) + (ABC )
Chapter 1.indd 47 4/9/2015 9:41:11 AM

32. An XNOR gate gives a HIGH output with the fol- 43. What combination of the inputs a JK flip-flop
lowing input combination toggles?
(a) A = LOW, B = HIGH (a) J = 0, K = 0 (b) J = 0, K = 1
(b) A = HIGH, B = LOW (c) J = 1, K = 0 (d) J = 1, K = 1
(c) A = LOW, B = LOW
(d) It is not possible to get a HIGH output in this gate 44. In a 6-bit Johnson counter sequence, the total
number of states or bit patterns is _________.
33. A shift register that has a parallel input, or a bidi-
rectional serial load and having internal shift fea-
tures, can be termed as Set 2
(a) Tristate device.
(b) Parallel-in shift-out (PISO) register. 1. The simultaneous equations on the Boolean vari-
(c) Universal shift register. ables x, y, z and w, are
(d) Bidirectional shift register. x+y+z=1
xy = 0
34. Which of the following logic families has the least
xz + w = 1
propagation delay?
xy + zw = 0
(a) 54LYY (b) BiCMOS
(c) ECL (d) 74SXX have the following solution for x, y, z and w,
respectively.
35. Convert binary 111111110010 into hexadecimal.
(a) 0100 (b) 1101
(a) (EF2)16 (b) (FF2)16
(c) 1011 (d) 1000
(c) (FE2)16 (d) (FD2)16
36. Convert the binary number (1001.0010)2 to decimal. 2. Given the following Karnaugh map, which one
of the following represents the minimal sum-of-
(a) 90.125 (b) 9.125
product of the map?
(c) 91.25 (d) 12.125
37. The alphanumeric code widely used for input and wx
yz 00 01 11 10
output to/from a computer?
(a) Biquinary Code (b) ASCII 00 0 X 0 X
(c) ANSI Code (d) EBCDIC 01 X 1 X 1
38. (8B3F)16 = (?)2
11 0 X 1 0
(a) 111101101111 (b) 01110101111
(c) 1011001111101111 (d) 1000101100111111 10 0 1 X 0
39. Solving -11 + (-2) will yield which 2’s comple-
(a) xy + y′z
ment answer?
(b) wx′y′ + xy + xz
(a) 10011001 (b) 1000 1001 (c) w ′x + y′z + xy
(c) 1111 0011 (d) 1110 1001 (d) xz + y
40. Dividing (1100010)2 by (0101)2 yields (X)10. The
value of X is _________. 3. Minimum SOP for f (w, x, y, z) shown in Karnaugh-
map below is
41. The solution to (0110)16 + (10010)16 = (?)16
wx
(a) 1012016 (b) 1112016 yz 00 01 11 10
(c) 1002016 (d) 00B016
00 0 1 1 0
42. Which of the following is correct for full adders?
01 X 0 0 1
(a) Full adders can be used for multiplication also.
(b) Full adders are used to make half adders and 11 X 0 0 1
quad adders.
(c) Full adders have three inputs including a carry 10 0 1 1 X
input.
(d) In a parallel full adder, the first stage may be (a) xz + y′z (b) xz ′ + zx ′
a half adder. (c) x′y + zx ′ (d) None of the above
Chapter 1.indd 48 4/9/2015 9:41:12 AM

4. Consider the following logic circuit whose inputs 8. Which of the following sets of component(s) is/
are functions f1, f2, f3 and output f. are sufficient to implement any arbitrary Boolean
function?
f1(x, y, z)
f2(x, y, z) f(x, y, z) (a) XOR gates, NOT gates
(b) 2 to 1 Multiplexors
(c) AND gates, XOR gates
(d) Three-input gates that output (A⋅B) + C for
f3(x, y, z) = ? the inputs A, B and C
Given that f1(x, y, z) = ∑(0, 1, 3, 5); f2(x, y, z) = 9. Which function does NOT implement the Karnaugh
∑(6, 7) and f3(x, y, z) = ∑(1, 4, 5) then f is map given in the following figure?
(a) ∑(1, 4, 5) (b) ∑(6, 7)

wz→ 00 01 11 10
(c) ∑(0, 1, 3, 5) (d) ∑(4, 5, 7)
xy↓
00 0 x 0 0
5. f(A, B) = A′ + B. Simplified expression for func-
tion f(f(x + y, y), z) is 01 0 x 1 1
(a) x′ + z (b) xyz 11 1 1 1 1
(c) xy′ + z (d) None of the above 10 0 x 0 0
6. Two NAND gates having open collector outputs

(a) (w + x)y (b) xy + yw
are tied together as shown in the following figure.
The logic function Y, implemented by the circuit is (c) (w + x)(w + y)(x + y) (d) None of the above
A 10. Consider the circuit shown below. The output of a

B 2:1 multiplexer is given by the function f ?
C
Y
D 1 a
E g a
2:1
MUX 2:1
0 b MUX f
(a) Y = ABC + DE (b) Y = ABC + DE c
(c) Y = ABC ⋅ DE (d) Y = ABC ⋅ DE
b c
7. What is the equivalent Boolean expression in prod- x1
uct-of-sums form for the Karnaugh map given in x2
the following figure.
(a) x1′x2′ + x1x2 (b) x1′x2 + x1x2′
AB
00 01 11 10 (c) x1′+ x2 (d) x1′ + x2′
CD
1 1 11. Consider the following multiplexer where I0, I1, I2,
00
I3 are four data input lines selected by two address
line combinations A1A0 = 00, 01, 10, 11, respec-
01 1 1
tively, and f is the output of the multiplexer, EN is
the Enable input.
11 1 1
10 1 1 x I0
I1
y I2
(a) BD + BD 4 to 1 f(x,y,z) ?
I3 MUX
(b) (B + C + D)(B + C + D) A1
(c) (B + D)(B + D) z A0
EM
(d) (B + D)(B + D)
Chapter 1.indd 49 4/9/2015 9:41:15 AM

The function f(x, y, z) implemented by the above b3 b2 b1 b0

circuit is
(a) xyz′ (b) xy + z
(c) x + y (d) None of the above F FF
12. Which of the given number has its IEEE-754 32-bit

floating-point representation as
(0 10000000 110 0000 0000 0000 0000 0000) Cin
4-Bit binary F
(a) 2.5 (b) 3.0 adder
(c) 3.5 (d) 4.5 Cout
13. Consider the following state table in figure for a

sequential machine. The number of states in the dy d3 d2 d1 d0
minimized machine will be
(a) Binary to Hex conversion
(b) Binary to BCD conversion
0 1 (c) Binary to grey code conversion
A D, 0 B, 1 (d) Binary to radix −12 conversion
B A, 0 C, 1 18. Given (224)r = (13)r ′
C A, 0 B, 1
The value of the radix r is
D A, 1 C, 1 (a) 10 (b) 8 (c) 5 (d) 6
19. Zero has two representations in:

(a) 4 (b) 3
(c) 2 (d) 1 (a) sign magnitude (b) 1’s complement
(c) 2’s complement (d) both (a) and (b)
14. A ROM is used to store the table for multiplication 20. The number 43 in 2’s complement representation is
of two 8-bit unsigned integers. The size of ROM
(a) 01010101 (b) 11010101
required is
(c) 00101011 (d) 10101011
(a) 256K × 16 (b) 64K × 8
(c) 4K × 16 (d) 64K × 16 21. Consider the values A = 2.0 × 1030, B = −2.0 ×
1030, C = 1.0 and the sequence
15. Booth’s algorithm for integer multiplication gives X: = A + B
worst performance when the multiplier pattern is Y: = A + C
(a) 101010……1010 (b) 100000…..001 X: = X + C
(c) 111111……1111 (d) 011111…..1110 Y: = Y + B
(a) X = 1.0, Y = 1.0 (b) X = 0.0, Y = 1.0
16. Consider the following floating point number
(c) X = 1.0, Y = 0.0 (d) X = 0.0, Y = 0.0
representation.
22. The 2’s complement representation of (−539)10 in
31 24 23 0
hexadecimal is
Exponent Mantissa
(a) ABE (b) DBC
The exponent is in 2’s complement representation (c) DE5 (d) 9E7
and mantissa is in the sign magnitude representation. 23. The decimal value 0.25
The range of the magnitude of the normalized
numbers in this representation is (a) is equivalent to binary 0.1
(b) is equivalent to binary 0.01
(a) 0 to 1 (b) 0.5 to 1
(c) 2−23 to 0.5 (d) 0.5 to (1 − 2−23)
(c) is equivalent to binary 0.00111
(d) cannot be represented precisely in binary
17. Consider the circuit in the following figure which
24. Sign extension is the step in
has a 4-bit binary number b3b2b1b0 as input and
a 5-bit binary number d4d3d2d1d0 as output. The (a) floating point multiplication.
circuit implements (b) signed 16-bit integer addition.
Chapter 1.indd 50 4/9/2015 9:41:15 AM

ANSWERS TO PRACTICE EXERCISES 51
(c) arithmetic left shift. 28. A flip-flop has a 4-ns delay from the time the clock
(d) converting a signed integer from one size to edge occurs to the time the output is complemented
another. in a 10-binary ripple counter. The ripple counter
operates reliably at which maximum frequency?
25. The state of a 12-bit register is 100010010111. It is
represented in BCD, what is its content? (a) 25 MHz (b) 30 MHz
(c) 32 MHz (d) 40 MHz
(a) 2199 (b) 3423
(c) 564 (d) 897 29. How many parity check bits must be included with
26. Simplification of given Boolean expressions = the data word to achieve single-error correction
{(a + b′ + c′)(a′ + c′)′′} to a minimum number of when data word contains 16 bits?
literals is (a) 4 Bits (b) 5 Bits
(a) (c′ + a′b′) (b) (a′ + b′c′) (c) 6 Bits (d) 7 Bits
(c) (b′ + a′c′) (d) (a′ + b′ + c′)
30. How many 32K × 8 RAM chips are needed to
27. Simplification of the following Boolean function provide a memory capacity of 256K bytes?
{F(A, B, C, D) = ∑(3, 7, 11, 13, 14, 15)} using
(a) 4 (b) 5
K-maps:
(c) 6 (d) 8
(a) CD + ABC + BCD (b) CD + ABC + ABD
(c) CD + ACD + ABD (d) CD + ABC + ACD
ANSWERS TO PRACTICE EXERCISES
Set 1
1. (b) 9. (b) 17. (b) 25. (c) 33. (c) 41. (a)
2. (a) 10. (a) 18. (d) 26. (a) 34. (c) 42. (d)
3. (a) 11. (b) 19. (c) 27. (b) 35. (b) 43. (d)
4. (c) 12. (d) 20. (c) 28. (a) 36. (b) 44. (12)
5. (a) 13. (b) 21. (b) 29. (d) 37. (b)
6. (d) 14. (c) 22. (c) 30. (b) 38. (d)
7. (a) 15. (b) 23. (d) 31. (a) 39. (c)
8. (d) 16. (d) 24. (c) 32. (c) 40. (3)
Set 2
1. (c) Considering each option However, given xy =0

Therefore, option (b) is not the solution.
(a) Given x = 0, y = 1, z = 0 and w = 0
i. x + y + z = 0 + 1 + 0 = 1 (c) Given x = 1, y = 0, z = 1 and w =1
ii. xy = 0 ⋅ 1 = 0 i. x + y + z = 0 + 0 + 1 = 1
iii. xz + w = 0 ⋅ 0 + 0 = 0 ii. xy = 1 ⋅ 0 = 0
but xz + w should be 1, so option (a) is incorrect. iii. xz + w = 1 ⋅ 1 + 1 = 1
(b) Given x = 1, y = 1, z = 0 and w = 1 iv. xy +z′w′ = 1 ⋅ 0 + 0 ⋅ 0 = 0
i. x + y + z = 1 + 1 + 0 = 1 It satisfies all the given conditions. So, (c) is the
ii. xy = 1 ⋅ 1 = 1 correct option.
Chapter 1.indd 51 4/9/2015 9:41:16 AM

2. (a) 7. (c) The Karnaugh map for POS can be given by

wx
00 01 11 10 CD
yz 01 11
00 10
AB
00 0 X 0 X
00 0 0 1 1
01 X 1 X 1
01 X X 1 x
11 0 X 1 0
11 1 1 1 0
10 0 1 X 0
10 0 1 1 0
SOP is xy + y′z.
POS form of f is given as
3. (c) F(C, D, A, B) = (B + D)(B + D)
wx
yz 00 01 11 10 8. (c) Functionally complete operation set is the set
00 0 1 1 1 of logic functions from which all other functions
01 x 0 0 0 can be realized.
11 x 0 0 0
10 0 1 1 x [XOR, AND] is a functionally complete set.
9. (c) For SOP form

Sum-of-product for given function is x′y + zx′.
wz
4. (a) xy 00 01 11 10
f1(x,y,z) f (x,y,z) 00 0 X 0 1
f2(x,y,z)
01 0 X 1 1
11 1 1 1 1
f3(x,y,z) 10 0 X 0 0
f1f2 = f1(x, y, z) NAND f2(x, y, z)

= ∑ (0, 1, 2, 3, 4, 6, 7) f in SOP can be given as
f3 = ∑ (1, 4, 5) f(x, y, w, z) = xy + yw
f3′ = ∑ (0, 1, 2, 3, 6, 7) or y(x + w)
f = f1f2 NAND f3′ = (f12f3′)′ Therefore, options (a) and (b) are true.
= (∑ (0, 1, 2, 3, 6, 7))′ = ∑ (1, 4, 5) For POS form
wz
5. (c) f(f(x + y, y), z)= f((x + y)′ + y, z) = f(x′y′ + y, z) xy 00 01 11 10
= f(x′ + y, z) = (x′ + y)′ + z = xy′ + z 0 X 0 0
00
01 0 X 1 1
6. (c)
A
B 11 1 1 1 1
C
Y 10 0 X 0 0
D
E
f(x, y, w, z) = y(x + w)
Therefore, Y = ABC ⋅ DE
Option (c) is incorrect.
Chapter 1.indd 52 4/9/2015 9:41:19 AM

10. (a) 14. (d) To store the table for multiplication of two
8-bit unsigned integers, size of ROM required is
a
22N× 2N = 216× 16 = 64K×16
1
g a
2:1
MUX 2:1 15. (a)
0 b MUX f
c Booth recording of a multiplier are
(a) 1 0 1 0 1 0 …… 1 0 1 0
b c ⇓
x1 −1 +1 −1 +1 −1 +1……−1 +1 −1 0
x2 (b) 1 0 0 0 0 0 ……0 0 0 1
⇓
The output g is g = 1 x1′ + 0 x1 = x1′ −1 + 0 0 0 0 0……−0 0 +1 −1
The output f is (c) 1 1 1 1 1 1 …… 1 1 1 1
f = gx2′ + x1x2
⇓

f = x1′x2′ + x1x2
0 0 0 0 0 0…… 0 0 0 −1
11. (a) (d) 0 1 1 1 1 1 …… 1 1 1 0
= EN(A1A0 I 0 + A1A0 I1 + A1A0 I2 + A1A0 I3 ) ⇓
1 0 0 0 0 0 …… 0 0 −1 0
= z ′(xyz ′ + yzx + 0 + y ′zy ′)
= xyz ′ + xyzz ′ + 0 ⋅ z ′ + z ′ ⋅ zy ′ Therefore, option (a) gives the worst performance
because it contains maximum number of 1’s.
= xyz ′ + 0 + 0 + 0 = xyz ′
12. (c) Sign bit S = 0 ⇒ positive number 16. (d)
E = 1000 0000B = 128D (in normalized form) 31 24 23 0

Fraction is 1.11B (with an implicit leading 1) Exponent Mantissa
= 1 + 1 × 2−1 + 1 × 2−2 = 1.75D
The number is + 1.75 × 2(128 − 127) = +3.5D The range of the magnitude of the normalized
13. (b) number: 0.5 to (1 − 2−23)
x
0 1 17. (d) Whenever both b3 and b2 is 1, then 01 00 is
added to the number.
A D, 0 B, 1
b3b2b1b0 = (1100)B = (12)10
B A, 0 C, 1
These two states So, in all the numbers greater than equal to
C A, 0 B, 1 work same 12, 0100 is added. Therefore, the circuit is binary
to radix −12 converter.
D A, 1 C, 1
18. (c)
Combining two states B and C into single state (a) When the radix r is 10
“BC ”. Then (224)10 ≈ 14.967 ≠ (13)10
x
0 1 (b) When the radix is 8
A D, 0 BC , 1 Then (224)8 ≈ (148)10 = (12.165)10 and
(13)8 = (11)10. Hence, (224)8 ≠ (13)8
BC A, 0 BC , 1 Option (b) is not correct.
(c) When the radix is 5
D A, 0 BC , 1
(224)5 ≈ (64)10 = (8)10 and (13)5 = (8)10 .
Therefore, (224)5 = (13)5
Therefore, the number of states in the minimized
machine will be 3. Option (c) is correct.
Chapter 1.indd 53 4/9/2015 9:41:21 AM

(d) When the radix is 6 Therefore, the decimal value 0.25 is equivalent to
binary 0.01.
(224)6 = (88)10 ≈ (9.38)10 and (13)6 = (9)10.
24. (d) Sign extension is the operation for increasing
Therefore, (224)6 ≈ (13)6 the number of bits of a binary number while pre-
Option (d) is not correct. serving the number’s sign (positive/negative) and
value.
19. (d) Zero has two representations in
25. (d)
Sign magnitude as (i) 0 0 0 0, (ii) 1 0 0 0
26. (a) (a + b′ + c′)(a′ + c′)
1’s Complement as (i) 0 0 0 0, (ii) 1 1 1 1
⇒ aa′ + ac′ + a′b′ + b′c′ + a′c′ + c′c′
20. (c) Positive 2’s complement numbers are repre- ⇒ c′ + b′(a′ + c′)
sented as same simple binary. So +43 in binary is ⇒ c′ + a′b′
0 0 1 0 1 0 1 1. 27. (b)
21. (a) Given that A = 2.0 × 10 , B = −2.0 × 10 ,
30 30
CD
C = 1.0 00 01 11 10
AB
m0 m1 m8 m2
(i) X = A + B = 2.0 × 1030 + (−2.0 × 1030) = 0.0 00
1
m4 m5 m7 m6
(ii) Y = A + C = 2.0 × 1030 + (1.0) 1
01
(iii) Y = X + C = 0.0 + 1.0 = 1.0 m12 m13 m15 m14
11 1 1 1 ABC
(iv) Y = Y + B = 2 .0 × 1030 + 1.0 + (−2.0 × m5 m9 m11 m10
1030) = 1.0 10 1
Values of X and Y are 1.0 and 1.0, respectively.

ABD CD
22. (c) Compute a binary representation of the magni- F = CD + ABC + ABD
tude of the number and then take 2’s complement
of binary representation. 28. (a) In worst case, all 10 flip-flops are comple-
mented. The maximum delay is 10 × 3 ns = 30 ns.
+539 = 0010 0001 1011 So, the maximum frequency is 109/40 = 25 MHz.
−539 = 1101 1110 0101
= (DE5)16 29. (b) Solving for n in terms of k, we obtain
2k - 1 - k ≥ n
23. (b)
0.25 × 2 = 0.50 0 For data word length (n) = 16 bit we require 5 bit.
0.50 × 2 = 1.00 1 30. (d) 256/32 = 8 chips
Chapter 1.indd 54 4/9/2015 9:41:23 AM

UNIT II: COMPUTER ORGANIZATION AND ARCHITECTURE
6
Number of questions
4 Marks 1
Marks 2
3 Total number of questions
2
0
2015 2014 2013 2012 2011 2010
Year Concepts
2015 Cache memory, Memory, Pipelining, Main memory, Pipeline, Anti-dependency,
Instruction format, Byte Addressable memory, Pipeline, Three-address code
2014 Memory access, pipelining
2013 Cache memory, hard disk, pipelining, registers, micro-operations
2012 Pipelining, ROM, cache, file system
2011 Cache memory, interrupts
2010 Memory, pipelining, registers, cache memory
Chapter 2.indd 55 4/9/2015 9:49:42 AM

Chapter 2.indd 56 4/9/2015 9:49:42 AM
CHAPTER 2
COMPUTER ORGANIZATION AND ARCHITECTURE
Syllabus: Machine instructions and addressing modes, ALU and data path, CPU control design, memory interface,
I/O interface (interrupt and DMA mode), instruction pipelining, cache and main memory, secondary storage
2.1 INTRODUCTION 2.2 COMPUTER ARCHITECTURE
Computer architecture and organization is the 2.2.1 Register Set

science of interconnecting hardware components, design-
ing and configuring the hardware/software interface to The computer needs registers for processing and manip-
fulfill functional and performance goals of a computer. ulating data and for holding memory addresses that
This chapter outlines the basic hardware structure of a are available to the machine-code programmer. Some
modern digital programmable computer, the basic laws registers for a basic computer are given in Table 2.1.
Table 2.1 | Types of registers and their functions

for performance evaluation, designing the control and
data path hardware for a processor, concept of pipelining
for executing machine instructions simultaneously and Register Register Function
designing fast memory and storage systems. Symbol Name
Computer architecture deals with the structure DR Data register Holds memory operand
and behaviour of computer system as viewed by the
user. It encompasses instruction formats, the instruction ACC Accumulator Special purpose
set architecture (ISA) and addressing modes. processor register
Computer organization deals with the operation AR Address Holds address for
and interconnection of the various hardware components. register memory
(Continued)
Chapter 2.indd 57 4/9/2015 9:49:42 AM

58 Chapter 2: Computer Organization and Architecture
Table 2.1 | Continued When the system consists of multiple frequent cases,
where i is the number of frequent cases:
Register Register Function
−1
Symbol Name 
( ) Fi 
Soverall =  1 − ∑ Fi + ∑ 
 S
IR Instruction Holds an instruction
register that is to be executed
Problem 2.1: Consider a hypothetical processor used
PC Program Holds address of
in mathematical model simulation. It consists of two
counter instruction to be
functional units, floating point and integer. The float-
executed next
ing point is enhanced then it runs two times faster,
TR Temporary Holds temporary data but only 10% of the instructions are floating point.
register if required What is the speed up?
INPR Input register Holds input character
Solution: Here S = 2, F = 0.1
OUTR Output register Holds output character
−1
 0.1 
Soverall = (1 − 0.1) +
2 
= 1.052
2.2.2 Quantitative Principles to Design 
High-Performance Processor
Amdahl’s law focused on performance gain after enhanc- 2.3 MACHINE INSTRUCTIONS AND
ing the system. The performance gain is denoted by ADDRESSING MODES
Soverall and ET stands for execution time.
Performance of the system with enhancement
Soverall = Machine instruction is an individual machine code. The
Performance of the system without enhancement complete set of all machine codes recognized by a partic-
1 ETnew ular processor makes its Instruction Set. Instructions can
Soverall = (2.1) be grouped according to the function they perform. The
1 ETold
number of ways by which arguments for these machine
ETold instructions can be specified constitutes the addressing
Soverall =
ETnew modes for a processor.
After enhancement, the system consists of two portions:
unenhanced and enhanced portion. 2.3.1 Machine Instructions
ETnew = ET of the unenhanced portion An instruction is a command to the microprocessor to

+ ET of enhanced portion perform a given task. Most computer instructions are
classified as follows:
To calculate ETnew, the following two factors are needed:
1. Data transfer instructions: These instructions
1. Fractionenhance (F): It indicates how much por- move data from one place to another in the com-
tion of the old system undergoes enhancement. puter without changing the data content. Example:
2. Speedenhance (S): It indicates how many times the LOAD, MOVE, IN, OUT, PUSH, STORE.
new portion is running faster than the old portion. 2. Data manipulation instructions: These instruc-
Performancenew F 1 ETnew F ETold F tions perform arithmetic, logical and shift opera-
S= = = tions on data. Example: ADD, SUB, MUL, DIV,
Performanceold F 1 ETold F ETnew F
INC, AND, XOR, OR, SHR, SHL, ROR, ROL.
ETold F 3. Program control instructions: These instruc-
So, ETnew F =
S tions may change the address value in program coun-
On the basis of the above factor, ter and cause the normal sequential flow to change.
ETold F
ETnew F = ETold (1 − F ) +
On the basis of the number of address fields in an instruc-
S tion, they are classified as follows:
Substitute the value of ET in Eq. (2.1): 1. Three-address instruction: Computer with
ETold F three-address instruction format can use each
ETnew F = ETold (1 − F ) + address field to specify two sources and a destina-
S
Let ETold = 1, tion, which can be either a processor register or a
−1 memory operand. It results in short program but
1  F
= (1 − F ) +  requires too many bits to specify three addresses.
(1 − F ) + (F S ) 
Soverall =
S
Example: ADD R1, A, B (R1 ← M[A] + M[B])
Chapter 2.indd 58 4/9/2015 9:49:45 AM

2.3 MACHINE INSTRUCTIONS AND ADDRESSING MODES 59
2. Two-address instruction: Each address field can 1. provide programming flexibility to users through
specify either a processor register or a memory word. use of pointers to memory, counter for loop control,
OV R1, A (R1 ← M[A]);
Example: M data indexing and program relocation.
MUL R1, R2 (R1 ← R1*R2) 2. reduce the size of the addressing field of the
instruction.
3. One-address instruction: It used an implied
accumulator (AC) register for all data manipula- Let us suppose [x] means contents at location x for all the
tion. The other operand is in register or memory. addressing modes.
OAD A (AC ← M[A]);

Example: L
2.3.2.1 Types of Addressing Modes
ADD B (AC ← AC + M[B])
4. Zero-address instruction: A stack organized 1. Implied mode: In this mode, the operands are
computer does not use an address field for the implicitly stated in the instruction. For example,
instruction ADD and MUL. register reference instructions such as CMA (comple-
ment accumulator), CLA (clear accumulator) and
zero-address instructions that use stack organization.
Problem 2.2:
2. Immediate mode: In this mode, the operand is
(a) ISA of a processor consists of 64 registers, 125 specified in the instruction itself, that is, address
instructions and 8 bits for immediate mode. In a field is replaced by an actual operand. Immediate
given program, 30% of the instructions take one mode instructions are useful for initializing regis-
input register and have one output register, 30% ters to a constant value. For example, used for ini-
have two input registers and one output register, tializing CPU registers to some constant value such
20% have one immediate input, and one output as MOV R1, #34.
register, and remaining have two immediate
Instruction with immediate mode
input, 1 register input and one output register.
Calculate the number of bits required for each Opcode Data Operand
instruction type. Assume that the ISA requires
that all instructions be a multiple of 8 bits in
3. Register (direct) mode: In this mode, the oper-
length.
ands are in CPU registers. An n-bit register field
(b) Compare the memory space required with that of
can specify any one of 2n registers. Example: ADD
variable length instruction set.
R1 will add the contents of an accumulator and
Solution: contents of R1, that is, ACC = [ACC] + [R1].
(a) Since there are 125 instructions so we need 7 bits Instruction with register direct mode
to differentiate them as 64 < 125 < 128. For 64
registers, we need 6 bits and 8 bits for immediate Opcode Register address
mode.
For Type 1, 1 reg in, 1 reg out: 7 + 6 + 6 = CPU register
19 bits ~ 32 bits
For Type 2, 2 reg in, 1 reg out: 7+ 6 + 6 + 6 = Operand
26 bits ~ 32 bits
For Type 3, 1 imm in, 1 reg out: 7 + 6 + 8 = 4. Register (indirect) mode: In this mode, the
21 bits ~ 32 bits instruction format specifies a CPU register which
For Type 4, reg in, 2 imm in, 1 reg out: 7 + 6 + contains an effective address of the operand resid-
8 + 8 + 6 = 35 bits ~48 bits ing in memory. This mode ensures less number of
(b) As the largest instruction type requires 48 bit bits to specify a register value than to specify a
instructions, the fixed-length instruction format memory location. Example: ADD @ R1 will add the
uses 48 bits per instruction. Variable length instruc- contents of an accumulator with contents of the
tion format uses 0.3 × 32 + 0.3 × 32 + 0.2 × 32 + register R1, that is, AC = [ACC] + [[R1]].
0.2 × 48 = 36 = bits on average, that is, 25% less
Instruction
Instruction with
with register
register indirect
indirect mode
mode
space.
Opcode
Opcode Register address
Register address
2.3.2 Addressing Modes Memory
Memory
The addressing mode specifies how effective address of CPU register

CPU register
an operand is calculated from an instruction. Computers Pointer to
Pointer to Operand
Operand
use various addressing mode techniques to: operand
operand
Chapter 2.indd 59 4/9/2015 9:49:47 AM

5. Auto-increment or Auto-decrement mode: 10. Base register addressing mode: In this mode,
This is similar to register indirect mode except the the effective address of an operand is obtained by
register containing effective address is incremented adding the content of a base register to the address
or decremented after (or before) its value is used to part of the instruction. This is somewhat similar to
access memory. the indexed addressing mode except that the base
6. Direct address mode: In this mode, the effec- register stores base or beginning address instead of
tive address of an operand is equal to the address an index register. It is used for program relocation.
part of the instruction. Example: ADD A instruc-
tion adds content of memory cell A to accumula-
tor, that is, ACC = [ACC] + M[A]. Problem 2.3: A two-word instruction LOAD is stored
Instruction with direct address mode at location 300 with its address field in the next loca-
tion. The address field has value 600 and value stored
at 600 is 500 and at 500 is 650. The words stored
Opcode Memory address
at 900, 901 and 902 are 400, 401 and 402, respec-
Memory tively. A processor register R contains the number
800 and index register has value 100. Evaluate the
effective address and operand if addressing mode of
the instruction is as follows:
Operand
1. Direct 4. Immediate
2. Indirect 5. Register indirect
7. Indirect address mode: In this mode, memory 3. Relative 6. Index
address specified by address field contains the Solution: Memory layout is as follows
address of (pointer to) the operand. Example:
ADD @ A will add the contents of the memory cell 300 LOAD
A, that is, ACC = [ACC] + M[M[A]]. 301 600
Instruction with indirect address mode
500 650
Opcode Memory address
Memory
600 500
Pointer to operand
700 900
800 700
Operand
900 400
901 401
8. Relative address mode: In this mode, the effective 902 402
address of an operand is obtained by adding the con-
tent of a program counter to the address part of the Addressing Effective Operand
instruction. The address part of the instruction can be Mode Address
either positive or negative represented in 2’s comple-
ment. The result obtained after adding the content of Direct 600 500
the program counter to the address field produces an Indirect 500 650
effective address whose position in memory is relative Relative 902 402
to the address of the next instruction. Immediate 301 600
9. Index address mode: In this mode, the effective Register indirect 800 700
address of an operand is obtained by adding the Index 700 900
content of an index register to the address part of
the instruction. The index register is a special CPU
register that stores an index value and the address
field of the instruction stores the base address of a Problem 2.4: A relative mode branch type instruc-
data array in the memory. The distance between tion is stored in memory at an address equivalent to
the base address and the address of the operand is decimal 600 and the branch is made to an address
the index value that is stored in the index register. equivalent to decimal 400. What is the value of the
The index register can be incremented to facilitate relative address field of the instruction (in decimal)?
access to consecutive operands stored in arrays
Solution: Relative address = 400 − 601 = −201
using the same instruction.
Chapter 2.indd 60 4/9/2015 9:49:48 AM

2.4 ARITHMETIC LOGIC UNIT 61
Table 2.2 | Arithmetic circuit function table

Select Input to Output of Binary Micro-Operation
Adder Y Adder
S1 S0 Cin
D = A + Y + Cin
0 0 0 0 D=A Transfer A
0 0 1 0 D=A+1 Increment A
0 1 0 B D=A+B Add
0 1 1 B D=A+B+1 Add with carry
1 0 0 B D = A+B Subtract with borrow
1 0 1 B D = A + B +1 Subtract
1 1 0 1 D=A−1 Decrement A
1 1 1 1 D=A Transfer A
2.4 ARITHMETIC LOGIC UNIT By controlling the output Y of multiplexers with two
selection inputs S1 and S0 and Cin either 0 or 1, we can
generate the eight arithmetic micro-operations (Table 2.2).
Arithmetic logic unit (ALU) is a combinational circuit
that performs all arithmetic and logic operations so that 2.4.2 Logic Micro-Operations
the entire register transfer operation from the source reg-
isters through the ALU and into the destination register Logic micro-operations such as AND, OR, Exclusive OR,
can be performed during one clock pulse period. etc., consider each bit of register separately and specify
binary operations for strings of bits (Table 2.3).
2.4.1 Arithmetic Micro-Operations
Table 2.3 | Types of micro-operations
The basic arithmetic micro-operations such as addition,
subtraction, increment, decrement and shift are performed Micro-operation Name
F←0
on numeric data stored in registers. The basic component
Clear
of arithmetic is parallel binary adder, and by controlling
the input to adder, different micro-operations can be F←A∧B AND
realized. Figure 2.1 depicts a 2-bit arithmetic circuit which F ←A∧B
includes two full-adder circuits and two multiplexers for
choosing different arithmetic micro-operations. There are F←A Transfer A
two 2-bit input numbers A and B and 2-bit output D. The F ←A∧B
F←B
two inputs from A go directly to X inputs of full adder.
Transfer B
The output of multiplexer goes to input Y of full adder.
F←A⊕B Exclusive OR
Cin F ← A ∨B OR
A0
S1
X0 C0 F ←A∨B NOR
S1 FA D0
S0 Y0 C1 F ← A⊕B Exclusive NOR
0 4×1
S0
B0 1 MUX F ←B Complement B
2
3
A1 F ←A∨B
X1 C1
S1 FA D1 F ←A Complement A
0 4×1
S0 Y1 C2
B1 1 MUX Cout F ←A∨B
2
3
0 1 F ←A∧B NAND
Figure 2.1 | A 2-bit arithmetic circuit. F ← all 1’s Set to all 1’s
Chapter 2.indd 61 4/9/2015 9:49:51 AM

Logical micro-operations are capable of manipulating 2.5 CPU CONTROL DESIGN

individual bits or a portion of word stored in CPU regis-
ters. Let us consider the data in a register A. In another
register, B is the operand that will be used to modify the Central processing unit (CPU), or the brain of a com-
contents of A using logic micro-operations. Some of the puter, performs the data processing operations. It consists
applications are as follows: of three major parts: register set that stores intermedi-
ate data during instruction execution, ALU performs the
1. Selective set operation: In this, If a bit in B is required micro-operations and control unit that super-
set to 1, that same position in A sets to 1, other- vises all other elements for the transfer of information
wise that bit in A retains its previous value. from one register to the other. The main function of a
1 1 0 0 At CPU is to fetch an instruction from the memory and exe-
1 0 1 0 B (To set some bits in A) cute it. CPU is divided into three types of organizations:
1 1 1 0 At +1(A ¬ A + B)
1. Single accumulator organization: In this, one
2. Selective complement operation: If a bit in operand is implied in the accumulator, a special
B is set to 1, that same position in A gets comple- purpose register, and the other operand is a register
mented from its original value, otherwise it remains or the memory. Example: ADD R1 (R1 ← AC +
unchanged. R1), LOAD A (AC ← A), STORE T (M[T] ← AC).
1 1 0 0 At 2. General register organization: In this, the
1 0 1 0 B (To complement some bits in A) CPU will have several general purpose regis-
0 1 1 0 At +1(A ← A ⊕ B) ters which lead to shorter and efficient programs
because registers are faster. Example: ADD R1, R2
3. Selective clear operation: If a bit in B is set to (R1 ← R1 + R2). Figure 2.2 shows bus organiza-
1, that same position in A sets to 0, otherwise it tion for three registers R1, R2 and R3. The output
remains unchanged. of these registers and one from the external input
is connected to two multiplexers A and B. The two
1 1 0 0 At
1 0 1 0 B (To clear some bits in A) External
0 1 0 0 At +1(A ← A ⋅ B ′) input
R1
4. Mask operation: If a bit in B is set to 0, that same R2
position in A sets to 0, otherwise it remains unchanged.
R3
1 1 0 0 At
1 0 1 0 B (To clear some bits in A)
At +1(A ¬ A × B)
SELA MUX MUX SELB
0 0 0 0
2×4
Decoder A bus B bus
5. Clear operation: If the bits in the same position
in A and B are the same, they are cleared in A, else
they are set in A. SELD OPR ALU
1 1 0 0 At
1 0 1 0 B
At+1 (A ¬ A Å B)
External output
0 1 1 0
Figure 2.2 | General register organization.
6. Insert operation: It is used to insert a specific bit
pattern into A register, leaving the other bit posi- select lines SELA and SELB from multiplexers A
tions unchanged. This is accomplished by two sub- and B select one of the input and feed to ALU.
operations: masking operation to clear the desired bit OPR specifies one of the possible operation codes
positions, followed by OR operation to introduce the that ALU will perform on the data inputs and the
new bits into the desired positions. Suppose you wanted output is transferred either to one of the registers
to introduce 10 into the low order two bits of A: using 2 × 4 decoder or to the external output say
1101 A (Original) and 1110 A (Desired) memory. The control word (Fig. 2.3) for the two-
operand instruction is as follows:
1 1 0 1 A (Original)
1 1 0 0 Mask 2 bits 2 bits 2 bits 4 bits
1 1 0 0 A (Intermediate)
SELA SELB SELD OPR
0 0 1 0 Added bits
1 1 1 0 A (Desired) Figure 2.3 | A control word.
Chapter 2.indd 62 4/9/2015 9:49:53 AM

2.5 CPU CONTROL DESIGN 63
3. Stack organization: Stack may consist of number repeated continuously for a complete program and is
of registers or a part of main memory in which data known as the fetch-execute cycle (Fig. 2.4). The fol-
items are stored in consecutive locations that are lowing steps are performed for executing an instruction:
accessed by LIFO (last in, first out) mechanism. As
there is limited number of registers, a part of memory Start
is implemented as stack for storage and retrieval of
intermediate data. Stack pointer (SP) keeps a track
of the top item of a stack. The process of inserting Load PC contents
a new item onto a stack is known as push accom- to MAR
plished by first incrementing stack pointer and then
inserting an item from the data register. Increment PC to
SP ← SP + 1 point to next
instruction
M[SP] ← DR
The process of removing an item from the top of a
Load the instruction
stack is known as pop performed by first transfer-
stored at MAR to IR
IR ← M[MAR]
ring data into DR and then decrementing SP.
DR ← M[SP]
SP ← SP − 1 Decode the
instruction
Problem 2.5: A system has CPU organized in the
form of general register organization consisting of 16 Load any data
registers, each storing 32-bit data. Assume the ALU required into MDR
has 35 operations.
(a) How many multiplexers are there in A bus and B
Check
bus, and what is the size of each multiplexer? Yes Set PC to value
for jump
(b) How many selection inputs are needed for MUX A from jump inset
instruction
and MUX B?
(c) How many inputs and outputs are there in a decoder?
(d) How many inputs and outputs are there in ALU No
for data, including input and output carries?
(e) Formulate a control word for the system. Execute the
instruction
Solution:
(a) 32 Multiplexers, each of size 16 × 1.
(b) 4 Inputs each, to select one of 16 registers.
(c) 4 to 16 − Line decoder Check for No
(d) 32 + 32 + 1 = 65 data input lines interrupts
(e) 32 + 1 = 33 data output lines
4 bits 4 bits 4 bits 6 bits Yes
SELA SELB SELD OPR Service the
interrupt
2.5.1 Instruction Execution Figure 2.4 | Instruction cycle.
A CPU generally executes one instruction at a time 1. Fetching the instruction: The next instruction
sequentially and a sequence of such instructions is is fetched from the memory address that is saved in
known as a program. The CPU executes the instructions the program counter, and memory content fetched
that reside in the main memory. In order to execute is stored in instruction register (IR). The program
an instruction, the CPU has to fetch the instruction counter then points to the next instruction that
first from the main memory into one of its registers. will be read in the next cycle.
It then decodes the instruction, that is, it decides what 2. Decode the instruction: During this cycle, the
the instruction intended to do, fetch operands required instruction inside the IR gets interpreted by the
and finally executes the instruction. This process is decoder.
Chapter 2.indd 63 4/9/2015 9:49:54 AM

3. Operand fetch: In case of a direct or indirect decoder generates a separate control line for each
memory instruction, the execution begins in the step in the control sequence. The encoder gets
next clock cycle. If the instruction has an indirect its input signal from the decoder, step decoder,
address, the effective address of the operand is read external input and condition codes and generates
from the main memory, and the required data is individual control signals. It is faster and more
fetched from the memory into memory data regis- efficient but less flexible and is difficult to add
ters. If the instruction has direct address, nothing new feature or correct mistakes in original design.
is done at this clock cycle.
4. Execute the instruction: The control unit of Clock Control step Reset
the CPU passes the instruction decoded by decoder counter
as a sequence of control signals to the different
functional units of the CPU to execute the tasks
required by the instruction such as reading values Step decoder
from registers or input devices, performing mathe-
T1 T2 Tn
matical or logic micro-operations by ALU, and writ- I1
ing the result back to a register or main memory. External
I2
inputs
Instruction
2.5.2 CPU Data Path IR Encoder
decoder
In Condition
CPU contains data paths that are responsible for routing codes
data between the functional units of a computer. The
following are the different data path structures available End
for routing: Control
1. Single bus structure: In this architecture, all CPU signals
Figure 2.5 | Block diagram of hardwired control unit.
registers are connected to the same bus. Data can be
transferred either between CPU registers or between
CPU register and ALU at a given clock pulse. The 2. Micro-programmed control: Control signals
speed of operation is slow as only one operand can be are generated by using programming known as
transferred in one clock cycle and addition operation micro-programs that constitutes micro-instructions
(R1 ← R2 + R3) occurs in three clock cycles. (control word) (Fig. 2.6). Memory that is part
2. Two bus structure: All general purpose CPU
registers are connected to both buses say bus A and IR External
Sequences inputs
bus B; but special purpose registers are divided into (starting and branch
two groups, say group 1 connecting bus A to pro- address generator) Condition
gram counter and one input of ALU and group 2 codes
connecting bus B to MDR (Memory Data Register)
and other input of ALU. The two operands are Control address
transferred to ALU in 1 clock cycle and the addition Clock
register
operation (R1 ← R2 + R3) occurs in 2 clock cycles. Address
3. Three bus structure: The performance can be
further be improved by using three buses such that Control
Read
addition operation (R1 ← R2 + R3) can occur in command memory
one clock cycle.
Control word
2.5.3 Control Unit Design Micro instruction

register
Control unit is considered as brain of a CPU that con-
trols various units in the data path. The performance of
control unit is important as it determines the clock cycle Decoder
of the processor. Control unit can be designed either by
hardwired or by microprogram. Control Control
1. Hardwired control: Control unit is made up signals signals to
of sequential and combinatorial circuits to gener- within CPU system bus
ate the control signals and interpret instructions
(Fig. 2.5). The instruction decoder decodes the Figure 2.6 | Block diagram of micro-programmed
instruction loaded in instruction register. The step control unit.
Chapter 2.indd 64 4/9/2015 9:49:55 AM

2.6 I/O INTERFACE (INTERRUPT AND DMA MODE) 65
of CPU is known as control memory and stores Table 2.4 | RISC versus CISC
micro-instructions. The micro-program sequencer
generates the address of micro-instruction accord- RISC (Reduced CISC (Complex
ing to instruction stored in instruction register. Instruction Set Instruction Set
The address of micro-instruction to be executed Computers) Computers)
is available in content addressable register. Micro- Rich register set Less number of registers
program sequencer issues read command to read
micro-instruction from control memory into micro- Supports less addressing Supports more number of
instruction register which on execution generates modes addressing modes
control signals for various parts of a processor. This Supports fixed length Supports variable length
control unit design is more flexible to accommodate instruction instruction
new features and less error prone but quite slower Successful pipeline with Unsuccessful pipeline
than the hardwired unit. one instruction per cycle
The format of the control word is Example: ARM, Example: Pentium
Motorola processors
Branch Flag Control Control memory
condition signal address
On the basis of the type of control word supported, it is 2.6 I/O INTERFACE (INTERRUPT
divided into two types: AND DMA MODE)
1. Horizontal micro-programmed control unit:
In this design, the control signals are represented in I/O interface bridges the differences between CPU and
the form of 1 bit per control signal and it supports peripheral devices and provides a method for transfer-
longer control word. ring information between internal storage and external
2. Vertical micro-programmed control unit: In I/O devices. There are the following three modes of I/O
this design, the control signal is represented by using transfer:
encoding format.
1. Programmed I/O: The I/O device does not
have direct access to memory. It requires execution
Problem 2.6: Consider a control unit which has 1024 of several instructions by the CPU and the CPU
control word memory; it supports 48 control signals has to wait for the I/O device to be ready for either
and 8 flag conditions. What is the size of the control reception or transmission of data.
word in bits and control memory in bytes? 2. Interrupt initiated I/O: In this, instead of
Solution: waiting, the control is transferred from a currently
running program to another service program as a
(a) Using horizontal programmed control unit result of an external/internal generated request.
0 bits 3 bits 48 bits 10 bits ••Hardware interrupts: These interrupts are
Branch Flag Control Control present in the hardware pins.
condition signal memory ••Software interrupts: These are the instruc-
tions used in the program whenever the required
Size of control word = 61 bits functionality is needed.
Control memory = (1024 × 61)/8 = 128 × 61 bytes ••Maskable interrupts: These interrupts may
(b) Using vertical programmed control unit be enabled or disabled explicitly.
••Non-maskable interrupts: These interrupts
0 bits 3 bits 48 bits 10 bits are always there in the enable state. We cannot
Branch Flag Control Control disable them by explicit conditions (flags).
condition signal memory ••Vectored interrupts: These interrupts are
log 48 ~ 6 bits associated with the static vector address.
••Non-vectored interrupts: These interrupts
Size of control word = 19 bits
Control memory = (1024 × 19)/8 = 128 × 19 bytes are associated with dynamic vector address.
••External interrupts: These interrupts are
generated by external devices such as I/O.
2.5.4 RISC versus CISC Processors ••Internal interrupts: These devices are gener-
ated by the internal components of the processor
The differences between reduced and complex instruc- such as temperature sensor, power failure, error
tion set computers is given in Table 2.4 instruction, etc.
Chapter 2.indd 65 4/9/2015 9:49:55 AM

••Synchronous interrupts: These interrupts block in memory is given by the address register,
are controlled by the fixed time interval. All and the length of the bytes to transfer is given by
the interval interrupts are called as synchronous the word count register. The controller decrements
interrupt. a word counter each time it moves a data byte.
••Asynchronous interrupts: These interrupts
There are several modes of operation of DMA:
are initiated based on the feedback of previ-
ous instructions. All the external interrupts are ••Burst or block transfer mode: In this mode,
called as asynchronous interrupt. the entire block of data is transferred once the
3. Direct memory access (DMA): It is one of DMA controller is granted access to the system
several methods for coordinating the data transfers bus by the CPU. The bytes of data in the block are
between an I/O device and the core processing unit transferred before releasing control of the system
or memory in a computer. It refers to transfer of buses back to the CPU. The only disadvantage of
data directly between a fast storage device and this mode is that it renders the CPU inactive for
memory bypassing CPU because of its limited some long periods of time.
speed. DMA provides a significant improvement ••Cycle stealing mode: In this mode, the DMA con-
in terms of latency and throughput as it allows troller obtains access to the system buses like burst
the I/O device to access the memory directly, mode; but after one byte of data transfer, the control
without using the processor. There are certain of the system bus is released back to the CPU via
advantages of using DMA for data transfer: BG. It is then continually requested again via BR,
••DMA saves processor’s MIPS as the core can transferring one byte of data per request, until the
operate in parallel. entire block of data has been transferred. This mode
••DMA saves power because it requires less cir- is suitable for the systems in which the CPU cannot
cuitry than the processor to transfer data. be disabled for the considerable length of time as in
••DMA has no modulo block size restrictions. burst transfer modes such as for controllers moni-
Direct memory access (DMA) controller takes over toring the data in real time. The advantage is that
the control of buses to manage the transfer directly CPU is not idled for as long as in burst mode, but
between the I/O device and memory. Bus request the data block is not transferred as quickly.
••Transparent mode: It is the slowest yet more ef-
(BR) and Bus grant (BG) signals are used by the
DMA controller to request the CPU to relinquish ficient data transfer mode in terms of overall system
control of the buses and get the control of system performance. The DMA controller transfers data only
buses (Fig. 2.7). The DMA controller consists of when the CPU is busy in performing operations that
3 different registers: an address register, a control do not use the system buses. So, the CPU never stops
register and a word counter register. To transfer a executing its programs but the biggest disadvantage
block of data between an I/O device and memory, is complex hardware circuitry that needs to deter-
the controller stores initial values in the address mine when the CPU is not using the system buses.
register. The DMA channel then transfers the A DMA read transfers data from the memory to
block of information from or to memory according the I/O device, while DMA write transfers data
to the control register. The starting address of the from an I/O device to memory. The functional
Address behaviour of a DMA transfer outlined in Fig. 2.8:
bus ••TheCPU transmits the following information to a
Data DMA controller:
bus Data bus (a) beginning address in memory which is stored in
Address bus
buffer buffer address register in DMA controller.
DMA (b) Number of words to transfer which is stored in
select DS
Register Address word count register in DMA Controller.
select RS register (c) direction (memory-to-I/O device or I/O device-
Read RD to-memory), port ID, DMA mode of transfer
Internal bus
Word count and end of block transfer either through inter-

Write WR register rupt request or no interrupt request which is
Bus
request BR stored in control register as command word.
Control
Bus register ••The processor them relinquishes control of address,
BG
grant data and control buses to DMA Controller and
Interrupt Interrupt I/O returns to other processing activities while the
device
DMA controller starts the data transfer between
Figure 2.7 | Block diagram of the DMA controller. I/O device and memory.
Chapter 2.indd 66 4/9/2015 9:49:56 AM

2.7 INSTRUCTION PIPELINING 67
Interrupt
BG Random Access
CPU
Memory (RAM)
BR
RD WR Address Data RD WR Address Data
Read control
Write control
Address
select Address bus
Data bus
RD WR Address Data
DS DMA acknowledge
RS Direct Memory I/O

Access (DMA) peripheral
DMA request
BR controller device
BG
Interrupt
Figure 2.8 | DMA controller interconnection with memory, CPU and I/O devices.
••When the DMA controller accesses memory, it execution time of a set of instructions and there is no
synchronizes this memory request with an idle need to wait of the most part of the processor circuits for
period of the processor, thus disabling the pro- the other parts of the processor to complete their part
cessor, or requesting a halt of the processor, and of execution. Pipeline speed is limited by the slowest
awaits an acknowledgement. pipeline stage.
••After the completion of the block transfer, the DMA Throughput of a processor is the rate at which opera-
controller either raises an interrupt request if the tions get executed. Latency is the amount of time that a
interrupts are enabled or indicates the completion single operation takes to execute. In an unpipelined com-
in its status register and the processor recognizes puter, throughput = 1/latency, as each operation exe-
I/O completion (either by interrupt signal or by cutes by itself and for pipelined computer, throughput
reading the status register) and gets its system > 1/latency, since execution of instruction is overlapped.
buses back and normal processing starts. The Consider a k-segment pipeline with a clock cycle time
device has to initiate a new data transfer through Tp used to execute n tasks (Fig. 2.9). An equivalent non-
DMA request signal which is again acknowledged pipelined system takes Tn time to complete each task.
by CPU through DMA acknowledge signal via The speed up of a pipelined system over a non-pipelined
DMA controller. system is given by the following relation:
n × Tn
S=
2.7 INSTRUCTION PIPELINING (k + n - 1) × Tp
Theoretically, maximum speed up that a pipelined

In early computers, each instruction completely finished system can achieve is given by the following equation:
before the execution of the next one began. The hard-
kTp
ware circuits needed to perform different operations S= =k
of an instruction cycle are different and most part of Tn
these processor circuits are idle at a given moment of Pipelining Hazards: These hazards reduce the
time. These processor circuits wait for the other parts ideal speed up gained by pipelining by preventing the
of the processor to complete its part of execution first. next instruction in the sequence from being executing
Instruction pipelining is a technique for overlapping the during its designated clock pulse. Hazards forces the
execution of several instructions to reduce the overall pipeline to be stalled. There are three types of hazards:
Chapter 2.indd 67 4/9/2015 9:49:57 AM

Fetch an Decode Fetch an Execute Write

instruction instruction operand instruction result back
One clock cycle

(a)
Write
Fetch an Fetch an Execute
Decode result
instruction operand instruction
back
Pipeline latch / overhead/delay
One clock One clock One clock One clock One clock
cycle cycle cycle cycle cycle
Pipeline Stages
(b)
step → 1 2 3 4 5 6 7 8
Segment ↓ Fetch I1 I2 I3 I4
Decode I1 I2 I3 I4
Fetch operand I1 I2 I3 I4
Execute I1 I2 I3 I4
Write back I1 I2 I3 I4
(c)
Figure 2.9 | (a) Unpipelined processor. (b) Pipelined five-stage processor.
(c) Timing diagram of a five-stage instruction pipeline.
1. Structural hazards: These result from resource instruction refers to a result which is yet not
conflicts when the hardware cannot support been calculated, that is, in this inst2 tries to
instructions that need simultaneous execution in read a source before inst1 writes to it. This
pipeling. situation arises if the read operation by instruc-
2. Data hazards: They arise when an instruction tion takes place before write done by other in-
depends on the result of a previous instruction and struction. For example,
that result is not yet calculated. inst1: R3 <-R1 + R2
There are three situations in which data hazards inst2: R4 <-R3 + R2
can occur: The first instruction calculates a value by adding
••Read
values in registers R1 and R2 and saves the result
after write (RAW), a true dependency
in register R3, and the second instruction uses
••Write after read (WAR), an anti dependency this saved value to calculate a result for regis-
••Write after write (WAW), an output dependency ter R4. However, in a pipeline, when operands
for the second operation are fetched, the results
Consider two instructions inst1 and inst2
from the first instruction will not have been
occurring, with inst1 occurring before inst2 in
saved yet, and so there arises a data depend-
the program order.
ency. It can be said that there is a data depend-
••Readafter write (RAW): A read after write ency with instruction inst2, as it is dependent
(RAW) data hazard is a situation in which an on the completion of instruction inst1.
Chapter 2.indd 68 4/9/2015 9:49:58 AM

••Write after read (WAR): A write after 3. Control hazards: They arise from the pipelining
read (WAR) data hazard refers to a situation in of branches and other instructions that change the
which there is a problem with concurrent execu- value of PC.
tion, that is, inst2 tries to write a destination Speed up from pipelining
before it is read by inst1. This situation arises
if write operation completes first by instruction Average instruction time unpipelined
=
before the read operation takes place by other Average instruction time pipelined
instruction. For example,
Speed up from pipelining
inst1: R4 <-R1 + R3
CPI unpipelined × Clock cycle pipelined
inst2: R3 <-R1 + R2 =
CPI pipelined × Clock cycle pipelined
If a situation arises in which there is a chance
that inst2 may get completed before inst1 CPI unpipelined
Ideal CPI =
(i.e., with concurrent execution) we must note Pipeline depth
that we do not store the result of register R3
before inst1 has had a chance to fetch the Speed up from pipelining
operands. Ideal CPI × Pipeline depth × Clock cycle unpipelined
=
••Write after write (WAW): A write after CPI pipelined × Clock cycle pipelined
write (WAW) data hazard refers to a situation
in which there is a concurrent execution envi- Speed up from pipelining
ronment, that is, inst2 tries to write an oper- Ideal CPI × Pipeline depth × Clock cycle unpipelined
and before it is written by inst1.This situation
(Ideal CPI + Pipeline stall) × Clock cycle pipelined
=
arises if write operation by an instruction occurs
in the reverse order of the intended sequence. Assuming ideal CPI as 1, speed up is:
For example,
Speed up from pipelining
inst1: R2 <-R1 + R3
inst2: R2 <-R4 + R5 Pipeline depth × Clock cycle unpipelined
(1 + Pipeline stall) × Clock cycle pipelined
=
The WB (write back) of inst2 must be delayed
until the execution of inst1. where CPI is cycles per instruction.
Problem 2.7: Consider a four-stage pipeline processor. The number of cycles needed by the four instructions I1, I2,
I3 and I4 in stages instruction fetch, decode, operand fetch and execute are shown below. Assume I2 is the branch
instruction. Draw the timing space diagram.
S1 S2 S3 S4
I1 2 1 1 1
I2 1 2 3 1
I3 1 1 1 2
I4 2 1 3 1
Solution:
STEP → 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17
Fetch I1 I1 I2 I3 - - - - - I3 I4 I4
Decode I1 I2 I2 - - - - - I3 - I4
Operand Fetch I1 I2 I2 I2 - - - I3 - I4 I4 I4
Execute I1 - - - I2 - - - I3 I3 - - I4
Chapter 2.indd 69 4/9/2015 9:49:59 AM

Problem 2.8: Assume a simple 5-stage pipeline (IF, ID, E, DF, W) each stage takes a single cycle. Assuming there
are no cache misses. How many cycles would the following code take to execute if there is no special hardware to
improve performance in the presence of hazards?
MOV edx,[ecx+100]
MOV ebx,[ecx+104]
ADD edx,ebx
MOV [ecx+108],ebx
MOV eax,[ecx+100]
ADD ebx,eax
Solution: The above code takes 14 cycles to execute, as shown below:
1 2 3 4 5 6 7 8 9 10 11 12 13 14
IF ID DF E W
IF ID DF E W
IF ID DF stall E W
IF ID stall DF stall W
IF ID stall DF stall stall E W
IF ID stall DF stall stall stall E W
Problem 2.9: In the below figure, calculate the total execution time after which the result of the fourth task enter-
ing the pipe above ready?
IF ID EX MEM WB
5 ns 5 ns 10 ns 10 ns 5 ns
Solution:
5 10 15 20 25 30 35 40 45 50 55 60 65
Inst1 IF ID EX EX MEM MEM WB
Therefore, the total execution time is 65 ns.
Problem 2.10: What is the mean overhead of a pipe- Problem 2.12: Calculate the time required to perform
line with 8 stages and an execution time per stage of 1000 operations in a 6-staged pipeline with an execu-
2 ns? tion time of 3 ns per stage?
Solution:
Solution: The mean overhead = (Stages - 1) ×
Execution time per stage = (8 - 1) × 2 = 7 × 2 = 14 ns Tp = (k - 1 + n) × T = (6 - 1 + 1000) × 3 = 3.015 µs
Problem 2.11: How many stages has a pipeline that Problem 2.13: Calculate the mean overhead of a pipeline
achieves a speed of 9.9 for 100 operations? with 7 stages and an execution time per stage of 2 ns?
Solution: Solution: Mean overhead of pipeline =

n ×k n × 90 (k × Tp - Tn )
Speed = ⇒ 9.9 = ⇒ n = 11 = (k - 1) × T = (7 - 1) × 2 = 12 ns
k -1 + n (90 - 1) + n k
Chapter 2.indd 70 4/9/2015 9:50:01 AM

Problem 2.14: Consider a pipeline with 5 stages: IF, ID, EX, M and W. Assume that each stage requires one clock
cycle. Show how the following program segment for adding 2 arrays is processed and compare the clock cycles
needed in non-pipelined system with pipelined system when result of the branch instruction i.e. content of is avail-
able after WB stage.
LOAD R4 #400
L1: LOAD R1, 0 (R4);
LOAD R2, 400 (R4);
ADD R3, R1, R2;
STORE R3, 0 (R4);
SUB R4, R4, #4;
BNEZ R4, L1;
Solution: Number of cycles = [Initial instruction + (Number of instructions in the loop L1) × Number of loop
cycles] × Number of clock cycles/instruction (CPI)
= [1 + (6) × 400/4] × 5 = 3005
Timing diagram for one loop iteration in a pipelined system is as follows:
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15
LOAD R4 #400 IF ID EX M W
LOAD R1, 0 (R4) IF ID EX M W
LOAD R2, 400 (R4) IF ID stall stall EX M W
ADD R3, R1, R2 IF ID stall stall EX stall M W
STORE R3, 0 (R4) IF ID stall DF stall stall E W
SUB R4,R4, #4 IF ID stall Ex M W
BNEZ R4, L1 IF stall ID stall stall EX M W
Number of cycles in the loop = 15

Number of clock cycles for segment execution on pipelined processor
= 1 + (Number of clock cycles in the loop L1) × Number of loop cycles
= 1 + 15 × 400/4 = 1501
Number of Clock cycles for the program execution on non-pipelined processor
Speedup =
Number of Clock cycles for the segment execution on pipelined processor
3005
= = 2 times
1501
Problem 2.15: Consider a 5-stage pipeline with stages: For all following questions we assume that: (a) Pipeline
contains stages: IF (Instruction Fetch), IS (Issue), FO (Fetch operand), E (Execute) and W (Write). (b) Each stage
except E requires one clock cycle and system has 4 Functional Units for floating point operations, FP load/store,
FP addition/subtraction, FP multiplication and FP division, (c) Execution stage for Load/Store operations requires
1 clock cycle, for ADD or SUB operations requires 1 clock cycle, for MUL operation requires 3 clock cycles and for
DIV operation requires 4 clock cycles. All memory references hit in cache. Pipeline has forwarding circuitry for all
FUs, except FP-Load/Store where operand is ready after W-stage.
Chapter 2.indd 71 4/9/2015 9:50:01 AM

Timing diagram of is presented below:
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16
LOAD F6, 20(R5) IF IS FO E W
LOAD F2, 28(R5) IF ISD FO E W
MUL F0, F2, F4 IF IS stall stall FO E E E W
SUB F8, F6, F3 IF IS FO E W
DIV F10, F0, F6 IF IS stall stall stall stall FO E E E E W
ADD F6, F8, F2 IF IS FO E W
STORE F8, 50(R5) IF IS FO E W
Identify the hazards in the following instructions from the following list (Structural, Data, Control, RAW, WAR,
WAW, None)
1. MULT F0, F2, F4 and STORE F8, 50(R5)
2. DIV F10, F0, F6 and ADD F6, F8, F2
3. MULT F0, F2, F4 and DIV F10, F0, F6
4. DIV F10, F0, F6 and ADD F6, F8, F2
Solution: 1. Structural; 2. Data; 3. RAW; 4. WAR.
2.8 MEMORY HIERARCHY and ROM (read only memory). Integrated RAM chips
are available in two modes:
The storage media can be categorized in hierarchy accord- 1. Static RAM: It stores the binary information in
ing to their speed and cost (Fig. 2.10). As we move down flip flops and information remains valid until power
the hierarchy, access time increases and cost per bit is supplied. It has faster access time and is used in
decreases. implementing cache memory.
2. Dynamic RAM: It stores the binary information
as a charge on the capacitor. It requires refreshing
circuitry to maintain the charge on the capacitors
CPU after few milliseconds. It contains more memory
registers cells per unit area as compared to SRAM.
Decreasing Cache Increasing
cost and memory cost and
speed Main memory speed 2.8.1.1 Memory Interfacing
Magnetic disks If the required memory for the computer is larger
Increasing Decreasing
size size than the capacity of one chip, it is necessary to
Magnetic tapes connect multiple RAM and ROM chips to a CPU
Figure 2.10 | Memory hierarchy.

through the data and address buses (Fig. 2.11). The
low-order address bus lines select the word within a
chip and other lines select a particular chip through
2.8.1 Main Memory its chip select inputs. Assume a computer system
needs 256 bytes of RAM and 512 bytes of ROM. The
It is the central storage unit that directly communicates configuration of RAM chip is 128 × 8 and ROM chip
with the CPU. It is designed using semiconductor- is 512 × 8. The RAM and ROM chips required are
integrated circuits and needs constant power supply to as follows:
maintain the information. It is expensive as compared to
auxiliary storage so it has limited capacity. Example: R/W Number of RAM chips = 256/128 = 2
(read/write) memory or RAM (random access memory) Number of ROM chips = 512/512 = 1
Chapter 2.indd 72 4/9/2015 9:50:02 AM

2.8 MEMORY HIERARCHY 73
The memory interconnection is depicted in the following diagram:
Chip select 1 CS1

Chip select 1 CS1
Chip select 2 CS2
128 × 8 Bidirectional Chip select 2 CS2 512 × 8
Data bus in
Read RD output mode only
RAM chip data bus ROM chip
Write WR
Address bus AD7 Address bus AD9
(a) (b)
Figure 2.11 | (a) RAM chip. (b) ROM chip.
Problem 2.16: A computer employs RAM chips of 256 × 8 and ROM chips of 1024 × 16. The computer system needs
2K bytes of RAM and 4K bytes of ROM and four interface units each with four registers. Draw a memory address
map for the system and give the address range in hexadecimal for RAM and ROM chips.
Solution: RAM 2048/256 = 8 chips; 2048 = 211; 256 = 28

ROM 4096/1024 = 4 chips; 4096 = 212; 1024 = 210
Interface 4 × 4 = 16 registers; 16 = 24
Component Address 16 15 14 13 12 11 10 9 8 7 6 5 4 3 2 1

RAM 0000-O7FF 0 0 0 0 0 ↔ x x x x x x x x
3×8
decoder
ROM 4000-4FFF 0 1 0 0 ↔ x x x x x x x x x x
2×4
decoder
Interface 8000-800F 1 0 0 0 0 0 0 0 0 0 0 0 x x x x
2.8.2 Secondary Memory coated with magnetized material on both sides.

Multiple disks are stacked over one another on the
Secondary memory, also known as auxiliary memory or spindle with read/write heads on each surface. Bits
external memory, can store a large amount of data at are stored as spots on magnetized surface along
lesser cost per byte than the main memory. They are concentric circles called tracks. Tracks are further
non-volatile in nature, that is, data is not lost when the divided into wedge-shaped sectors.
device is powered off. The most common auxiliary stor- 4. Magnetic tapes: It consists of tape made up of
age devices used in consumer systems are flash memory, plastic covered with magnetic oxide coating. Tapes
optical disks and magnetic disks. are mounted on reels. Bits are recorded as magnetic
1. Flash memory: Flash memory is an electronic spots on tape along several tracks. R/W heads are
non-volatile fastest computer storage device that mounted in each track so that data can be recorded
can be electrically erased and reprogrammed. and read as a sequence of characters. Seven or nine
Example: flash drives and solid state drive. bits are recorded to form a character together with
2. Optical disk: Optical disks are low-cost mass a parity bit. Data is recorded in contiguous blocks
storage devices from which read and write opera- separated by inter-record gaps.
tions are performed using laser technology. Optical
disks can store huge amounts of data up to 6 GB 2.8.3 Cache Memory
(6 billion bytes). Different types of optical disks
are CD-ROM (compact disk read-only), WORM It is a special memory that compensates the speed
(write-once read-many), EO (erasable optical mismatch between processor and main memory access
disks) and DVD. time. It temporarily stores frequently used instructions
3. Magnetic disk: A magnetic disk is composed and data for faster processing by the CPU. Cache hit
of a circular platter made of metal or plastic and ratio is calculated to measure its performance. If a data
Chapter 2.indd 73 4/9/2015 9:50:02 AM

item requested by the CPU is found in cache it is called 2.8.4.1 Direct Mapping

hit otherwise it is a miss. Hit ratio is defined as ratio
of number of hits divided by total CPU references to In this technique, each block from the main memory has
memory. only one possible location in the cache memory. In this
example, say a block from main memory maps onto a
Number of hits block (i mod 128) of the cache. If there are 2n words in the
Hit ratio (h) =
Number of hits + Number of misses cache memory and 2m words in the main memory, then
Average access time = Hit ratio × Tc m-bit main memory address is divided into two fields: n
bits for index field to access the cache and (m − n) bits
+ (1 - Hit ratio)(Tc + Tm )
for the tag field. Each word in cache consists of the data
where Tc is cache access time and Tm is the main memory and the associated tag. Whenever a new block is brought
access time. into cache, tag is stored along with data bits. Index field
is further divided into block and word if there are mul-
2.8.3.1 Elements of Cache Design tiple words (say k) in a block. The lower k bits select one
of the k words in a block known as word field. The block
The various elements of cache design are as follows: field is used to distinguish a block from other blocks.
1. Cache size: It should be optimum, small enough Tag (m − n) bits Index (n bits)
to keep average cost per bit close to the main
memory and large enough to keep overall average Tag (m - n) bits Block (n - k) bits Word (k bits)
access time close to the cache memory.
2. Mapping function: It describes the mapping of When CPU generates a memory request, the block field
main memory block to cache block. There are three points to a particular block location in the cache. The
different mapping techniques: fully associative, direct high-order tag field is compared with tag bits associated
mapped and set associative cache organization. with that cache location. If they match, then the desired
3. Replacement algorithm: When a new memory word is in that block of cache. If there is no match, then
block is required in cache, one of the existing blocks the block containing the required word must be loaded
must be replaced by a new block. Example: FIFO to cache first (Fig. 2.13).
(first in, first out), LRU (least recently used). Main memory address Main memory
4. Write policy: Cache memory follows write-
through and write-back updating policies. In 5 7 4 Block 0
write-through policy, cache controller copies data Tag Block Word Block 1
immediately to main memory as data is written in
≈ ≈
cache. The data in main memory is always valid,
but this approach reduces system performance. In Cache memory Block 127
write back, update to memory block is delayed until Tag 0
Block 0 Block 128
the updated cache block is replaced by a new block. (5 bits) Data
≈ ≈ ≈ ≈
2.8.4 Cache Mapping Techniques Block 255
The cache memory can store a reasonable number of Block 127 Tag 3 Data Block 3968
blocks, but this number is always small as compared to
blocks in the main memory to keep average cost per bit
low. The correspondence between memory blocks and
≈ ≈
cache block is specified by the following mapping tech- Block 4095
niques. Consider a cache memory consisting of 2K words
with 128 blocks of 16 words each. Number of bits required Figure 2.13 | Direct mapped cache organization.
to address a cache block is 11 bits. Main memory has 64K
The demerit of direct mapping is that hit ratio drops
words and bits required to address is 16 (Fig. 2.12).
considerably if two or more words having same index and
different tags are accessed consecutively one after the other.
Main memory CPU 2.8.4.2 Fully Associative Mapping

64K × 8 Cache memory
2K × 8 In this technique, a main memory block can be placed
into any cache block location. It is the most flexible cache
Figure 2.12 | Cache mapping example. organization. The main memory address is divided into
Chapter 2.indd 74 4/9/2015 9:50:04 AM

2.8 MEMORY HIERARCHY 75
two fields: word and tag. The associative memory stores contain the desired block. The high-order tag field is
both the address (tag) and data of the main memory. then compared associatively to the tags corresponding
Figure 2.14 shows the mapping of different blocks into to the matched set. If a match occurs, the corresponding
cache. High-order 12 bits of CPU address is placed in the word is read from cache else main memory is referred
argument register of the associative memory and com- and block containing that word is brought into cache for
pared to tag bits of each block of the cache to see if the future reference (Fig. 2.15).
desired block is present. Once the desired block is pres-
ent, 4-bit word is used to extract necessary word from Tag Set Word
the cache. 6 6 4 Main memory
Main memory address Main memory address Block 0
12 4 Block 1
≈
Tag 0
Main memory ≈
Tag Word Cache memory
Block 0 Block 63
Set 0 Tag 0 Data Tag 2 Data
Cache memory Block 1 6 bits Block 62
Tag 0 Date
(12 bits)
Block i
Set 63 Tag 3 Date Tag 61 Date 4033 Tag 63
4095
Block 4095 Figure 2.15 | Set-associative mapped cache organization.
Figure 2.14 | Associative mapped cache organization.

Problem 2.17: Consider a memory hierarchy
It is necessary to compare high-order bits of main system containing a cache, a main memory and a
memory with all tag bits corresponding to each block to virtual memory. Assuming, cache access time of 5
find whether a given block is present in cache, so it is ns, and 80% hit ratio . The access time of the main
the most expensive. memory is 100 ns, and it has a 99.5% hit rate.
The access time of the virtual memory is 10 ms.
2.8.4.3 Set-Associative Mapping Calculate the average access time of the memory
hierarchy.
As fully associative mapping is an expensive solution and
direct mapping does not allow words with same index
Solution: As we know, the hit rate of virtual memory
but different tag to exist in cache, set associative map-
is 100%, the average access time for requests that
reach the main memory as (l00 ns × 0.995) + (10 ns ×
ping is a combination of both. It is an improvement over
direct mapping where contention problem is solved by
0.005) = 50,099.5 ns. Given this, the average access
time for requests that reach the cache is (5 ns × 0.80) +
having several choices for block placement. The figure
(50,099.5 ns × 0.20) = 10,024 ns.
below shows two-way set associative cache because each
block of main memory has two choices for block place-
ment in cache. A block i in the main memory can be
in any block belonging to set i mod S of cache, where
S is the number of sets. The block 0, 64, 128, … and so Problem 2.18: A computer uses RAM chips of 1024 ×
on of main memory can map into any of the two blocks 1 capacity.
in set 0. (a) How many chips are needed to provide a memory
The main memory address is divided into three fields: capacity of 16K bytes?
low-order bits for word field, set field to determine the (b) How many of these lines will be common to all
desired block from all possible sets and high-order bits chips?
for the tag field. Each word in cache consists of data and
the associated tag. Solution:
(a) Chips are needed to provide a memory capacity of
Tag Set Word 16K bytes = 16 × 8 = 128 chips
(b) Using 14 address lines (16K = 214), we have 10
lines specifying the chip address which is common
When the CPU generates a memory request, the set
to all chips.
field points to a particular set of the cache which might
Chapter 2.indd 75 4/9/2015 9:50:04 AM

Problem 2.19: Consider a 2-way set associative cache Problem 2.20: The access time of a cache memory
consisting of 256 blocks of 8 words each, and assume is 200 ns and that of main memory is 2000 ns. It is
that the main memory is addressable by 16-bit address estimated that 70% of the memory requests are for
and it consists of 4K blocks. Calculate the number read and remaining 30% for write. The hit ratio for
of bits in each of the TAG, BLOCK/SET and word read accesses only is 0.9. A write-through procedure
fields for different mapping techniques? is used.
Solution: For direct mapping, word field is of 3 (a) What is the average access time of the system
bits to identify 8 different words in a block (23 = 8). considering only memory read cycles?
As cache memory consists of 256 blocks so (28 = 256) (b) What is the average access time of the system for
8 bits are required to address a block because there is both read and write requests?
one-to-one correspondence of block k in main memory (c) What is the hit ratio taking into consideration the
to block (k mod 256) in cache memory. The remain- write cycles also?
ing 5 (16 − 8 - 3) high-order address bits are tag bits.
Thus, the main memory address for direct mapping is Solution:
divided as follows: (a) Average access time = 0.9 × 200 + 0.1 × 2200 =
180 + 220 = 400 ns
Tag Block Word (b) Average access time = 0.3 × 2000 + 0.7 × 400 =
5 bits 8 bits 3 bits 600 + 280 = 880 ns
(c) Hit ratio = 0.7 × 0.9 = 0.63
For fully associative mapping, number of word
bits are same, that is, 3 bits. Cache memory stores
both tag and data. The high-order tag bits of an
address generated by CPU are compared with tag
bits of each block so number of block bits is zero. All Problem 2.21: A 4-way set-associative cache memory
remaining bits (except word bits) are identified as tag uses blocks of four words. The cache can accommo-
bits. Thus, the main memory address for fully asso- date a total of 1024 words from the main memory.
ciative mapping is divided as follows: The main memory size is 128K × 32.
(a) Formulate all pertinent information required to
Tag Word
construct the cache memory.
13 bits 3 bits (b) What is the size of the cache memory?
For two-way set-associative mapping, cache Solution:

memory is mapped with two blocks per set. The word (a) Main memory is 128K = 217 so 17 bits address is
field has same number of bits, 3 bits. There are 128 generated. For a set size of 4, the number of sets
(256/2) sets so 7 bits are required to uniquely identify in cache is 256/4 = 64, so 6 bits are required. Each
these 128 sets. The remaining 6 (16 - 7 - 3) bits are block consists of 4 words so 2 bits are required for
used as tag bits. Thus, the main memory address for the word field.
set-associative mapping is divided as follows:
Tag Set Word
Tag Set Word
9 bits 6 bits 2 bits
(b) Since cache is 4-way set associative, 4 blocks per set are stored in cache memory.
Tag 1 Data 1 Tag 2 Data 2 Tag 3 Data 3 Tag 4 Data 4

9 bits 32 bits 9 bits 32 bits 9 bits 32 bits 9 bits 32 bits
Set 0
- Tag Data Tag Data Tag Data Tag Data
Set 128
Size of cache memory is 128 × 4(9 + 32) = 5248 bits.
Chapter 2.indd 76 4/9/2015 9:50:04 AM

SOLVED EXAMPLES 77
Problem 2.22: Suppose physical memory is of 2GB (b) Since cache is 2-way set associative, 2 blocks per
and each word is of 16 bits. There is a cache contain- set are stored in cache memory. So, the number
ing 2K words of data, and each cache block contains of sets is 128/2 = 64.
16 words. For each of the direct mapped and 2-way
set associative cache configurations, specify how the 21 bits 6 bits 4 bits
address would be partitioned. Tag Set Word
Solution:
(a) For direct mapping, the word field is of 4 bits
Problem 2.23: Consider a direct mapped cache of size
identify 16 different words in a block (24 = 16).
32 KB with block size 32 bytes. The CPU generates
As the cache memory consists of 2K words which
32-bit addresses. What are number of bits needed for
is equivalent to 2K/16 = 128 blocks so (27 =
addressing block in cache and number of tag bits?
128) 7 bits are required to address a block. The
remaining 20 (31 − 7 − 4) high-order address
Solution:
bits are tag bits. Thus, the main memory address
for direct mapping is divided as follows: Tag Block Word
32 − 10 − 5 bits 10 bits 5 bits
Tag Block Word The number of bits needed for addressing block in
cache and number of tag bits are 10, 17, respectively.
IMPORTANT FORMULAS
••Amdahl’s law ••Hit ratio (h)

−1 Number of hits
 Fi 
Soverall = (1 − ∑ Fi ) + ∑ 
=
Number of hits + Number of misses
 S
••Average access time
••The speed up of pipelined system over non-pipelined
= Hit ratio × Tc + (1 − Hit ratio)(Tc + Tm )
system is given by:
here Tc is cache access time and Tm is main
w
n × Tn memory access time.
S=
(k + n − 1) × Tp ••Direct mapping
••Maximum speed up that a pipelined system can Tag (m - n) bits Block (n - k) bits Word (k bits)
achieve is given by:
••Set-associative mapping
kTp
S= =k Tag Set Word
Tn
SOLVED EXAMPLES
1. The principle of locality is used in 2. Which memory unit has lowest access time?
(a) Interrupt (b) Registers (a) Cache (b) Registers
(c) DMA (d) Cache memory (c) Optical disk (d) Main memory
Solution: It is used in cache memory to help the Solution: Registers are used for processing
program access small amounts of address space at and manipulating data and for holding memory
any instant. addresses that are available to the machine-code
programmer. So, they have lowest access time.
Ans. (d)
Ans. (b)
Chapter 2.indd 77 4/9/2015 9:50:06 AM

3. During DMA transfer, the DMA controller takes (c) a processor interrupt.
over the buses to manage the transfer (d) a clock interrupt.
(a) Directly from CPU to memory Solution: Hardware interrupt is present in hard-
(b) Directly from memory to CPU ware pins.
(c) Directly between the memory and registers Ans. (b)
(d) Directly between the I/O device and memory
9. Priority is provided by for access to memory
Solution: DMA controller manages transfer by various I/O channels and processors.
between I/O device and memory.
(a) a register
Ans. (d)
(b) a counter
4. Booth’s algorithm is used for the arithmetic opera- (c) the processor scheduler
tion of (d) a controller
(a) addition. (b) subtraction. Solution: Controller sets priority for memory
(c) multiplication. (d) division. access by various I/O devices and processes.
Solution: It is a multiplication algorithm that Ans. (d)
multiplies two signed binary numbers in 2’s com- 10. By applying the principle of temporal locality,
plement notation. processes are likely to reference pages that
Ans. (c)
(a) have been referenced recently.
5. The reason for improvement in CPU performance (b) are located at address near recently referenced
during pipelining is pages in memory.
(c) have been preloaded into memory.
(a) reduced memory access time.
(d) have to be reloaded into memory.
(b) increased clock speed.
(c) introduction of parallelism. Solution: Temporal locality refers to reuse of
(d) increase in cache memory. resources referenced within a short time frame.
Solution: Instruction-level parallelism is imple- Ans. (a)
mented within a single processor to allow faster 11. Which of the following is a correct statement
CPU throughput. related to L2 cache memory?
Ans. (c)
(a) T
he level 1 cache is always faster than the level
6. Use of cache memory enhances 2 cache.
(b) The level 2 cache is used to mitigate the dynamic
(a) I/O access time
slowdown every time a level 1 cache miss occurs.
(b) memory access time.
(c) Level 2 cache comes as on board only.
(c) effective memory access time.
(d) In modern day computer, the level 2 cache is
(d) secondary storage access time.
considered an internal cache.
Solution: Cache memory compensates the speed
Solution: L2 level of cache is placed between the
mismatch between processor and main memory
L1 and RAM. The L1 cache is always the fastest.
access time.
Ans. (c) Ans. (a)
7. An instruction cycle refers to 12. What is the control unit’s function in the CPU?
(a) fetching an instruction. (a) To decode program instructions
(b) executing an instruction. (b) To transfer data to primary storage
(c) fetching, decoding and executing an instruction. (c) To perform logical operations
(d) reading and executing an instruction. (d) To store arithmetic operations
Solution: It involves fetching, decoding and exe- Solution: Control unit controls several units of
cuting the instruction. CPU and helps decode program instructions.
Ans. (c)
Ans. (a)
8. A hardware interrupt is also called
13. CPU fetches the data and instructions from
(a) an internal interrupt.
(a) ROM (b) control unit
(b) an external interrupt.
(c) RAM (d) coprocessors chip
Chapter 2.indd 78 4/9/2015 9:50:06 AM

SOLVED EXAMPLES 79
Solution: The CPU fetches data and instructions Solution:

from RAM and executes it.
Ans. (c) 0
14. Which of the following affects the processing power 1
of the CPU? 0 2
Set
1 3
(a) Data bus, addressing schemes
(b) Clock speed, addressing schemes
(c) Clock speed, data bus
(d) Clock speed, data bus, addressing schemes 2c — 2 2cm — 2
Set
Solution: CPU processing speed is affected by 2c — 1 2cm — 1
clock cycle, data bus for routing data and address- Set of associate cache Main memory
ing modes.
Ans. (d) Number
Number of setsofinsets in cache
cache
Number of blocksblocks
Number of in cache
in cache 2c 2c
15. The contents of the flag register after execution of the = = = =
following program by 8085 microprocessor will be: NumberNumber of blocks
of blocks in oneinset
one set
2 2
Program Number of sets in cache = c

SUB A So to map block k, main memory maps to the set
MVI B, (01)H (k mod c) of the cache.
DCR B Ans. (b)
HLT
(a) (54)H (b) (00)H (c) (01)H (d) (45)H 18. Which of the following is/are advantage of virtual
memory?
Solution:
(a) Faster memory to memory on an average
SUB A Subtract contents of memory loca-
(b) Processes can be given protected address
tion whose contents are stored in A
spaces
MVI B, (01)H Move B with (01)H (c) Linker can assign addresses independent of where
DCR B Decrement B gives (54)H the program will be loaded in physical memory
(d) Programs larger than the physical memory size
HLT
can be run
Ans. (a)
16. An N-bit carry look-ahead adder, where N is a mul- Solution: Virtual memory allows programs to
tiple of 4, employs ICS 74181 (40-bit ALU) and have a larger space than the physical memory size
74182 (4-bit carry look-ahead generator). to run. So, the programmer does not have to worry
about physical memory size.
The minimum addition time using the best archi-
Ans. (d)
tecture for adder is
19. The number of full and half adders required to add
(a) Proportional to N
16-bit numbers is
(b) Proportional to log N
(c) A constant (a) 8 half adders, 8 full adders
(d) None of the above (b) 1 half adder, 15 full adders
Solution: The addition time for N-bit carry look- (c) 16 half adders, 0 full adder
ahead adder is always a constant. (d) 4 half adders, 12 full adders
Ans. (c) Solution: To add N-bit numbers:
Number of half adders required = 1
Number of full adders required = N − 1
17. The main memory of a computer has 2 cm blocks
while the cache has 2c blocks. If the cache uses the
Therefore, to add 16-bit number only, 1 half adder
set-associative mapping scheme with two blocks per
and 15 full adders are required.
set, then block k of the main memory maps to the set:
Ans. (b)
(a) (k mod m) of the cache
20. Which of the following requires a device driver
(b) (k mod c) of the cache
(c) (k mode 2c) of the cache (a) Register (b) Cache
(d) (k mod 2cm) of the cache (c) Main memory (d) Disk
Chapter 2.indd 79 4/9/2015 9:50:07 AM

Solution: Disk is the I/O device attached exter- 23. Advantage of synchronous sequential circuits over
nally to the processor. Therefore, disk requires a asynchronous ones is
device driver.
(a) faster operation.
Ans. (d)
(b) ease of avoiding problems due to hazards.
21. More than one word are put in one cache block to (c) lower hardware requirement.
(d) better noise immunity.
(a) exploit the temporal locality of reference in a
program. Solution: Because of less delay, synchronous
(b) exploit the spatial locality of reference in a sequential circuits have faster operation than asyn-
program. chronous ones.
(c) reduce the miss penalty. Ans. (a)
(d) none of the above.
24. The total size of address space in a virtual memory
Solution: There are two types of locality of references system is limited by
temporal and spatial locality.
(a) the length of MAR.
The concept of spatial locality, instead of fetching
(b) the available secondary storage.
just one item from the main memory to the cache,
(c) the available main memory.
is useful to fetch several items that reside at adja-
(d) all of the above.
cent address as well.
So, option (b) is correct. Solution: Virtual memory depends only on the
Ans. (b) available size of the secondary memory.
22. Which of the following statements is false? Ans. (b)
(a) Virtual memory implements the translation of a 25. Comparing the time T1 taken for a single instruc-
program’s address space into physical memory tion on a pipelined CPU with time T2 taken on a
address space. non-pipelined but identical CPU, we can say that
(b) Virtual memory allows each program to exceed (a) T1 ≤ T2
the size of the primary memory. (b) T1 ≥ T2
(c) Virtual memory increases the degree of (c) T1 < T2
multiprogramming. (d) T1 plus T2 is the time taken for one instruction
(d) Virtual memory reduces the context-switching fetch cycle
overhead.
Solution: In case of one instruction, non-
Solution: Virtual memory increases the context- pipelined CPU takes less time as compared to pipe-
switching overhead. lined CPU. This is due to buffer delays for pipelining.
Ans. (d) Ans. (b)
1. For a pipelined CPU with a single ALU, consider Solution: All the three statements cause hazards.
the following situations: Ans. (d)
I. The j + 1-st instruction uses the result of the
Common Data Questions 2 and 3: Consider the
jth instruction as an operand.
following assembly language program for a hypo-
II. The execution of a conditional jump instruction.
thetical processor. A, B and C are 8-bit registers.
III. The j-th and j + 1-st instructions require the
The meanings of various instructions are shown as
ALU at the same time.
comments:
Which of the above can cause a hazard?
MOV B, #0; B ← 0
(a) I and II only (b) II and III only MOV C, #8; C ← 8
(c) III only (d) All the three Z: CMP C, #0; Compare C with 0
(GATE 2003: 1 Mark) JZ X; Jump to X if zero flag is set
Chapter 2.indd 80 4/9/2015 9:50:07 AM

SUB C, #1; C ← C − 1

RRC A, #1; Right rotate A through carry Instruction Operation Instruction
by one bit. Thus: Size
; if the initial values of A and (in words)
the carry flag are a7..a0 and; c0
respectively, their values after the
MOV R1, 1 ← Memory
;R 2
5000 [5000]
execution of this; instruction will
be c0 a7..a1 and a0, respectively.
JC Y; Jump to Y if carry flag is set
MOV R2 2 ← Memory
;R 1
(R1) [(R1)]
JMP Z; Jump to Z
Y: ADD B, #1; B ← B + 1
JMP Z; Jump to Z
ADD R2, R3 2 ← R2 + R3
;R 1
X:
MOV 6000, ; Memory 2
2. If the initial value of register A is A0, the value of R2 [6000] ← R2
register B after the program execution will be
HALT ; Machine halts 1
(a) The number of 0 bits in A0
(b) The number of 1 bits in A0
(c) A0 5. Consider that the memory is byte addressable with
(d) 8 size 32 bits, and the program has been loaded
(GATE 2003: 2 Marks) starting from memory location 1000 (decimal).
Solution: After execution, register B will contain If an interrupt occurs while the CPU has been
number of 0 bits in A0. halted after executing the HALT instruction, the
Ans. (a) return address (in decimal) saved in the stack
will be
3. Which of the following instructions when inserted
at location X will ensure that the value of register (a) 1007 (b) 1020
A after program execution is the same as its initial (c) 1024 (d) 1028
value? (GATE 2004: 2 Marks)
(a) RRC A, #1 Solution:
(b) NOP; no operation
(c) LRC A, #1; left rotate A through carry flag by MOV R1, 5000 ;2 1000 to 1007
one bit
(d) ADD A, #1 MOV R2 (R1) ;1 1008 to 1011
ADD R2, R3 ;1 1012 to 1015
Solution: RRC A, #1
Ans. (a) MOV 6000, R2 ;2 1016 to 1023
4. Which of the following addressing modes are suit-
HALT ;1 1024 to 1027
able for program relocation at run time?
(i) Absolute addressing (ii) Base addressing Ans. (c)
(iii) Relative addressing (iv) Indirect addressing 6. Let the clock cycles required from various opera-
(a) (i) and (iv) (b) (i) and (ii) tions be as follows:
(c) (ii) and (iii) (d) (i), (ii) and (iv) Register to/from memory transfer: 3 clock cycles ADD
(GATE 2004: 1 Mark) with both operands in register: 1 clock cycle Instruction
fetches and decodes: 2 clock cycles per word
Solution: Both base addressing and relative
The total number of clock cycles required to exe-
addressing modes are suitable.
cute the program is
Ans. (c)
(a) 29 (b) 24
following program segment for a hypothetical CPU (c) 23 (d) 20
having three-user registers R1, R2 and R3. (GATE 2004: 2 Marks)
Chapter 2.indd 81 4/9/2015 9:50:07 AM

Solution: 10. Consider a multiplexer with X and Y as data inputs

and Z as control input. Z = 0 selects input X, and
Clock Cycles
Z = 1 selects input Y. What are the connections
MOV R1, 5000 ;2 8 required to realize the 2-variable Boolean function
MOV R2 (R1) ;1 5 f = T + R, without using any additional hardware?
ADD R2, R3 ;1 1 (a) R to X, 1 to Y, T to Z
(b) T to X, R to Y, T to Z
MOV 6000, R ;2 8 (c) T to X, R to Y, 0 to Z
HALT ;1 2 (d) R to X, 0 to Y, T to Z
Total 24 clock cycles are required. Solution: Connect R to X, 1 to Y and T to Z.

Ans. (b) Ans. (a)
7. Consider a small two-way set-associative cache 11. The microinstructions stored in the control
memory, consisting of four blocks. For choosing memory of a processor have a width of 26 bits.
the block to be replaced, use the least recently used Each microinstruction is divided into three fields: a
(LRU) scheme. The number of cache misses for the micro-operation field of 13 bits, a next address field
following sequence of block addresses 8, 12, 0, 12, 8 is (X), and a MUX select field (Y). there are 8 status
bits in the inputs of the MUX. How many bits are
(a) 2 (b) 3 (c) 4 (d) 5 there in the X and Y fields, and what is the size of
(GATE 2004: 2 Marks) the control memory in number of words?
Solution: Sequence is: 8, 12, 0, 12, 8 (a) 10, 3, 1024 (b) 8, 5, 256
(c) 5, 8, 2048 (d) 10, 3, 512
12 12 12 Load
8 8 0 8 Control address
register
Total number of miss = 4
Ans. (c) Increment
8. A hard disk with a transfer rate of 10 MB/s is con- Control
stantly transferring data to memory using DMA. memory
The processor runs at 600 MHz, and takes 300 and 13
Mux Y
900 clock cycles to initiate and complete DMA Micro operations
transfer, respectively. If the size of the transfer is 8
20 KB, what is the percentage of processor time Status bits
X
consumed for the transfer operation?
Solution: MUX has 8 input bits, 3 select lines are
(a) 5.0% (b) 1.0% (c) 0.5% (d) 0.1%
required. So, Y = 3. Total bits = 26
(GATE 2004: 2 Marks) X = 26 — 13 — 3 = 10. So, memory required = 210 = 1024
Ans: (a)
Solution: Given that data transfer rate = 10 MB/s
and size of transfer = 20 KB 12. Which one of the following is true for a CPU
So, % processor time consumed = 20 KB × 100/10 having a single interrupt request line and a single
MB/s = 0.1% interrupt grant line?
Ans. (d) (a) N either vectored interrupt nor multiple inter-
9. A 4-stage pipeline has the stage delays as 150, 120, rupting devices are possible.
160 and 140 ns, respectively. Registers that are (b) Vectored interrupts are not possible but mul-
used between the stages have a delay of 5 ns each. tiple interrupting devices are possible.
Assuming constant clocking rate, the total time taken (c) Vectored interrupts and multiple interrupting
to process 1000 data items on this pipeline will be devices are both possible.
(a) 120.4 µs (b) 160.5 µs (c) 165.5 µs (d) 590.0 µs
(d) Vectored interrupt is possible but multiple
interrupting devices are not possible.
(GATE 2005: 1 Mark)
Solution: Maximum stage delay = 160 + 5 = 165 ns Solution: Vectored interrupts are not possible but
T = [4 + (1000 -1)] 165 = 165.5 µs multiple interrupting devices are possible.
Ans. (c) Ans. (b)
Chapter 2.indd 82 4/9/2015 9:50:08 AM

13. Normally user programs are prevented from handling

(1) A[I] = B[J] (a) Indirect addressing
I/O directly by I/O instructions in them. For CPUs
having explicit I/O instructions, such I/O protection (2) while (*A++) (b) Indexed addressing
is ensured by having the I/O instructions privileged. (3) int temp =*x (c) Auto increment
In a CPU with memory mapped I/O, there is no
explicit I/O instruction. Which one of the following (a) (1, c), (2, b), (3, a) (b) (1, a), (2, c), (3, b)
is true for a CPU with memory mapped I/O? (c) (1, b), (2, c), (3, a) (d) (1, a), (2, b), (3, c)
(a) I/O protection is ensured by operating system (GATE 2005: 2 Marks)
routine(s). Solution: A[I] = indexed addressing
(b) I/O protection is ensured by a hardware trap. *A++ = auto increment
(c) I/O protection is ensured during system Temp = *x = indirect addressing
configuration. Ans. (c)
(d) I/O protection is not possible.
16. Consider a direct mapped cache of size 32 KB with
(GATE 2005: 1 Mark) block size 32 bytes. The CPU generates 32-bit
Solution: I/O protection is ensured by operating addresses. The number of bits needed for cache
system routine(s). indexing and the number of tag bits are, respectively
Ans. (a) (a) 10, 17 (b) 10, 22 (c) 15, 17 (d) 5, 17
14. Consider a three word machine instruction ADD (GATE 2005: 2 Marks)
A[R0], @B
Solution: Cache size = 32 KB = 15 bits are
The first operand (destination) “A [R0]” uses required
indexed addressing mode with R0 as the index regis-
ter. The second operand (source) “@B” uses indirect Index bits = 15
addressing mode. A and B are memory addresses Tag bits 32 − 15 = 17
residing at the second and the third words, respec- Ans. (a)
tively. The first word of the instruction specifies 17. A 5-stage pipelined CPU has the following sequence
the opcode, the index register designation and the of stages:
source and destination addressing modes.
IF: Instruction fetch from instruction memory
During execution of ADD instruction, the two operands
RD: Instruction decodes and register read
are added and stored in the destination (first operand).
EX: Execute ALU operation for data and address
The number of memory cycles needed during the computation
execution cycle of the instruction is:
MA: Data memory access for write access, the
(a) 3 (b) 4 register read at RD state is used.
(c) 5 (d) 6 WB: Register write back.
(GATE 2005: 2 Marks) Consider the following sequence of instructions:
Solution: A[R0] = require 1 memory cycle I1: L R0, loc 1; R0 ⇐ M[loc1]
The 3rd and 4th operands are indirectly addressed I2: A R0, R0 1; R0 ⇐ R0 + R0
I3: S R2, R0 1; R2 ⇐ R2 − R0
so each requires 2 cycles = 4 memory cycles.
1 memory cycle will be required to store the result.
So, total of 6 memory cycles are required. Let each stage take one clock cycle. What is the
Ans. (d) number of clock cycles taken to complete the above
sequence of instructions starting from the fetch of I1?
15. Match each of the high level language statements
given on the left-hand side with the most natural (a) 8 (b) 10 (c) 12 (d) 15
addressing mode from those listed on the right- (GATE 2005: 2 Marks)
hand side. Solution: 12 clock cycles are taken, as shown below
Clocks 1 2 3 4 5 6 7 8 9 10 11 12
I1 IF RD EX MA WB
I2 IF - - - RD EX MA WB
I3 IF - - - RD - - EX MA WB
Ans. (c)
Chapter 2.indd 83 4/9/2015 9:50:08 AM

18. A device with data transfer rate 10 KB/s is con- fetch cycle of the first word of the instruction, its
nected to a CPU. Data is transferred byte-wise. register transfer interpretation is
Rn ⇐ PC+1;
Let the interrupt overhead be 4 s. The byte trans-
PC ⇐ M[PC];
fer time between the device interfaces register and
CPU or memory is negligible. What is the mini-
mum performance gain of operating the device The minimum number of CPU clock cycles needed
under interrupt mode over operating it under pro- during the execution cycle of this instruction is
gram-controlled mode? (a) 2 (b) 3 (c) 4 (d) 5
(a) 15 (b) 25 (GATE 2005: 2 Marks)
(c) 35 (d) 45 Solution: The minimum number of CPU clock
(GATE 2005: 2 Marks) cycles needed during the execution cycle = 4. This
is because
Solution: Data transfer rate = 10 KB/s 1 cycle is required to transfer already incremented
Interrupt overhead = 4 × 10−2 s value of PC
10 KB is sent = 1 s 2 cycles for getting data in MDR
1 B is sent = 1/10K = 100 − 10−2 s 1 to load value of MDR in PC
Minimum performance gain = 100 × 10−2/4 × 10−2 Ans. (c)
= 25
21. Consider a disk drive with the following
Ans. (b)
specifications:
Common Data Questions 19 and 20: Consider the 16 surfaces, 512 tracks/surface, 512 sectors/track, 1
following data path of a CPU. The ALU, the bus KB/sector, rotation speed 3000 rpm. The disk is oper-
and all the registers in the data path are of identi- ated in cycle stealing mode whereby whenever one 4
cal size. All operations including incrementation of byte word is ready it is sent to memory; similarly, for
the PC and the GPRs are to be carried out in the writing, the disk interface reads a 4 byte word from
ALU. Two clock cycles are needed for memory read the memory in each DMA cycle. Memory cycle time
operation - the first one for loading address in the is 40 nsec. The maximum percentage of time that the
MAR and the next one for loading data from the CPU gets blocked during DMA operation is:
memory bus into the MDR.
(a) 10 (b) 25
MAR MDR (c) 40 (d) 50
Solution:
Data transfer in one rotation = 512 × 1024 Bytes
S T 60
1 rotation takes = s
IR PC 3000
GPRs ALU 60
512KB is transferred in = s
3000
60
1 byte will be transferred = × 512 × 1024
19. The instruction “add R0, R1” has the register 3000
transfer interpretation R0 ⇐ R0 + R1. The mini- 4 bytes will be transferred
4
= 60 × × 512 × 1024 = 152.58 ns
mum number of clock cycles needed for execution
cycle of this instruction is 3000
(a) 2 (b) 3 40
Block % = = 26%
(c) 4 (d) 5 152.28
Ans. (b)
22. A CPU has 24-bit instructions. A program starts at
Solution: There will be three cycles-(1) R1out, address 300 (in decimal). Which one of the following
Sin, (2) R2out, Tin and (3) Sout, Tout, ALUadd, Rin. is a legal program counter (all values in decimal)?
Ans. (b) (a) 400 (b) 500
20. The instruction “call Rn, sub” is a two-word instruc- (c) 600 (d) 700
tion. Assuming that PC is incremented during the (GATE 2006: 1 Mark)
Chapter 2.indd 84 4/9/2015 9:50:09 AM

Solution: Size of instruction = 24 bits; Start and the bits are numbered 0 to 31, bit in position
address = 300. Legal address will be multiple of 0 being the least significant. Consider the following
three, that is, 300. emulation of this instruction on a processor that
Ans. (c) does not have bbs implemented.
23. A CPU has a cache with block size 64 bytes. temp ← reg & mask
The main memory has k banks, each bank being Branch to label if temp is non-zero.
c bytes wide. Consecutive c byte chunks are The variable temp is a temporary register. For correct
mapped on consecutive banks with wrap around. emulation, the variable mask must be generated by
All the k banks can be accessed in parallel, but
two accesses to the same bank must be serialized. (a) mask ← 0 × 1 pos
A cache block access may involve multiple itera- (b) mask ← 0 × ffffffff pos
tions of parallel bank accesses depending on the (c) mask ← pos
amount of data obtained by accessing all the k (d) mask ←0×f
banks in parallel. Each iteration requires decoding (GATE 2006: 2 Marks)
the bank numbers to be accessed in parallel and
this takes k/2 ns. The latency of one bank access is Solution: As there is only one bit with pos, the other
80 ns. If c = 2 and k = 24, the latency of retriev- bits need to be set to 0 in temp. The mask register
ing a cache block starting at address zero from the must have 1 in pos position, for which pos number
main memory is of left shifts over 1 need to be made.
Ans. (a)
(a) 92 ns (b) 104 ns
(c) 172 ns (d) 184 ns Common Data Questions 26 and 27: Consider two
cache organizations. The first one is 32 KB, 2-way
(GATE 2006: 2 Marks) set associative with 32-byte block size. The second
Solution: one is of the same size but direct mapped. The size
Time for one parallel process = k/2 + latency of an address is 32 bits in both cases. A 2-to-1 mul-
Time for one byte = 24/2 + 80 = 92 tiplexer has a latency of 0.6 ns while a k-bit com-
Total time for c bytes = 2 × 92 = 184 ns parator has a latency of k/10 ns. The hit latency of
Ans. (d) the set-associative organization is h1 while that of
the direct mapped one is h2.
24. A CPU has a five-stage pipeline and runs at 1
GHz frequency. Instruction fetch happens in the 26. The value of h1 is
first stage of the pipeline. A conditional branch (a) 2.4 ns (b) 2.3 ns
instruction computes the target address and (c) 1.8 ns (d) 1.7 ns
evaluates the condition in the third stage of the
pipeline. The processor stops fetching new instruc- (GATE 2006: 2 Marks)
tions following a conditional branch until the Solution:
branch outcome is known. A program executes Address bits = 32
109 instructions out of which 20% are conditional Block size = 32 B = 5 bits
branches. If each instruction takes one cycle to Size of cache = 32KB/32 = 1KB = 10 bits
complete on average, the total execution time of For 2-way set-associative memory = index bits = 9,
the program is tag bits = 18
(a) 1.0 s (b) 1.2 s k/10 = 18/10 = 1.8 + latency = 1.8 + 0.6 = 2.4
(c) 1.4 s (d) 1.6 s Ans. (a)
(GATE 2006: 2 Marks) 27. The value of h2 is
Solution: (a) 2.4 ns (b) 2.3 ns
Total execution time of the program (c) 1.8 ns (d) 1.7 ns
= 109 + 0.20 × 2 × 109 = 1.4 s (GATE 2006: 2 Marks)
Ans. (c)
Solution:
25. Consider a new instruction named branch-on-bit- In direct memory access: Tag bits = 17; Index = 10
set (mnemonic bbs). The instruction “bbs reg, pos, bits; word = 5 bits
label” jumps to label if bit in position pos of reg- k/10 = 17/10 = 1.7 + latency = 1.7 + 0.6 = 2.3
ister operand reg is one. A register is 32 bits wide Ans. (b)
Chapter 2.indd 85 4/9/2015 9:50:10 AM

Common Data Questions 28 and 29: A CPU has Solution: For 64 words, log264 = 6 bits are required.
a 32-KB direct-mapped cache with 128-byte block For lines = 128/4 = 32 lines, 5 bits are required
size. Suppose A is a two-dimensional array of size
512 × 512 with elements that occupy 8 bytes each. Tag bits Line Word
Consider the following two C code segments, P1 9 5 6
and P2. Ans. (d)
P1: for (i=0; i<512; i++) { 31. Consider a pipelined processor with the following
for (j=0; j<512; j++) { four stages:
x +=A[i][j];
} IF: Instruction fetch
} ID: Instruction decode and operand fetch
P2: for (i=0; i<512; i++) { EX: Execute
for (j=0; j<512; j++) { WB: Write back
x +=A[j][i]; The IF, ID and WB stages take one clock cycle each
} to complete the operation. The number of clock cycles
}
for the EX stage depends on the instruction. The
ADD and SUB instructions need 1 clock cycle and
P1 and P2 are executed independently with the
the MUL instruction needs 3 clock cycles in the EX
same initial state, namely, the array A is not in the
stage. Operand forwarding is used in the pipelined
cache and i, j, x are in registers. Let the number
processor. What is the number of clock cycles taken
of cache misses experienced by P1 be M1 and that
to complete the following sequence of instructions?
R2 ← R1 + R0
for P2 be M2.
ADD R2, R1, R0
28. The value of M1 is MUL R4, R3, R2 R4 ← R3 × R2
(a) 0 (b) 2048 SUB R6, R5, R4 R 6 ← R 5 − R4
(c) 16384 (d) 262144 (a) 7 (b) 8 (c) 10 (d) 14
Solution: Solution:
Memory = 32 KB Clock 1 2 3 4 5 6 7 8
Block size = 128 B
I1 IF ID EX WB
Number of blocks = 256
Number of elements in block = 256/8 =16 I2 IF ID EX EX EX WB
P1 cache misses = M1 : 512 × 512/16 = 16384 I3 IF ID - - EX WB
Ans. (c)
Using operand forwarding, 8 clock cycles are required.
29. The value of the ratio M1/M2 is
Ans. (b)
(a) 0 (b) 1/16
32. In a simplified computer, the instructions are
(c) 1/8 (d) 16
OP RJ, Ri - Performs RJ OP Ri and stores the
result in register Ri.
Solution: OP m, Ri - Performs val OP Ri and stores the
P2 number of cache misses = M2 = 512 × 512 result in Ri. val denotes the content of memory
Ratio of M1:M2 = 1:16 location m.
Ans. (b) MOV, mRi - Moves the content of memory loca-
tion m to register Ri.
30. Consider a 4-way set-associative cache consist- MOV Ri, m - Moves the content of register Ri to
ing of 128 lines with a line size of 64 words. The memory location m.
CPU generates a 20-bit address of a word in main
memory. The number of bits in the TAG, LINE The computer has only two registers, and OP is either
and WORD fields are, respectively: ADD or SUB. Consider the following basic block:
t1 = a+b
(a) 9, 6, 5 (b) 7, 7, 6
t2 = c+d
e − t2
(c) 7, 5, 8 (d) 9, 5, 6
t3 =
(GATE 2007: 1 Mark) t4 = t1 − t3
Chapter 2.indd 86 4/9/2015 9:50:11 AM

Assume that all operands are initially in memory. The Solution: Given that R1 = 10, so the loop will
final value of the computation should be in memory. run 10 times
What is the minimum number of MOV instructions 10 × 2 + 1 = 21
in the code generated for this basic block? Ans. (d)
(a) 2 (b) 3
(c) 5 (d) 6 34. Assume that the memory is word addressable.
After the execution of this program, the content of
(GATE 2007: 1 Mark) memory location 2010 is
Solution: The instructions generated in the code (a) 100 (b) 101
for this basic block are as follows: (c) 102 (d) 110
MOV a, Ri (GATE 2007: 1 Mark)
ADD b, Ri
MOV c, Rj Solution: It will remain 100, because the loop will
ADD d, Rj exit as the value in R1 becomes 0 when address in
SUB e, Rj R3 becomes 2010.
SUB Ri, Rj Ans. (a)
MOV m, Ri 35. Assume that the memory is byte addressable and
Ans. (b) the word size is 32 bits. If an interrupt occurs
Common Data Questions 33-35: Consider the during the execution of the instruction “INC R3”,
following program segment. Here R1, R2 and R3 what return address will be pushed on to the stack?
are general purpose registers. (a) 1005 (b) 1020
(c) 1024 (d) 1040
Instruction Operation Instruction
Size (No. (GATE 2007: 1 Mark)
of Words) Solution: Memory is byte addressable, take 4 bytes
per word. So at INC R3, stack will contain 1024.
MOV R1, R1 ← m[3000] 2
(3000) Ans. (c)
LOOP: MOV R2 ← M[R3] 1 36. Consider a disk pack with 16 surfaces, 128 tracks
R2, (R3) per surface and 256 sectors per track. 512 bytes of
R2 ← R1 + R2
ADD R2, R1 1 data are stored in a bit serial manner in a sector.
The capacity of the disk pack and the number of
MOV (R3), R2 M[R3] ← R2 1 bits required to specify a particular sector in the
disk are respectively:
INC R3 R3 ← R3 + 1 1
(a) 256 Mbyte, 19 bits (b) 256 Mbyte, 28 bits
DEC R1 R1 ← R1 - 1 1 (c) 512 Mbyte, 20 bits (d) 64 Gbyte, 28 bits
BNZ LOOP Branch on not 2 (GATE 2007: 1 Mark)
zero
Solution:
Disk capacity = 16 surfaces × 128 tracks ×
HALT Stop 1
256 sectors × 512 bytes = 256 MB
Assume that the content of memory location 3000 Total number of sectors = 16 × 128 × 256 = 219
is 10 and the content of the register R3 is 2000. The Ans. (a)
content of each of the memory locations from 2000 to
2010 is 100. The program is loaded from the memory Linked Answer Questions 37 and 38: Consider a
location 1000. All the numbers are in decimal. machine with a byte addressable main memory of
162 bytes. Assume that a direct mapped data cache
33. Assume that the memory is word addressable. The consisting of 32 lines of 64 bytes each is used in the
number of memory references for accessing the system. A 50 × 50 two-dimensional array of bytes is
data in executing the program completely is stored in the main memory starting from memory
location 1100H. Assume that the data cache is ini-
(a) 10 (b) 11
tially empty. The complete array is accessed twice.
(c) 20 (d) 21
Assume that the contents of the data cache do not
(GATE 2007: 1 Mark) change in between the two accesses.
Chapter 2.indd 87 4/9/2015 9:50:11 AM

37. How many data cache misses will occur in total? exception occurs, so an exception is not allowed to
execute. Option (d) is the correct option.
(a) 48 (b) 50 (c) 56 (d) 59
Ans. (d)
41. For a magnetic disk with concentric circular tracks,
Solution: the seek latency is not linearly proportional to the
Main memory = 216 B seek distance due to
Block size = 64 B (a) non-uniform distribution of requests
Number of blocks = 32 (b) arm starting and stopping inertia
Number of elements = 50 × 50 = 2500 (c) higher capacity of tracks on the periphery of
Starting from location 1100 means from 68th block the platter
Number of blocks = 2500/64 = 40 blocks (d) use of unfair arm scheduling policies
Initially cache is empty for 32 misses, then 8 are
remaining from total 40 for one access (GATE 2008: 2 Marks)
Array is traversed twice, so data cache misses = Solution: Tracks on magnetic disks are concentric
40 + 8 + 8 = 56 and seek latency from one sector to other which
Ans. (c) may or may not be in different tracks. This seek
38. Which of the following lines of the data cache will distance is not proportional to latency since the
be replaced by new blocks in accessing the array tracks at periphery have higher diameter, and
for the second time? hence higher capacity to store data.
Ans. (b)
(a) line 4 to line 11 (b) line 4 to line 12
(c) line 0 to line 7 (d) line 0 to line 8 42. Which of the following are NOT true in a pipelined
processor?
Solution:
Applying k mod c to find the location: 68 mod I. Bypassing can handle all RAW hazards.
32 = 4 to 11 II. Register renaming can eliminate all register
Ans. (a) carried WAR hazards.
III. C ontrol hazard penalties can be eliminated by
39. Which of the following is/are true for the auto- dynamic branch prediction.
increment addressing mode?
(a) I and II only (b) I and III only
I. It is useful in creating self-relocating code. (c) II and III only (d) I, II and III
II. If it is included in an Instruction Set Architec
ture, then an additional ALU is required for (GATE 2008: 2 Marks)
effective address calculation. Solution: All the statements are true.
III. T he amount of increment depends on the size Ans. (d)
of the data item accessed.
43. For inclusion to hold between two cache levels L1
(a) I only (b) II only and L2 in a multi-level cache hierarchy, which of
(c) III only (d) II and III only the following are necessary?
(GATE 2008: 2 Marks) I. L1 must be a write-through cache
Solution: Only statement (III) is true. II. L2 must be a write-through cache
Ans. (c) III. T
he associativity of L2 must be greater than
40. Which of the following must be true for the RFE that of L1
(Return from Exception) instruction on a general IV. The L2 cache must be at least as large as the
purpose processor? L1 cache
(a) IV only (b) I and IV only
I. It must be a trap instruction. (c) I, II and IV only (d) I, II, III and IV
II. It must be a privileged instruction.
III. A
n exception cannot be allowed to occur during (GATE 2008: 2 Marks)
execution of an RFE instruction. Solution: L1 and L2 cache are placed between
(a) I only (b) II only CPU and they can be both write through cache
(c) I and II only (d) I, II and III only but not necessarily.
Associativity does not matter.
L2 cache must be at least as large as L1 cache,
Solution: RFE (Return from Exception) is a since all the words in L1 are also in L2.
privileged trap instruction that is executed when Ans. (a)
Chapter 2.indd 88 4/9/2015 9:50:11 AM

44. The use of multiple register windows with over- Solution: Total elements can come in one slot 2048.
lap causes a reduction in the number of memory After 2048 elements, same cache index will be on
accesses for [2][0] and [4][0].
Ans. (b)
I. Function locals and parameters
II. Register saves and restores 47. The cache hit ratio for this initialization loop is
III. Instruction fetches
(a) 0% (b) 25% (c) 50% (d) 75%
(a) I only (b) II only (c) III only (d) I, II and III
Solution: Cache hit ratio is found out as follows:
Solution: Multiple register windows with over-
1024 1
lap causes a reduction in the number of memory = = 50%
accesses for register saves and restores. 2048 2
Ans. (b) As we can see in the above, there will be 50% hits.
Ans. (c)
45. Consider a machine with a 2-way set-associative
data cache of size 64 KB and block size 16 bytes. Linked Answer Questions 48 and 49: Delayed
The cache is managed using 32 bit virtual addresses branching can help in the handling of control
and the page size is 4 KB. A program to be run on hazards.
this machine begins as follows:
48. For all delayed conditional branch instructions,
double ARR [1024] [1024]; irrespective of whether the condition evaluates to
int i, j;
true or false:
/* Initialize array ARR to 0.0 */
for(i=0; i<1024; i++) (a) T he instruction following the conditional
for(j=0; j<1024; j++) branch instruction in memory is executed.
ARR [i] [j] =0.0; (b) The first instruction in the fall through path is
executed.
The size of double is 8 bytes. Array ARR is located (c) The first instruction in the taken path is
in memory starting at the beginning of virtual page executed.
0xFF000 and stored in row major order. The cache (d) The branch takes longer to execute than any
is initially empty and no pre-fetching is done. The other instruction.
only data memory references made by the program
are those to array ARR.
The total size of the tags in the cache directory is Solution: The first instruction following the
branch instruction is always executed (irrespective
(a) 32 Kbits (b) 34 Kbits
of whether the branch is taken or not).
(c) 64 Kbits (d) 68 Kbits
Ans. (b)
49. The following code is to run on a pipelined proces-
Solution: sor with one branch delay slot:
Virtual address = 32 bits
2-way cache size = 64 KB I1: ADD R2 ← R7 + R8
1 set will contain = 32 KB entries = 15 bits I2: SUB R4 ← R5 − R6
Block size = 16 bytes = 4 bits I3: ADD R1 ← R2 + R3
Tag bits Set bits Word bits I4: STORE Memory [R4] ← R1
17 11 4 BRANCH to Label if R1 == 0
Tag size = 17 × 2 × 1024 = 34 kbits Which of the instructions I1, I2, I3 or I4 can legiti-
Ans. (b) mately occupy the delay slot without any other
program modification?
46. Consider the data given in the above question.
Which of the following array elements has the (a) I1 (b) I2 (c) I3 (d) I4
same cache index as ARR[0][0]? (GATE 2008: 2 Marks)
(a) ARR[0][4] (b) ARR[4][0] Solution: Instruction I2 contains delayed slot. I4
(c) ARR[0][5] (d) ARR[5][0] has data dependency in I2.
(GATE 2008: 2 Marks) Ans. (b)
Chapter 2.indd 89 4/9/2015 9:50:12 AM

50. How many 32K × 1 RAM chips are needed to pro- 53. Consider a 4-way set-associative cache (initially
vide a memory capacity of 256 K bytes? empty) with total 16 cache blocks. The main
memory consists of 256 blocks and the request for
(a) 8 (b) 32
memory blocks is in the following order:
(c) 64 (d) 128
0, 255, 1, 4, 3, 8, 133, 159, 216, 129, 63, 8, 48, 32,
Solution: 73, 92, 155
256K × 8 Which one of the following memory block will NOT
Number of chips required = = 64
32K × 1 be in cache if LRU replacement policy is used?
Ans. (c) (a) 3 (b) 8 (c) 129 (d) 216
51. A CPU generally handles an interrupt by execut-
Solution:
ing an interrupt service routine
To decide the location (address), mod 4 is applied.
(a) as soon as an interrupt is raised.
(b) by checking the interrupt register at the end of Set 0 0, 4, 8,216 → 48, 32, 8, 92
the fetch cycle. Set 1 1, 133, 129, 73
(c) by checking the interrupt register after finishing Set 2 155
the execution of the current instruction. Set 3 255, 3, 159, 63
(d) by checking the interrupt register at fixed time 216 will not be there in cache.
intervals. Ans. (d)
Common Data Questions 54 and 55: A hard disk has
Solution: Interrupts are handled by checking the 63 sectors per track, 10 platters each with 2 record-
interrupt register after finishing the execution of ing surfaces and 1000 cylinders. The address of a
current instruction. sector is given as a triple <c, h, s>, where c is the
Ans. (c) cylinder number, h is the surface number and s is
the sector number. Thus, the 0th sector is addressed
as <0, 0, 0>, the 1st sector as <0, 0, 1>, and so on.
52. Consider a 4-stage pipeline processor. The number
of cycles needed by the four instructions I1, I2, I3,
I4 in stages S1, S2, S3, S4 is shown below: (GATE 2009: 2 Marks)
54. The address <400, 16, 29> corresponds to sector
S1 S2 S3 S4 number:
I1 2 1 1 1 (a) 505035 (b) 505036
(c) 505037 (d) 505038
I2 1 3 2 2
I3 1 1 1 3 Solution:
Total surfaces = 10 × 2 = 20
Address is: 400 × 20 × 63 + 16 × 63 + 29 = 505037
I4 1 2 2 2
Ans. (c)
What is the number of cycles needed to execute the 55. The address of 1039th sector is
(a) <0, 15, 31> (b) <0, 16, 30>
following loop?
for (i=1 to 2) {I1; I2; I3; I4;} (c) <0, 16, 31> (d) <0, 17, 31>
Solution:
(a) 16 (b) 23 (c) 28 (d) 30 Address of 1039th sector = 16 × 31 + 31 = 1039
Solution:
Clock 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15
I1 S1 S1 S2 S3 S4
I2 S1 S2 S2 S2 S3 S3 S4 S4
I3 S1 S1 S2 S3 S4 S4 S4
I4 S1 S2 S2 S3 S3 S4 S4
For two iterations = 15 × 2 = 30 clock cycles
Ans. (d)
Chapter 2.indd 90 4/9/2015 9:50:12 AM

56. A main memory unit with a capacity of 4 mega- Here we have to find X in terms of Y. So,
bytes is built using 1M × 1-bit DRAM chips. Each  a a a -1 
X = a1 + 2 + 3 + + (nn- =Y
 2 2) 
DRAM chip has 1 K rows of cells with 1K cells
2 4
in each row. The time taken for a single refresh
operation is 100 ns. The time required to perform  
If a0 + 1 + 2 + … + n-1 
a a a
one refresh operation on all the cells in the memory  2 4 2( n -1) 
unit is
 a -1 
< a1 + 2 + 3 + … + (nn-
a a

(a) 100 nanoseconds  2 4 2 2) 
(b) 100 × 210 nanoseconds
(c) 100 × 220 nanoseconds OR
(d) 3200 × 220 nanoseconds  a 1 
X = a0 + 1 + 2 + + (nn -
a a
 = a0 +
Y
(GATE 2010: 1 Mark)  2 4 2 -1) 
 2
Solution:  
If a0 + 1 + 2 + … + (nn −−11) 
a a a
 
Main memory = 4 MB 2 4
Number of DRAM chips = 4 MB/ 1M × 1 bit = 32
2
Total cells = 32 × 1K × 1K  
> a1 + 2 + 3 + … + (nn −−12) 
a a a
Time taken to refresh all the cells = 32 × 1K ×  2 4 2 
1K × 100 ns Hence, we sum up as
Ans. (d) X = MAX(Y, a0 + Y/2)
57. The weight of a sequence a0, a1/2, … an−1 of real Ans. (b)
numbers is defined as a0 + a1/2 + … + an−1/2n−1.
58. A 5-stage pipelined processor has instruction fetch
A subsequence of a sequence is obtained by delet-
(IF), instruction decode (ID), operand fetch (OF),
ing some elements from the sequence, keeping the
perform operation (PO) and write operand (WO)
order of the remaining elements the same. Let X
stages. The IF, ID, OF and WO stages take 1 clock
denote the maximum possible weight of a subse-
quence of a0, a1, …, an−1 and Y the maximum pos-
cycle each for any instruction. The PO stage takes
sible weight of a subsequence of a1, a2, … an−1.
1 clock cycle for ADD and SUB instructions, 3 clock
cycles for MUL instruction and 6 clock cycles for
Then X is equal to
DIV instruction, respectively. Operand forwarding
(a) max(Y, a0 + Y ) (b) max(Y, a0 + Y/2) is used in the pipeline. What is the number of clock
(c) max(Y, a0 + 2Y ) (d) a0 + Y/2 cycles needed to execute the following sequence of
(GATE 2010: 2 Marks) instructions?
Solution: The concepts involve the Dynamic Instruction Meaning of Instruction
Programming in Algorithms.
Given that I0: MUL R2, R0, R1 R2 ← R0 × R1
X = max weight from the sequence (a0, a1, a2, … I1: DIV R5, R3, R4 R5 ← R3/R4
a a a -1
an−1) = a0 + 1 + 2 + + n I2: ADD R2, R5, R2 R2 ← R5 + R2
2 4 2(n -1)
Y= max weight from the sequence (a1, a2, … an−1) I3: SUB R5, R2, R6 R5 ← R2 − R6
a2 a3 a -1
= a1 + + + + (nn -2) (a) 13 (b) 15 (c) 17 (d) 19
2 4 2 (GATE 2010: 2 Marks)
Solution:
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15
I0 IF ID OF PO PO PO WO
I1 IF ID OF - - PO PO PO PO PO PO WO
I2 IF ID OF - - - - - PO WO
I3 IF ID OF PO WO
With operand forwarding, 15 clock cycles are needed.
Ans. (b)
Chapter 2.indd 91 4/9/2015 9:50:14 AM

59. The program below uses six temporary variables a, Solution: Access to L1 cache = 2 ns
b, c, d, e, f. Access to L2 cache = 20 ns
a = 1 Block size of L2 is 16 words; data bus size is 4 words.
b = 10 So, time taken for this transfer = 4 × 22 = 88 ns.
c = 20 Ans. (d)
d = a + b 61. When there is a miss in both L1 cache and L2 cache,
e = c + d
first a block is transferred from main memory to
f = c + e
b = c + e L2 cache, and then a block is transferred from L2
e = b + f cache to L1 cache. What is the total time taken for
d = 5 + e these transfers?
return d + f (a) 222 nanoseconds (b) 888 nanoseconds
(c) 902 nanoseconds (d) 968 nanoseconds
Assuming that all operations take their operands
from registers, what is the minimum number of regis- (GATE 2010: 2 Marks)
ters needed to execute this program without spilling? Solution: Block transfer from main memory to L2
cache = 2 + 4 × (20 + 200) = 2 + 880 = 882 ns
(a) 2 (b) 3
L2 to L1 = 882 + 20 + 66 = 968 ns
(c) 4 (d) 6 Ans. (d)
(GATE 2010: 2 Marks) 62. A computer handles several interrupt sources of
which the following are relevant for this question.
Solution: Let us take three Registers R1, R2 and R3
••Interrupt from CPU temperature sensor (raises
R1 R2 R3 interrupt if CPU temperature is too high)
••Interrupt from Mouse (raises interrupt if the
a=1 b = 10 c = 20
mouse is moved or a button is pressed)
d = 11 b = 10 c = 20 ••Interrupt from Keyboard (raises interrupt when
d = 11 e = 21 c = 20 a key is pressed or released)
••Interrupt from Hard Disk (raises interrupt when
f = 41 e = 21 c = 20
a disk read is completed)
b = 41 e = 21 c = 20
Which one of these will be handled at the HIGHEST
e = 42 b = 41 f = b = 41 priority?
d = 47 e = 42 f = 41 (a) Interrupt from Hard Disk
(b) Interrupt from Mouse
All the operations will be completed using three
(c) Interrupt from Keyboard
registers only.
(d) Interrupt from CPU temperature sensor
Common Data Questions 60 and 61: A computer Solution: Interrupt from CPU temperature sensor
system has an L1 cache, an L2 cache, and a main will be handled at the highest priority.
memory unit connected as shown below. The block Ans. (d)
size in L1 cache is 4 words. The block size in L2 63. Consider a hypothetical processor with an instruction
cache is 16 words. The memory access times are 2 of type LW R1, 20 (R2), which during execution reads
nanoseconds, 20 nanoseconds and 200 nanoseconds a 32-bit word from memory and stores it in a 32-bit
for L1 cache, L2 cache and main memory unit, register R1. The effective address of the memory loca-
respectively. tion is obtained by the addition of a constant 20 and
Data bus Data bus the contents of register R2. Which of the following
L1 L2 Main best reflects the addressing mode implemented by
Cache Cache memory this instruction for the operand in memory?
4 words 4 words
(a) Immediate addressing
60. When there is a miss in L1 cache and a hit in L2 (b) Register addressing
cache, a block is transferred from L2 cache to (c) Register indirect scaled addressing
L1 cache. What is the time taken for this transfer? (d) Base indexed addressing
(GATE 2011: 1 Mark)
(a) 2 nanoseconds (b) 20 nanoseconds
(c) 22 nanoseconds (d) 88 nanoseconds Solution: Effective address = contents of register
R2 + 20.
(GATE 2010: 2 Marks) Ans. (d)
Chapter 2.indd 92 4/9/2015 9:50:15 AM

64. On a non-pipelined sequential processor, a program Memory size for tag bits = 19 + 2 = 21
segment, which is a part of the interrupt service Total size of memory for tags = 21 × 256 = 5376 bits
routine, is given to transfer 500 bytes from an I/O Ans. (d)
device to memory.
66. Consider an instruction pipeline with four stages
Initialize the address register (S1, S2, S3 and S4) each with combinational circuit
Initialize the count to 500 only. The pipeline registers are given in the figure.
LOOP: Load a byte from device
Pipeline register (delay 1 ns)

Store in memory at address given by address register
Increment the address register
Decrement the count Stage Stage Stage Stage
If count!= 0 go to LOOP S1 S2 S3 S4
Assume that each statement in this program is delay delay delay delay
equivalent to a machine instruction which takes 5 ns 6 ns 11 ns 8 ns
one clock cycle to execute if it is a non-load/store
instruction. The load-store instructions take two
clock cycles to execute.
The designer of the system also has an alternate
approach of using the DMA controller to implement What is the approximate speed up of the pipeline in
the same transfer. The DMA controller requires 20 steady state under ideal conditions when compared
clock cycles for initialization and other overheads. Each to the corresponding non-pipeline implementation?
DMA transfer cycle takes two clock cycles to transfer (a) 4.0 (b) 2.5 (c) 1.1 (d) 3.0
one byte of data from the device to the memory.
What is the approximate speed up when the DMA Solution:
controller based design is used in place of the inter- Time taken by non-pipelined
rupt-driven program-based input output? Speed up =
Time taken by pipelined
(a) 3.4 (b) 4.4 5 + 6 + 11 + 8
(c) 5.1 (d) 6.7 = = 2.5
11 + 1
Solution: By using load store instructions: 1 + 1 67. The decimal value 0.5 in IEEE single-precision
+ 500 × (2 + 2 + 1 + 1 + 1) = 2 + 500 × 7 = 3502 floating-point representation has
By using DMA: 20 + 500 × 2 = 1020
Speed up = 3502/1020 = 3.4 (a) Fraction bits of 000…000 and exponent value of 0
Ans. (a) (b) Fraction bits of 000…000 and exponent value
of −1
65. An 8 KB direct-mapped write-back cache is orga- (c) Fraction bits of 100…000 and exponent value of 0
nized as multiple blocks, each of size 32 bytes. The (d) No exact representation
processor generates 32-bit addresses. The cache (GATE 2012: 1 Mark)
controller maintains the tag information for each
cache block comprising of the following: Solution: Binary representation of 0.5 is 0.1000.
which can be written as = 1.000 × 2−1
Here, exponent = −1 and fraction bits are 000…000.
1 Valid bit
1 Modified bit
Ans. (b)
As many bits as the minimum needed to identify
the memory block mapped in the cache. 68. Register renaming is done in pipelined processors
What is the total size of the memory needed at the (a) A
s an alternative to register allocation at com-
cache controller to store metadata (tags) for the cache? pile time
(a) 4864 bits (b) 6144 bits (b) For efficient access to function parameters and
(c) 6656 bits (d) 5376 bits local variables
(GATE 2011: 2 Marks) (c) To handle certain kinds of hazards
(d) As part of address translation
Solution: Number of blocks = 8K/32 = 256 blocks (GATE 2012: 1 Mark)
Tag bits Block bits Word bits Solution: Register renaming is done to handle
WAR/WAW hazards.
19 8 5
Ans. (c)
Chapter 2.indd 93 4/9/2015 9:50:15 AM

69. The amount of ROM needed to implement a 4-bit 73. Consider the following sequence of micro-operations.
multiplier is MBR ← PC
(a) 64 bits (b) 128 bits (c) 1 Kbits (d) 2 Kbits MAR ← X
PC ← Y
Solution: Amount of ROM required Memory ← MBR
= 22k × 2k (where k = number of bits) Which one of the following is a possible operation
= 22×4 × 2 × 4 performed by this sequence?
= 2 Kbits (GATE 2013: 2 Marks)
Ans. (d)
(a) Instruction fetch
70. A computer has a 256 KB, 4-way set-associative, (b) Operand fetch
write-back data cache with block size of 32 bytes. (c) Conditional branch
The processor sends 32-bit addresses to the cache (d) Initiation of interrupt service
controller. Each cache tag directory entry contains,
in addition to address tag, 2 valid bits, 1 modified Solution: Program counter value is stored in
bit and 1 replacement bit. memory by MBR and gets a new address by Y. This
indicates initialization of interrupt service routine.
The number of bits in the tag field of an address is Ans. (d)
(a) 11 (b) 14 (c) 16 (d) 27
74. Consider a hard disk with 16 recording surfaces
(0-15) having 16384 cylinders (0-16383) and each
Solution: cylinder contains 64 sectors (0-63). Data storage
256 KB capacity in each sector is 512 bytes. Data are orga-
Number of blocks = = 213 blocks nized cylinder-wise and the addressing format is
32 B
<cylinder no., surface no., sector no.>. A file of size
Due to 4-way set associative = 213/22 = 211 42797 KB is stored in the disk and the starting disk
32 bits location of the file is <1200, 9, 40>. What is the
cylinder number of the last sector of the file, if it is
Ans. (c) stored in a contiguous manner?
71. The size of the cache tag directory is (GATE 2013: 2 Marks)
(a) 160 Kbits (b) 136 bits (a) 1281 (b) 1282 (c) 1283 (d) 1284
(c) 40 Kbits (d) 32 bits Solution: Number of sectors required to store the
(GATE 2012: 2 Marks) 42797 × 1024
file = = 85594 sectors
512
Solution: Tag directory contains: Tag bits + 4 Number of sectors in a cylinder = 16 × 64 = 1024
additional bits = 20 bits
85594
Size of cache tag directory = 20 × 213 = 160 Kbits Total number of cylinders required = = 84
Ans. (a) 1024
Last sector will be stored on 1284th cylinder.
72. In a k-way set associative cache, the cache is divided Ans. (d)
into v sets, each of which consists of k lines. The lines
of a set are placed in sequence one after another. 75. Consider an instruction pipeline with five stages
The lines in set s are sequenced before the lines in without any branch prediction: Fetch Instruction
set (s+1). The main memory blocks are numbered (FI), Decode Instruction (DI), Fetch Operand
0 onwards. The main memory block numbered j (FO), Execute Instruction (EI) and Write Operand
must be mapped to any one of the cache lines from (WO). The stage delays for FI, DI, FO, EI and WO
are 5 ns, 7 ns, 10 ns, 8 ns and 6 ns, respectively.
(a) (j mod v) * k to (j mod v) * k + (k-1) There are intermediate storage buffers after each
(b) (j mod v) to (j mod v) + (k-1) stage and the delay of each buffer is 1 ns. A pro-
(c) (j mod k) to (j mod k) + (v-1) gram consisting of 12 instructions I1, I2, I3, … I12 is
(d) (j mod k) * v to (j mod k) * v + (v-1) executed in this pipelined processor. Instruction I4
(GATE 2013: 1 Mark) is the only branch instruction and its branch target
is I9. If the branch is taken during the execution of
Solution: Number of sets = v
this program, the time (in ns) needed to complete
Number of main memory blocks = j
the program is
Number of lines = k (from 0 to k-1)
Position will be (j mod v)*k to (j mod v)*k +k-1 (a) 132 (b) 165 (c) 176 (d) 328
Ans. (a) (GATE 2013: 2 Marks)
Chapter 2.indd 94 4/9/2015 9:50:16 AM

Solution: Clock pulse duration if (R2 > R1)

= Maximum stage delay + Buffer delay {
= 11 × 5 + (4 − 1) × 11 = 88 ns R2 ← R1 * R1 [y = a*a]
n + k − 1 = 88 + 88 − 11 = 165 }
Ans. (b) else
{
Common Data Questions 76 and 77: The follow- R2 ← c [read value c from memory]
ing code segment is executed on a processor which R2 ← R2 * R1 [d = c * a;]
allows only register operands in its instructions. R2 ← R2 * R2 [d = d * d;]
Each instruction can have atmost two source oper- R2 ← R2 * R1 [e = c + a;]
ands and one destination operand. Assume that all R2 ← R2 * R2 [e = e * e;]
variables are dead after this code segment. }
c = a + b; Only `c’ is spilled variable here, which needs to be
d = c * a; stored and loaded from memory.
e = c + a; Ans. (b)
x = c * c;
77. What is the minimum number of registers needed
if (x > a) {
y = a * a; in the instruction set architecture of the proces-
} sor to compile this code segment without any spill
else { to memory? Do not apply any optimization other
d = d * d; than optimizing register allocation.
e = e * e;
} (a) 3 (b) 4 (c) 5 (d) 6
76. Suppose the instruction set architecture of the Solution: Let R1 = a, R2 = b. The machine code is
processor has only two registers. The only allowed c = a + b; R2 ← R1 + R2
compiler optimization is code motion, which moves
d = c * a; R3 ← R2 * R1
statements from one place to another while pre-
serving correctness. What is the minimum number e = c + a; R4 ← R2 + R1
of spills to memory in the compiled code? x = c * c; R2 ← R2 * R2
if (x > a) { R2 > R1
(a) 0 (b) 1
y = a * a; R1 ← R1 * R1
(c) 2 (d) 3
(GATE 2013: 2 Marks) }
else {
Solution: Suppose R1 and R2 hold a and b, respec- d = d * d; R3 ← R3 * R3
tively. Here for code optimization, code motion is e = e * e; R4 ← R4 * R4
allowed. The machine code is; }
R2 ← R1 + R2 (c = a + b)
R2 ← R2 * R2 (x = c * c) Four registers are required.
Ans. (b)
PRACTICE EXERCISES
Set 1 (c) The width of the address bus

(d) The total number of general purpose CPU
1. The three different kinds of buses supporting the registers
architecture of a computer are
3. Which of the following statement is false about
(a) Address bus, data bus, I/O bus CISC ARCHITECTURE?
(b) Front bus, data bus, address bus
(c) Control bus, data bus, address bus (a) C
ISC machine instructions may include com-
(d) I/O bus, address bus, instruction bus plex addressing modes, which require many
clock cycles to carry out.
2. The term `WORD’ of a CPU equals to (b) CISC control units are typically micro-pro-
(a) The maximum addressable memory size grammed, allowing the instruction set to be
(b) The width of a CPU register (integer or float point) more flexible.
Chapter 2.indd 95 4/9/2015 9:50:16 AM

(c) In the CISC instruction set, all arithmetic/logic 11. The most relevant addressing mode to write posi-
instructions must be register based for fast tion-independent code is
processing. (a) direct mode (b) auto mode
(d) CISC architectures may perform better in net- (c) relative mode (d) indexed mode
work centric applications than RISC.
12. A CPU uses 24-bit instruction. A program starts at
4. The register that holds the address of the loca- address 300 (in decimal). Which one of the follow-
tion to or from which data are to be transferred is ing is a legal program counter content (all values
called in decimal)?
(a) Index register (a) 324 (b) 512
(b) Accumulator (c) 600 (d) 700
(c) Memory address registers
(d) Memory data registers 13. An attempt to access a location not owned by a
program is called
5. Which one of the following is not a type of I/O
(a) data fault (b) address fault
channel?
(c) instruction fault (d) page fault
(a) Multiplexer (b) Selector
14. Which of the following statement about relative
(c) Block multiplexer (d) None of the above
addressing mode is FALSE?
6. The performance of a pipelined processor is (a) It enables reduced program code
degraded if (b) It allows indexing of array element with same
(a) the pipeline stages have different delays instruction
(b) consecutive instructions are to be executed serially (c) It enables easy relocation of data
(c) the pipeline stages share hardware resources (d) It enables faster address calculation than abso-
(d) all of the above lute addressing
7. The minimum time delay between the initiation of 15. Compared to CISC processors, RISC processors contain
two independent memory operations is called (a) more register and smaller instruction set
(a) Access time (b) Cycle time (b) larger instruction set and less registers
(c) Rotational time (d) Latency time (c) less registers and smaller instruction set
(d) more registers and larger instruction set
8. The register which keeps track of the execution of a
16. Micro-programmed control cannot be implemented
program and which contains the memory address of
in RISC architecture because
the instruction currently being executed is known
as (a) it tends to slow down the processor.
(b) it consumes more chip areas and large instruc-
(a) index register
tion set.
(b) memory address register
(c) handling a large number of registers is impos-
(c) program counter
sible in micro-programmed system.
(d) instruction registers
(d) the 1 instruction/cycle timing requirement
9. For interval arithmetic, the best rounding tech- for RISC is difficult to achieve in micro-pro-
nique used is grammed based architecture.
(a) rounding to plus and minus infinity 17. Relocation of the code is easier in irrespec-
(b) rounding to zero tive of the program code
(c) rounding to nearest zero (a) indirect addressing
(d) rounding to the next number (b) indexed addressing
10. Hardwired control unit are faster than micro- (c) base register addressing
programmed control unit because (d) absolute addressing
(a) they do not consist of slower memory elements. 18. In inverted page table organization, the size of the
(b) they do not have slower elements such as gates, page table depends on
flip flops and registers. (a) the number of processes
(c) they consist of elements based on VVLSI design (b) the size of page
technology. (c) the size of main memory
(d) they contain high-speed digital components. (d) the number of frames in the main memory
Chapter 2.indd 96 4/9/2015 9:50:17 AM

19. When using the concept of locality of reference, the 25. A device employing INTA line for device inter-
page reference being made by a process rupt puts the CALL instruction on the data bus
(a) will always be to the page used in the previous while
page reference (a) INTA is active. (b) HOLD is active.
(b) is likely to be one of the pages used in the past (c) READY is active. (d) READY is active.
few page references
(c) will always to be one of the pages existing in 26. On receiving an interrupt from an I/O device, the CPU
the main memory (a) b ranches off to halt (or wait) for a predeter-
(d) will always lead to page fault mined time
(b) branches off to the interrupt service after com-
20. If the new version of processor is not made com-
pletion of the current instruction
patible to programs written for its older version, it
(c) branches off to the interrupt service routine
could be able to process at a faster speed
immediately
(a) the statement is true. (d) hands over control of address bus and data bus
(b) the statement is false. to the interrupting device
(c) the speed cannot be predicted.
(d) speed has nothing to do with the compatibility. 27. Using large block size in a fixed block size file
system leads to
21. A certain snooping cache can snoop only on address
(a) b etter disk throughput but poorer disk space
line. Which of the following is true?
utilization
(a) This would adversely affect the system if the (b) better disk throughput and better disk space
write-through protocol is used. utilization
(b) It would run well if the write-through protocol (c) it does not matter as the total memory size is
is used. same
(c) Data snooping is mandatory to be implemented (d) poorer disk throughput but better disk space
on data line. utilization
(d) Data snooping may not be required.
28. Which of the following statements are true about
22. When the frequency of the input signal to a CMOS paging?
gate is increased, the average power dissipation (a) It divides memory into units of equal size.
(a) decreases exponentially (b) It permits implementation of virtual memory.
(b) increases (c) It suffers from internal fragmentation.
(c) decreases (d) It suffers from external fragmentation.
(d) increases exponentially
29. The number of entries in an inverted page table
23. The disadvantage of hardwired control units with (a) is equal to the number of processes.
flip flop is (b) is equal to the number of page frames in the
(a) design becomes complex main memory.
(b) it requires more number of flip flops (c) is equal to the size of the page frame.
(c) control circuit speed does not match with flip (d) is equal to the number of page frames in cache
flops memory.
(d) flip—flops can handle the data unit not the
30. In a virtual memory system, the addresses used by
control unit
the programmer belongs to
24. In a vectored interrupt, (a) memory space (b) physical space
(a) the branch address is assigned to a fixed loca- (c) address space (d) main memory space
tion in memory.
31. Power consumption of processors can be vastly
(b) the interrupting source supplies the branch
reduced by making use of based transistors
information to the processor through an inter-
to implement the ICs.
rupt vector.
(c) the branch address is obtained from a register (a) NMOS only
in the processor. (b) TTL Schottky and PMOS
(d) the branch address is obtained from program (c) PMOS only
counter. (d) NMOS and PMOS
Chapter 2.indd 97 4/9/2015 9:50:17 AM

32. Address symbol table is generated by the 41. In a multiprogramming system, which of the fol-
(a) memory management software lowing concepts is used?
(b) assembler (a) Data parallelism (b) Paging
(c) table match of associative memory (c) L1 cache (d) DMA
(d) generated by CPU
42. PAL circuit can be defined as
33. How many 128 × 8 RAM chips are needed to have
(a) fixed OR and programmable AND logic.
a total RAM of 2048 bytes?
(b) programmable OR and programmable AND logic.
(a) 8 (b) 16 (c) 24 (d) 32 (c) fixed AND and programmable OR logic.
(d) fixed OR and fixed AND logic.
34. In 8085 microprocessor, how many I/O devices can
be interfaced in I/O mapped I/O technique? 43. If the clock input applied to a cascaded Mod-6 and
Mod-4 counter is 48 kHz. Then the output of the
(a) Either 256 input devices or 256 output devices
cascaded arrangement shall be of:
(b) 8 I/O devices
(c) 256 input devices and 256 output devices (a) 4.8 kHz (b) 12 kHz
(d) 512 input-output devices (c) 8 kHz (d) 48 kHz
44. If there are four ROM ICs of 8K and two RAM ICs
35. After reset, the CPU starts the execution of instruc-
of 4K words, then the address range of Ist RAM is
tion from memory address
(assume initial addresses correspond to ROMs)
(a) 1111H (b) 8000H
(c) 0000H (d) FFFFH (a) (8000)H to (9FFF)H (b) (5000)H to (7FFF)H
(c) (8000)H to (8FFF)H (d) (5000)H to (9FFF)H
36. In a microprocessor system, suppose TRAP, HOLD
45. The method for updating the main memory as soon
and RESET pin got activated at the same time,
as a word is removed from the cache is called
while the processor was executing some instruc-
tions, the system will (a) Write-through (b) Write-back
(c) Write-save (d) Cache-save
(a) execute the TRAP instruction
(b) execute the HOLD instruction Set 2
(c) execute the RESET instruction
(d) none of these instructions will be executed 1. The most appropriate matching for the following
pairs
37. In 8085 microprocessor, the programmer cannot
access which flag directly? X. Indirect addressing 1. Loops
Y. Immediate addressing 2. Pointers
(a) Sign flag (b) Carry flag Z. Auto-decrement addressing 3. Constants
(c) Auxiliary carry flag (d) Parity flag
(a) X − 3 Y − 2 Z − 1 (b) X - 1 Y - 3 Z - 2
38. Which of the following is a pseudo-instruction for (c) X − 2 Y − 3 Z − 1 (d) X - 3 Y - 1 Z - 2
8085?
2. Which of the following is not a form of memory?
(a) SPHL (b) CMP
(c) NOP (d) END (a) Instruction cache
(b) Instruction register
39. The term “cycle stealing” refers to: (c) Instruction opcode
(d) Translation look aside buffer
(a) Interrupt-based data transfer
(b) DMA-based data transfer 3. In serial data transmission, every byte of data is
(c) Polling mode data transfer padded with a `0’ in the beginning and one or two
(d) Clock cycle overriding 1’ s at the end of byte because
(a) Receiver is to be synchronized for byte reception.
40. Which of the following architecture is not suitable
(b) Receiver recovers lost `0’ and `1’ from these
for the following SIMD architecture?
padded bits.
(a) Vector processor (b) PLA-based processor (c) Padded bits are useful in parity computation.
(c) Von Neumann (d) PAL-based processor (d) None of these.
Chapter 2.indd 98 4/9/2015 9:50:17 AM

4. In 2’s complement addition, overflow (c) H

orizontal microprogramming, vertical micro-
programming, hardwired control.
(a) is flagged whenever there is carry from sign bit
(d) Vertical microprogramming, horizontal micro-
addition
programming, hardwired control.
(b) cannot occur when a positive value is added to
a negative value 10. A processor needs software interrupt to
(c) is flagged when the carries from sign bit and (a) test the interrupt system of the processor
previous bit match (b) implement co-routines
(d) none of the above (c) obtain system services which need execution of
5. The performance of a pipelined processor suffers if privileged instructions
(d) return from subroutine
(a) the pipeline stages have different delays
(b) consecutive instructions are dependent on each 11. Which is the most appropriate match for the items
other in the first list with the items in the second list?
(c) the pipeline stages share hardware resources
(d) all of the above List-I List-II
X. Indirect addressing I. Array implementation
6. Match the pairs in the following questions
By writing the corresponding letters only Y. Index addressing II. Writing relocatable
code
(A) IEEE 488 (P) Specifies the interface for con-
necting a single device Z. B
ase register III. Passing array as
(B) IEEE 796 (Q) Specified the bus standard addressing parameter
for connecting a computer to
(a) (X, III) (Y, I) (Z, II) (b) (X, II) (Y, III) (Z, I)
other devices including CPUs
(c) (X, III) (Y, II) (Z, I) (d) (X, I) (Y, IIII) (Z, II)
(C) IEEE 696 (R) Specifies the standard for an
instrumentation bus 12. Consider the following data path of a simple non-
(D) RS232-C (S) Specifies the bus standard for the piplelined CPU. The registers A, B, A1, A2, MDR,
“back plane” bus called multibus the bus and the ALU are 8-bit wide. SP and MAR
are 16-bit registers. The MUX is of size 8 × (2:1)
(a) (A) − (P); (B) − (R); (C) − (S); (D) − (Q)
and the DEMUX is of size 8 × (1.2). Each memory
(b) (A) − (P); (B) − (Q); (C) − (R); (D) − (S)
(c) (A) − (Q); (B) − (S); (C) − (R); (D) − (P)
operation takes 2 CPU clock cycles and uses MAR
(d) (A) − (R); (B) − (S); (C) − (P); (D) − (R)
(memory address register) and MDR (memory
data register). SP can be decremented locally
7. When an interrupt occurs, an operating system
(a) Ignores the interrupt
A2 A1
(b) Always change state of interrupted process MUX
B A DEMUX
after processing the interrupt 1:2 1:2
(c) Always resumes execution of interrupted pro-
cess after processing the interrupt
(d) May change state of interrupted process to dcr SP MAR MDR
`blocked’ and schedule another process.
8. RAID configuration of disks are used to provide The CPU instruction “push r”, where r = A or B
has the specification
(a) Fault-tolerance (b) High speed
(c) High data density (d) None of the above M[sp] ← r
SP ← SP − 1
9. Arrange the following configuration for CPU in
decreasing order of operating speeds: hardwired How many CPU clock cycles are needed to execute
control, vertical microprogramming, horizontal the “push r” instruction?
microprogramming. (a) 2 (b) 3 (c) 4 (d) 5
(a) Hardwired control, vertical microprogramming, 13. In the absolute addressing mode,
horizontal microprogramming.
(b) Hardwired control, horizontal microprogram- (a) the operand is inside the instruction.
ming, vertical microprogramming. (b) the address of the operand is inside the
instruction.
Chapter 2.indd 99 4/9/2015 9:50:18 AM

(c) the register containing the address of the oper- (a) A - 4, B - 3, C -1, D - 2
and is specified inside the instruction. (b) A - 2, B - 1, C -3, D - 4
(d) the location of the operand is implicit. (c) A - 4, B - 3, C -2, D - 1
(d) A - 2, B - 3, C -4, D - 1
14. What are the states of the auxiliary carry (AC)
and carry flag (CY) after executing the following 19. I/O redirection
8085 program? (a) implies changing the name of a file
MVI H, 5DH (b) can be employed to use an existing file as input
MIV L, 6BH file for a program
MOV A, H (c) implies connecting two programs through a pipe
ADD L (d) none of the above
(a) AC = 0, CY = 0 (b) AC = 1, CY = 1 20. The main difference(s) between a CISC and a RISC
(c) AC = 1, CY = 0 (d) AC = 0, CY = 1 processor is/are that a RISC processor typically
15. Horizontal microprogramming (a) has fewer instruction

(b) has fewer addressing modes and more registers
(a) does not require use of signal decoders (c) is easier to implement using hardwired control logic
(b) results in larger-sized microinstructions than (d) all of these
vertical microprogramming
(c) uses 1 bit for each control signal 21. How big is a 4-way set-associative cache made up of
(d) all of the above 16 bits word, 4 words per line and having 1024 sets?
(a) 32 KB (b) 64 KB
16. For the daisy chain scheme of connecting I/O
(c) 128 KB (d) 256 KB
devices, which of the following statements is true?
21. Consider the expression to be evaluated as X =
(M + N × O)/(P × Q). How many three address
(a) It gives non-uniform priority to various devices.
(b) It gives uniform priority to all devices.
instruction are required to evaluate the above
(c) It is only useful for connecting slow devices to
given expression?
a processor device.
(d) It requires a separate interrupt pin on the pro- (a) 3 (b) 4 (c) 5 (d) 6
23. In X = (M + N × O)/(P × Q), how many one-
cessor for each device.
17. A microprogram control unit is required to gener- address instructions are required to evaluate it?
ate a total of 25 control signals. Assume that during (a) 4 (b) 6 (c) 8 (d) 10
any microinstruction at most two control signals
are active. Minimum number of bits required in 24. A decimal number has 64 digits. The number of bits
the control word to generate the required control needed for its equivalent binary representation is
signal is . (a) 200 (b) 213 (c) 246 (d) 277
18. The correct matching for the following pairs is 25. Determine the speed up obtained from pipelining if
latencies for each stage in single cycle processor is
A. DMA I/O 1. High-speed RAM given as:
B. Cache 2. Disk
C. Interrupt I/O 3. Printer IF ID ALU MEM WB
D. Condition code 4. ALU 45 ns 20 ns 52 ns 44 ns 18 ns
register
Set 1
1. (c) 3. (c) 5. (d) 7. (b) 9. (a)

2. (b) 4. (c) 6. (d) 8. (c) 10. (a)
Chapter 2.indd 100 4/9/2015 9:50:18 AM

11. (c) 18. (d) 25. (a) 32. (b) 39. (b)
12. (c) 19. (b) 26. (b) 33. (b) 40. (c)
13. (b) 20. (a) 27. (a) 34. (c) 41. (b)
14. (d) 21. (a) 28. (c) 35. (c) 42. (a)
15. (a) 22. (b) 29. (b) 36. (d) 43. (a)
16. (a) 23. (b) 30. (c) 37. (c) 44. (c)
17. (c) 24. (a) 31. (d) 38. (d) 45. (b)
Set 2 Hardwired control > horizontal microprogramming

> vertical microprogramming
1. (c) Indirect addressing mode → Pointer access
10. (c) A processor needs software interrupts to obtain
Immediate addressing mode → Constant access system services which need execution of privileged
Auto decrement addressing mode → Loops instructions.
2. (c) An opcode is the portion of a machine lan- 11. Pointers are used for accessing indirect addresses.
guage instruction that specifies the operation to be
Indirect addressing mode → Pointer access
performed.
In case of immediate addressing, the constant rep-
3. (a) Bits in the beginning and at the end of byte are resent the effective address.
known as start and stop bits, respectively. Start Immediate addressing mode → Constant access
and stop bits are used for synchronization purpose. Auto decrement/increments are used for loop
4. (b) When two 2’s complement numbers are added, variables.
overflow occurs in the following situations: Auto decrement addressing mode → Loops
Both operands are positive and the result is negative. 12. (a) Push `r’ memory operation needs 2 clocks.
Both operands are negative and the result is positive. 13. (b) In absolute addressing mode, the address of the
So option (b) is true, overflow does not occur when operand is inside the instruction.
positive and negative numbers are added.
14. (c)
5. (d) Different hazards are caused due to various
dependencies. Different dependencies for pipelined Carry Auxilary carry
processor are as follows: 0 1 1 1 1 1 1 1
Structural dependency is due to different delays in (5D) # (0 1 0 1 1 1 0 1)B
pipelined stages. +(6D) # + (0 1 1 0 1 0 1 1)B
Control dependency is due to consecutive instruc-
tions are dependent on each other. (1 1 0 0 1 0 0 0)B
Data dependency is due to hardware resources
AC = 1 and C Y = 0
sharing.
15. (d) Features of horizontal microprogramming are
6. (c) (A) IEEEE 488 - (Q)
(B) IEEEE 796 - (S) (i) It does not require use of signal decoders
(C) IEEEE 696 - (R) (ii) It results in larger-sized microinstructions
(D) RS232-C - (P) than vertical microprogramming
(iii) It uses 1 bit for each control signal
7. (c) When interrupt is caused, the execution of cur-
rent instruction is stopped. After handling inter- 16. (a) The daisy chaining method of establishing
rupt, the program resumes its execution. priority consists of a serial connection of all devices
that request an interrupt. The device with the
8. (a) RAID is random array of independent disks highest priority is placed in the first position, fol-
that combines multiple disk drive components into lowed by lower-priority devices up to the device
a logical unit. RAID configuration provides fault- with the lowest priority, which is placed last in
tolerance and high speed. the chain. The farther the device is from the first
9. (b) Horizontal micro programming has high paral- position, the lower is its priority. Therefore, daisy
lelism than vertical. So, the speed order is: chain gives non-uniform priority to various devices.
Chapter 2.indd 101 4/9/2015 9:50:18 AM

17. (10) To generate 25 control signals, 5 bits are 22. (b) Three-address instructions are as follow:
required. To generate two control signals, the fol-
MUL R1, N, O
lowing scheme will be used.
MUL R2, P, Q
ADD R3, M, R1
5 bits to identify first and 5 bits to identify second DIV X, R3, R2
including the case when one of them is not present.
So total bits required = 10. 23. (c) LOAD P (AC ← P)
18. (b) DMA I/O - Disk MPY Q (AC ← AC × Q)
Cache - High-speed RAM STORE X (X ← AC)
Interrupt I/O - Printer LOAD N (AC ← N)
Condition code register - ALU MPY O (AC ← AC × O)
19. (c) I/O redirection implies connection two pro- ADD M (AC ← AC + M)
grams through a pipe. DIV X (AC ← AC/X)
20. (d) The major characteristics of a RISC processor are STORE X (X ← AC)
(i) Relatively few instruction 24. (b) The number of bits is
(ii) Relatively few addressing modes
(iii) More registers 1064 − 1 = 2x − 1
(iv) Hardwired rather than microprogrammed ⇒ 1064 = 2x = x = log2 1064 ≈ 213
control
21. (a) Number of bytes per line = 16 bit × 4 byte = 25. (3.44) Speed up obtained by pipelining
8 bytes
(45 + 20 + 52 + 44 + 18)
Cache size = 8 × 4 × 1024 = 32 KB =
52
= 3.44
Chapter 2.indd 102 4/9/2015 9:50:19 AM

UNIT III: PROGRAMMING AND DATA STRUCTURES
12
Number of questions
10
8
6 Marks 1
Marks 2
4
2
0
2015 2014 2013 2012 2011 2010
Year Concept
2015 Binary tree, Minimum-maximum heap, B+tree, Functions, Recursion, Kth Smallest
element in C, Hashing, Stack, C strings, Postfix equation, Binary search tree,
Merge sort, Pointers, Switch-case, Storage classes, Quick sort, Array, Complexity
2014 Loops, Inorder traversal, Binary search tree, Tree traversal
2013 Binary search tree, Heap sort, Control statements, Recursion, Queues, Array and
Pointer
2012 Switch case, Queue, Trees, Functions, Storage classes
2011 Array, Heaps, Binary search tree, Recursion
2010 Binary tree, Pointers, Heaps, Recursion, Link list
Chapter 3.indd 103 4/9/2015 9:51:28 AM

Chapter 3.indd 104 4/9/2015 9:51:28 AM
CHAPTER 3
PROGRAMMING AND DATA STRUCTURES
Syllabus: Programming and data structures: programming in C; functions, recursion, parameter passing, scope,
binding; abstract data types, arrays, stacks, queues, linked lists, trees, binary search trees, binary heaps.
3.1 INTRODUCTION worked in an appropriate manner. Usually, efficient data

structures are a key in designing efficient algorithms. In
programming, efficient data structures are used to design
A computer is a machine that manipulates information. efficient programs, both in terms of time and memory.
Computer science is the study of how information is In the following text, the programming part is discussed
organised, manipulated and can be utilised. Data struc- followed by data structure.
ture is a way of organising data in computer’s memory
or disk so that it can be used efficiently. Examples of
several common data structures are arrays, linked lists, 3.2 BASIC TERMINOLOGY
queues, stacks, binary trees and hash tables. Different
kinds of data structures are suited to different kinds
This section defines the basic terms used in computer
of applications, and some are highly specialised to spe-
programming.
cific tasks. Different types of data structures are used
in different types of applications, which are specialised
to perform the specific task. Data structures provide a 3.2.1 Programming Languages
means to manage large amounts of data efficiently, such
as large databases and Internet indexing services. Data A computer performs the tasks commanded by the user.
structures are designed to manage a large amount of data The command is given through a sequence of lines of
to suit a specific purpose so that it can be accessed and code that are understood by the computer. The art of
Chapter 3.indd 105 4/9/2015 9:51:28 AM

106 Chapter 3: PROGRAMMING AND DATA STRUCTURES
writing such a code (or a program) is called program- 1. Procedural Languages: These are those
ming. According to famous programmer Niklaus Wirth, languages which follow a certain procedure. These
are three different types-:
•• Algorithmic languages: The programmer specifies
PROGRAMS = DATA + ALGORITHMS
Every programming language has two important pil- the steps (algorithm) to be followed for accom-
lars- syntax (refers to grammatical representation) and plishing a particular task. Such languages have
semantic (refers to the meaning). A correct program built-in expressions, functions and procedures.
should be correct from syntactic view as well as from For example, C, COBOL, PASCAL.
semantic view. The different kinds of translators check •• Object-oriented languages: Instead of using func-
only the syntactic representation and the programmer tions, objects are used. The data and the func-
takes care of the semantic representation. There are dif- tions are combined in defining an object. Some
ferent generations of programming languages: of the important features are: Abstraction,
1. First Generation/1GL/(Machine level pro- Encapsulation, Polymorphism and Inheritance.
gramming): The program is written in a sequence For example, C++, Java.
of 0s and 1s (binary number system). Such a •• Scripting languages: It is a kind of language that
language is more closer to the machine and the combines different components together to per-
programs are machine dependent and difficult to form a difficult task. For example, Visual Basic
debug. The only advantage is that there is no need script, PERL.
for translator software (as the program is already 2. Non-Procedural Languages: 4GL languages
in machine code) and execution is very fast. are defined as non-procedural languages. These are
2. Second Generation /2GL/(Assembly level two different types-:
•• Functional language: The program consists of
programming): The program is written using
instructions based on mnemonics and hexadecimal
functions, such languages are used in the field
number system. Such a language is also machine-
of artificial intelligence. For example, LISt Pro-
dependent. Assembler is the translator used to
cessing (LISP).
•• Logic-based language: It is a set of rules that
convert assembly level code to machine level code.
The programs more reader friendly so are easier to
are defined and the answer is retrieved using
debug.
if−then rules. For example, PROgramming in
3. Third Generation /3GL/(High-level pro-
LOGic (PROLOG).
gramming): The program is written using
English-like statements and decimal number 3. Problem-Oriented Languages: Especially designed
system. Such a language is more closer to the user to solve certain kind of problems. These are of two dif-
and the programs are machine independent. There ferent types:
is a need of translator (compiler/interpreter) to •• Numerical problems: These contains the built-in
convert the high- level code to machine level code. functions, used to solve mathematical problems.
Such languages are user-friendly. Such languages For example, Mathematica.
are file-oriented. •• Publishing: It contains the built-in functions,
4. Fourth Generation /4GL/(Non-procedural used in publishing industry. For example,
programming): Such languages allow the user to LATEX.
specify the result instead of the describing how the
result is to be obtained. All kinds of query lan- 3.2.3 Procedural Programming and Object-
guages belong to this generation. Such languages Oriented Programming
are database-oriented.
5. Fifth Generation /5GL/(Natural Language 3.2.3.1 Procedural Programming
Programming): They are similar to 4GL and
eliminates the user or programmer to learn a spe- It uses a list of instructions in a stepwise manner to
cific language. The language used, resembles the computer. It relies on procedures. Procedures are also
human speech closely. Example- CLOUT, Q & A, known as routines or subroutines. A procedure consists
HAL (Human Access Language). of a series of computational steps that has to be carried
out. Procedural programming languages are also known
as top-down languages.
3.2.2 Classification of High-Level Languages
3.2.3.2 Object-Oriented Programming
Depending upon their usage, the high-level languages
are classified as: procedural, non-procedural and problem Object-oriented programming (OOP) is a problem-
oriented languages. solving approach in which all computations are carried
Chapter 3.indd 106 4/9/2015 9:51:29 AM

3.2 BASIC TERMINOLOGY 107
out using objects. An object is an entity of a program. 2. Modularization: Dividing the program into
It is used to perform actions and interactions with other small independent modules.
elements of the program. These are the fundamental
units of OOP, for example, a person. 3.2.4 Data Types
The other fundamental concepts of OOP are as
follows: It constitutes the type of values a variable can take and
a set of operations that can be applied to those values.
1. Class: It is a blueprint for an object. It does not
Data types in programming language can be broadly
define any data, rather it will define the object of
classified as shown in Fig. 3.1.
the class it consists and the operations performed
on the object.
2. Abstraction: It is used to provide essential char
information in a program without presenting the integer type
details. For example, in a database system, only int
data is shown, all the information regarding creat-
ing, storing, etc., is hidden. float
3. Encapsulation: It is used to place data and func- Basic data
tions at the same place. types float type
4. Inheritance: It is the process of forming a new class double
from an existing class. The new class is known as the bool
derived class and the existing class is known as the
base class. It uses the concept of reusability of code. void
5. Polymorphism: It is used to provide different
meanings or functions to the existing operators or array
functions.
6. Overloading: It is a branch of polymorphism. Data type Derived data
types function
When the existing operator or function is made to
operate on a new data type, then it is known as
overloading. pointer
3.2.3.3 Structured Programming
structure
A kind of programming approach that revolutionized
the programming concept, proposed by E. Dijkstra’. union
User defined
In 1968, Dijkstra submitted the famous article in data types
ACM Journal, highlighting the drawbacks of using the class
GOTO statement. He advocated that program written
without the use of GOTO statement (or JUMP state- enumeration
ment) were more stable. Nowadays, more and more
Figure 3.1 | Classification of data types in
programmers are following such kind of approach in
writing programs. Other notable attributes of struc- programming languages.
tured programming are:
1. Top-down analysis: A large problem is sub- Table 3.1 gives detail about basic data types with its
divided into smaller sub-problems. storage sizes and value ranges.
Table 3.1 | Basic data types, their storage size and value ranges
Type Storage Size Value Range (Depends on Compiler Type)
char 1 byte −128 to 127 or 0 to 255
unsigned char 1 byte 0 to 255
signed char 1 byte −128 to 127
int 2 or 4 bytes −32,768 to 32,767 or −2,147,483,648 to 2,147,483,647
unsigned int 2 or 4 bytes 0 to 65,535 or 0 to 4,294,967,295
(Continued)
Chapter 3.indd 107 4/9/2015 9:51:29 AM

Type Storage Size Value Range (Depends on Compiler Type)

short 2 bytes −32,768 to 32,767
unsigned short 2 bytes 0 to 65,535
long 4 bytes −2,147,483,648 to 2,147,483,647
unsigned long 4 bytes 0 to 4,294,967,295
float 4 bytes 3.4E − 38 to 3.4E + 38
double 8 bytes 1.7E − 308 to 1.7E + 308
long double 10 bytes 3.4E − 4932 to 1.1E + 4932
C language is a weakly typed language so it allows or take decisions and branches. For that purpose, any
implicit and explicit type conversions (Fig. 3.2). The programming language provides control flow statements
compiler promotes each term in a binary expression to that specifies the control flow of program or the order of
the highest precision operand. execution of statements (Fig. 3.3).
3.2.5 Control Flow Statements 3.2.5.1 Conditional Statements
Programs consist of several statements and execution is These are used to execute a statement or a group of
not limited to a linear sequence of statements. During statements on the basis of certain conditions. These con-
its execution, a program may repeat segments of code, ditions are as follows:
char
int
unsigned int
Wide conversion Narrow conversion
(High precision) long
(Low precision)
unsigned long
long long
unsigned long long
float
double
long double
Figure 3.2 | Type conversion in C.
Chapter 3.indd 108 4/9/2015 9:51:29 AM

Control statements
Sequenital Branch Iterative or loop

statements statements statements
Conditional/ Unconditional
while
Decision control branch
if goto for
if-else break do-while
nested if continue
else-if return
switch
Figure 3.3 | Control flow statements.
1. If: When the condition is satisfied then the state- 3.2.5.2 Loops
ments inside the parenthesis {} will be executed.
2. If-Else: If the condition is true then the state- These are used to execute a block of code several number
ments between if and else will be executed. If it of times. The statements are executed sequentially.
is false then the statement after else will be exe- Loop executes a statement or a group of statements
cuted. In nested if, one or more if-else block(s) lie multiple times. When the execution of statements in
within the parent if statement. the loop is unknown, then this concept is known as
3. Else-If: If the condition is true then the state- odd loop.
ments between if and else if will be executed. If it Loops are of the following three types:
is false then the condition in else if is checked, and
if it is true it will be executed. 1. while loop: It repeats a statement or group of
4. Switch: It is also known as matching case state- statements while a given condition is true. It tests
ments. It is used to test for equality against a list the condition before executing the loop body.
of values. It matches the value in variable with 2. for loop: It executes a sequence of statements
any of the case inside the switch. If it matches the multiple times.
case defined in the switch then that match will be 3. do … while loop: It is similar to while statement,
executed. If none of the cases match the statement except that it tests the condition at the end of the
then default will be executed. loop body.
Chapter 3.indd 109 4/9/2015 9:51:30 AM

Loops can be nested, that is, execution of one loop inside Note that the first element is numbered 0 and so on.
another loop. Here, the size of the array age is five times the size of
There are various control statements used for changing int because there are five elements.
execution from its normal sequence. These are as follows: Starting address of an array is also known as the base
1. break statement: It terminates the loop or address of the array. Suppose, the starting address of
switch and transfers the execution to the statement age[0] is 1000 and the size of int is 4 bytes. Then, the
immediately following the loop or switch. next address (address of age[1]) will be 1004, address of
2. continue statement: It skips the remainder of age[2] will be 1008 and so on.
the loop body and immediately begins execution
from the top.
3.2.6.2 Important Things to Remember in
3. goto statement: It transfers control to the
labelled statement. C Arrays
4. return statement: It transfers control from the
Following are the important points for arrays:
function to the calling function.
1. Suppose you want to declare an array of 10 stu-
3.2.6 Array dents. For example: arr[10]. Only array members
from arr[0] to arr[9] are used. But, if element arr[10],
Array handles similar types of data. For example, stor- arr[13], etc. are used, then the compiler may not
ing marks of 100 students. show error but may cause fatal error during pro-
An array is a set of homogenous elements, which uses gram execution.
contiguous memory locations. The lowest address cor- 2. The size of the array may not be defined during
responds to the first element and the highest address to initialisation. For example,
the last element. int age[] = {2, 4, 35, 5, 8};
3.2.6.1 Declaring an Array In this case, the compiler determines the size of the
array by calculating the number of elements of an
Array can be declared as follows: array.
type arrayName [arraySize]; 3. If we do not provide value to any position in the
where array it will store 0 by default. For example,
1. type represents the type of element an array stores. int age[5] = {2, 4, 35, 5};
For example, if array stores integer elements then
In this above case, first four places in the array will
type of array is ìnt’.
be filled by given values and the last will be filled
2. arrayName represents the name given to the array.
by 0.
It can be any string.
3. arraySize represents the number of elements the int age[5]= { };
array stores. It is showed in [].
Now array age is initialised with 0 at every
For example, an array of 10 integers can be defined as position.
follows:
int arr[10] 3.2.6.3 Address Calculation in an Array
Arrays are of the following two types:
The address can be calculated as follows:
1. One-dimensional arrays
2. Multi-dimensional arrays
One-Dimensional Arrays
Initialisation of One-Dimensional Array In one dimension, an array À’ is declared as follows:
Arrays can be initialised at declaration time in this A[lb… …ub]
source code as follows (Fig. 3.4):
int age[5] = {2, 4, 35, 5, 8} where lb is the lower bound of the array and ub is the
upper bound of the array.
age[0] age[1] age[2] age[3] age[4]
Suppose we want to calculate the ith element address,
2 4 35 5 8 then
Figure 3.4 | Initialisation of one-dimensional array. Address (arr[i]) = BA + (i − lb) * c;
Chapter 3.indd 110 4/9/2015 9:51:30 AM

where BA is the base address of array, lb is the lower 3.2.7 Function and Recursion
bound of array and c is the size of each element.
A function is a group of statements, which is used to
Two-Dimensional Arrays perform a task and recursion is a programming tech-
nique, which is used to express operations in terms of
In two dimensions, an array À’ is declared as follows: themselves.
A[lb1 … …ub1][lb2 … … ub2]
where lb1 is the lower bound for row, lb2 is the lower 3.2.7.1 Function
bound for column, ub1 is the upper bound for row and
ub2 is the upper bound for column. A function is a set of statements that together perform
a task. Every C program has at least one function which
1. Row Major Order: is main(), and programs can have additional functions.
Address (arr[i][j]) = BA+ [i-lb1) * Nc+ In other words, a function is a self-contained block,
(j-lb2)] * c which takes some input and gives output after processing
2. Column Major Order: (Fig. 3.5). A function is known with various names such
as a method or a sub-routine or a procedure, etc.
Address (arr[i][j]) = BA + [j - lb2) * Nr +
(i - lb1)] * c
Input x
where BA is the base address, Nr is the number of
rows = (lb2 − lb1 + 1) and Nc is the number of
columns = (ub2 − ub1 + 1). Function F
Output ƒ(x)
Problem 3.1: Consider the following array:
a[1 … … 500] Figure 3.5 | A view of function.
Base address = 1000
Size of each element = 3 bytes
Find address of a[397]. Example 3.1
Solution: intavg(inta,intb,intc){
float result;
Address (a[397]) = BA + (i − lb) * c result = (a+b+c)/3;
= 1000 + (397 − 1) * 3 return result;
}
= 1000 + 396 * 3
= 2188
Parts of a Function
A function has the following parts:
Problem 3.2: Consider the following array:
1. Function prototype
a[50 … … 299][299 … … 500] 2. Definition of a function
Base address = 0 3. Function call
4. Actual and formal arguments
Size of each element = 5 5. Return statement
Find the address of a[250][499] by row major order.
These parts have been discussed in detail as follows:
Solution: 1. Function prototype: Function prototype con-
Address(arr[i][j])= BA +[i−lb1)* Nc tains function declaration, which tells the compiler
+(j−lb2)]* c about a function name, return type, argument
Address(a[250][499])= BA +[(250−50)* 201 types and function calling. Function declaration
is necessary before calling it. A full declaration
+(499 − 299)] * 5
includes the return type and the type of arguments.
= 0 +(200 * 202 + 200)* 5 Function declaration will be as follows:
= 203000
return_type function_name (parameter list);
Chapter 3.indd 111 4/9/2015 9:51:30 AM

For the above definition of function, the avg() 5. Return statement: When a return statement
function declaration is as follows: gets executed, then the function returns the value
intavg(int a, int b, int c); and transfers the control to the calling function.
Execution continues in the calling function by
In function declaration, the arguments are not following the remaining statements. A return state-
important, only their type is required. So, the ment can also return a value to the calling function.
following declaration is also correct:
intavg(int, int, int); Types of C Functions
2. Definition of a function: Function definition The following are two types of functions in C on the
consists of the following parts: basis of whether it is defined by a user or not:
•• Function name: This is the actual name of the
1. Pre-defined function: The function whose
function. It is unique. Function signature is con-
definition is already stored in the library of the
stituted with the function name and parameter
respective language is called pre-defined func-
list together.
tion. These functions are provided in the system
•• Parameters: At the time of function declara-
or particular language library. For example, in C
tion, the parameters are passed into the func-
language, printf(), scanf(), gets(), puts() are avail-
tion. The parameter list shows the type, value
able in stdio.h header file. These functions are also
of parameter and number of arguments in the
called library functions or built-in functions.
function. Parameters field is optional so that a
2. User-defined function: User-defined functions
function may contain no parameters.
are functions created by the user at the time of
•• Function body: The function body contains a
writing the program for modularity purpose. For
collection of statements that defines what the
example,
task is performing by that function.
•• Return type: A function returns a value. The int max (int num1, int num2){
return type is that data type in which a function int result;
should return a value. Some functions perform if(num1 > num2)
the desired operations without returning a value. result = num1;
There are two possibilities, the function either else
returns a value or not. When the function does result = num2;
not return a value then the void data type used. return result;
}
In the above example, the function name is avg,
the function has three integer-type parameters and
the function calculates the average of those three Function Advantages
numbers. Function has the following advantages:
3. Function call: In function call, the control is 1. It provides the reusability of code in an easy manner.
transferred to the called function when the pro- 2. It makes debugging and editing easier.
gram calls the respective function. The name of 3. It reduces the size of a program.
the function must match with the function exactly 4. It makes the logic of the program easier to design
with the name defined in the function prototype. A and understand.
called function performs a defined task, and when
its return statement is executed, it returns program 3.2.8.2 Recursion
control back to the main program. To call the func-
tion, the first step is to pass the arguments with When the function calls itself then it is called recursion.
function name, and then the return value stores in In C language, there is provision for the function to call
a specific manner. itself. When using the recursion, it is necessary to have
4. Actual and formal arguments: A parameter is exit condition from the function; otherwise, the result
a special kind of variable used in a function to refer will be infinite.
to the data provided as input to the function. These
pieces of data are called arguments. Syntactically, Writing Recursive Functions
we can pass any number of parameters to a function.
Parameters are specified within a pair of parenthe- A recursive function has the following two parts:
sis. These parameters are separated by commas (,). 1. Base case: It is the non-recursive solution and a
Parameter written in a function definition is called stopping condition.
`formal parameter’. Parameter written in a func- 2. Recursive case: It is the recursive solution of the
tion call is called àctual parameter’. problem.
Chapter 3.indd 112 4/9/2015 9:51:30 AM

It is simply a specification of the general recursive func- 3.2.8 Pointers

tion we have seen many times.
A pointer is a variable which stores the address of some
return_typefunction_name(Pass appropriate
other variables. The declaration of a pointer is done as
arguments)
{ follows:
if a simple case, return the simple value Data type *variable_name;
/* base case or stopping condition*/
else call function itself with simpler
version of problem Example 3.3
} int *p; /* pointer to an integer */
int b = 20;
Example 3.2 p = &b; /* address of variable b is stored
in p */
int sum(int n){
if(n==0) /* base case or stopping printf ("Address of variable b is: %d", &b);
condition */ printf ("Address of variable b is: %d, p);
return n; printf ("Value of variable b is: %d", *p);
else /* *p displays the value at stored address */
returnn+sum(n-1); /*self-call to function
sum() */
} The operating systems may stop to access memory and
cause the program to crash. To avoid such problem,
pointers should always be initialized before use. These
In this program, sum() function is called itself in else are initialized using free memory in the following way:
condition. If n is equal to 0 then the function returns the
value passing in the argument, but if n is not equal to 1. Memory allocation to pointers:
0 then the function calls itself. Let n = 5 initially, then int *p = malloc (sizeof (*p));
next time 4 is passed to the function and the value of 2. De-allocating memory of pointer:
the parameter is decreased by 1 until the condition is
free (p);
satisfied. At the end, when n becomes 0, the value of n
is returned, which is from 5 to 1. 3. Pointer to pointer:
sum(5)
Address Address Value
= 5 + sum(4)
= 5 + 4 + sum(3)
int *p /* pointer to an integer */
= 5 + 4 + 3 + sum(2) int **ptr /* pointer to pointer */
= 5 + 4 + 3 + 2 + sum(1) int b = 20;
= 5 + 4 + 3 + 2 + 1 + sum(0) p = &val; /* address of variable b is
stored in p */
=5+4+3+2+1+0 ptr = &p; /* address of pointer is
=5+4+3+2+1 stored in ptr */
=5+4+3+3 To print value of b:
=5+4+6 *p = 20;
**ptr = 20;
= 5 + 10
= 15 Pointer Arithmetic
1. Incrementing a pointer: Incrementing a
Advantages of Recursion
pointer variable depends on data type of the
Recursion has the following advantages: pointer variable. It is generally used in array in
1. Unnecessary calling of functions is avoided. which elements are stored at contiguous memory
2. In recursion, the function calls anywhere in the locations. For example,
program so it if flexible in nature. ptr ++ = ptr + 1 = ++ ptr = ptr + 1 *
3. In some situations, the recursion is compulsory in size_of(data type);
a programming. For example, to find the factorial (address + 1= ++ address = address++
of a given number. gives an address value)
Chapter 3.indd 113 4/9/2015 9:51:31 AM

#include<stdio.h> Actual Result = (ptr2 - ptr1) / Size

int main(){ of Data Type
int *ptr=(int *)200; /*Assume integer
takes 4 bytes*/ Suppose the address of variable n is
ptr=ptr+1; 200 and integer takes 4 bytes.
printf("New Value of ptr : %u",ptr); #include<stdio.h>
return 0; int main()
} {
int n, *ptr1, *ptr2;
Output ptr1 = &n;
New Value of ptr: 204 ptr2 = ptr1 + 2;
printf("Difference is: %d",ptr2 –
2. Decrementing a pointer: Decrementing a
ptr1);
pointer to a data variable will cause its value to be
return 0;
decremented by size of variable because memory }
required to store a variable (integer) varies from
compiler to compiler.Example Output
ptr-- = ptr - 1 = --ptr = ptr - 1 * Difference is 2
size_of(data type); ptr1 = 200
(address - 1= --address = address-- ptr2 = 200 + 2 * 4 = 208
gives an address value) ptr2-ptr1 = (208-200)/4 = 2
Both pointers numerically differ by 8
3. Adding an integer value with pointer: In C
and technically by 2 objects
Programming, it is legal to add any integer number
to pointer variable.
ptr + n = (ptr) + (n * size of data type); 3.2.8.1 Dangling Pointer
#include<stdio.h> Dangling pointer arises when an object is deleted or de-

int main() allocated, without modifying the value of pointer, so that
{ the pointer still points to the memory location of the de-
int *ptr=(int *)100; allocated memory. In other words, a pointer pointing to
ptr=ptr+3; a non-existing memory location is known as a dangling
printf("New Value of ptr : %u",ptr); pointer. For example,
return 0;
} void main ()
{
Output
int *p;
New Value of ptr : 112 {
4. Subtracting an integer value from pointer: int i=10; //This is a Block.
In C programming, it is legal to subtract any inte- p=
&i;//Variable i is local
so i’s lifetime is only
ger number from pointer variable.
within the block.
ptr - n = (ptr) - (n * size of data type) }
#include<stdio.h> pr
intf ("%d", *p); //p will point to
int main() de-allocated memory.
{ *p=*p+100;
int *ptr=(int *)100; printf ("%d",*p);
ptr=ptr - 3; }
printf("New Value of ptr : %u",ptr);
return 0; This program is perfectly running and giving output as
} 100 200. Actually this should not have happened because
in this program p points to a location i which has been
Output deleted from the memory so p becomes a dangling
New Value of ptr : 88 pointer but still the program is running.
5. Subtracting two pointers: It means subtracting Sometimes, the dangling pointer constitutes an unde-
two pointers and it gives total number of two fined behaviour as may be the reason that memory has
objects between them. not been overwritten by anything else. For example,
Chapter 3.indd 114 4/9/2015 9:51:31 AM

int *demoFun()//returns the address of {

variable i int a = 10, b = 20;
{ printf("\na = %d b = %d", a, b);
int i=10; exchange(a, b);
return &i; printf("\na = %d b = %d", a, b);
} }
int main()
{ exchange(int x, int y)
int *p; {
p=demoFun(); /*function returning int t;
address is assigning to pointer t = x;
variable.*/ x = y;
printf("%d",*p); //10 y = t;
*p=*p+20; printf("\nx = %d y = %d", x, y);
printf("%d",*p); // Garbage Value }
}
Output Before calling the function exchange(), values of a and

- 10 b are 10 and 20, respectively. After calling exchange(), x
and y also receive 10 and 20, respectively. In exchange(),
Some Garbage Value
we are interchanging values of x and y. So, the new values
According to the program, variable i have been deleted of x and y are 20 and 10, respectively. After successful
after completion of demoFun() function but still the execution of the function exchange(), control transfers to
pointer is able to access the memory location of i because the main function, but in the main function, values of a
this memory is not overwritten by another variable. It and b remain unchanged, that is, 10 and 20, respectively.
must be the same reason that dangling pointer consti- So, the output of the above program is as follows:
tutes an undefined behaviour. But when we perform
another operation, complier gives some garbage value a=10 b=20
because base pointer p does not get any data to that x=20 y=10
particular location. a=10 b=20
3.2.9 Parameter Passing In the call-by-value mechanism, only the formal argu-
ment (variables receiving value in called function from
There are various types of parameter-passing techniques calling function, that is, x and y in the above program)
in programming languages. The parameter passed is gets changed and do not reflect in actual argument (vari-
based on the program requirement. According to the able in function call, that is, a and b in function call to
parameter passing, the passing parameter depends on exchange(a, b)).
the way the calling function calls the called function.
The C programming language supports the following 3.2.9.2 Call-by-Reference Parameter-
two types of parameter-passing techniques: Passing Technique
1. Call by value In call by reference, the parameter is passing through the
2. Call by reference actual address of the value. The address of the variable
never changes throughout the execution of the program.
3.2.9.1 Call-by-Value Parameter-Passing There is a possibility to access and modify the value of
Technique the variable by using a pointer. In case of any changes
in the parameter inside the function, there is a change of
When we call a function, we pass the parameter to the the actual parameter. For example, the declaration of a
called function and it is by default passed by the value variable is as follows:
of variable in the calling function. This is called `call by
int x = 20; which tells the compiler:
value’.
1. To allocate 2-byte memory location on stack frame
of that particular function in which it is declared.
Example 3.4
2. To hold value of type integer and also associate
void exchange(int x, int y); this address with name x.
main() 3. This memory location is initialised with value 20.
Chapter 3.indd 115 4/9/2015 9:51:31 AM

printf("\na = %d b = %d", a, b);

}
void exchange(int *x, int *y)
X Location name
{
int t;
20 Location value t = *x;
*x = *y;
*y = t;
1000 Address
}
Output
Figure 3.6 | Memory map in call by reference. a = 10 b = 20
a = 20 b = 10
The selection of address for a particular variable is com-
piler dependent and is always a whole number. In Fig. 3.6, Consider the declaration of a variable. int *x; does
the value 20 can be accessed by using variable name x not mean that x is going to hold integer value, rather it
and address 1000. The symbol `&’ is used to represent simply means it is going to hold address (memory loca-
the àddress of’. tion) of integer type of variable. We are passing address
For example, expression &x in the above case returns of variables a, b to function exchange(), so while receiv-
address of variable x, which happens to be 1000. ing we are using two pointers to integer.
An address is always positive, we use operator %u for Generally, we use call-by-value mechanism if we do
unsigned integer. *’ is a special operator in C used as not need the changes made in calling function to reflect
`value at address’. It is also called indirection operator. in the called function. But if we require the changes to
For example, *(&x), which means value at reflect in the called function, we use the call-by-reference
address(&x), that is, value at address (1000). The exam- mechanism.
ple given below returns the address of a variable, the Also, if we want to return more than one value to the
value of the variable and the value of variable using called function from the calling function, we go for call-
indirection operator. by-reference value, which is not possible ordinarily.
Problem 3.3: Find the output of the following

Example 3.5 programs
#include<stdio.h>
void main( ) (a)
{ #include<stdio.h>
int x = 20; void main()
printf("\nAddress of x = %u", &x); {
printf("\nValue of x = %d", x); int a=3, *b;
printf("\nValue of x = %d", *( &x )); b = &a;
} printf("%d\n", a**b*a+*b);
}
Output (b)
Address of x = 1000 #include<stdio.h>
Value of x = 20 int main()
Value of x = 20 {
int a=10, *b, *c;
/* Call by Reference Parameter passing b=
&a; /* Assume address of a is 100
technique */ and integer is 4 byte size */
#include<stdio.h> c=b;
void exchange(a,b); *b++=*c++;
void main( ) a++;
{ pr
intf("a=%d, b=%d, c=%d\n", a,
int a = 10, b = 20; b, c);
printf("\na = %d b = %d", a, b); return 0;
exchange(&a, &b); }
Chapter 3.indd 116 4/9/2015 9:51:32 AM

enum month {jan, feb, mar, apr, may, jun,

(c) jul, aug, sep, oct, nov, dec};
#include<stdio.h> enum month this_month;
int main()
{ this_month = apr;
char s[10] = "Point";
char *const p=s; 3.2.12 Scoping
*p=’J’;
printf("%s\n", s); The scope is a part of a program where the identifier
return 0; may have direct access, and beyond that scope, a vari-
}
able cannot be accessed. In programming, variables can
be declared in the following three ways:
Solution:
1. Inside the main function
The outputs of the programs are: 2. Outside the main function
(a) 30 (b) a=11, b=104, c=104 (c) Point 3. Function parameter definition
3.2.12.1 Local Variables
3.2.10 Structures and Unions
A local variable is that variable which is declared in
Structures are used to represent a record (e.g. library). A the body of the main function. It can be used only by
union is a special data type in C, which is used to store the function in which it declares. Other function has no
different data types in the same memory location. Table information about local variable. Formal parameter of a
3.2 shows the comparison of the structures and union. function is also treated as a local variable to that func-
tion. In the following example, num, sum, r are local
Table 3.2 | Comparison between structure and union variables.
Structure Union
Example 3.6
1. T
o define a structure, 1. T
o define union, the
the `struct’ keyword ùnion’ keyword is #include<stdio.h>
is used. used. int main(){
/* Local Variable Declaration*/
2. W
hen a structure 2. W
hen a union int num,sum=0,r;
variable is created, variable is created, printf("Enter a number: ");
the compiler allocates the compiler allocates scanf("%d",&num);
memory for this memory for that while(num){
variable equal to variable equal to the r=num%10;
the sum of size of maximum size of num=num/10;
structure elements. union element. sum=sum+r;
3. E
ach structure element 3. Individual members }
within a structure is of the union share the printf("Sum of digits of number: %d",sum);
assigned to unique memory allocated. return 0;
}
memory location.
4. M
ore than one 4. O
nly one member
structure element can can be initialised at a Variables num, sum and r are local to the function main(),
be initialised at once. time. and cannot be accessible outside the main function.
5. Individual members 5. O
nly one member can
can be accessed at a be accessed at a time. 3.2.12.2 Global Variables
time.
Global variables are those variables which are defined
outside the body of the main function. The global
3.2.11 Enumerated Data Types variables will hold their value throughout the lifetime
of the program. They can be accessed inside any of the
Enumerated data types assume values which are previ- functions defined for the program. Following is an exam-
ously declared. For example, ple using local variable and global variable.
Chapter 3.indd 117 4/9/2015 9:51:32 AM

Example 3.7 1. Register

2. Auto
#include<stdio.h> 3. Static
/* Global Variable Declaration*/ 4. Extern
int r;
int main(){ Storage, default value, scope and life of these four stor-
/* Local Variable Declaration*/ age classes are given in Table 3.3.
int num,sum=0;
printf("Enter a number: ");
scanf("%d",&num); Table 3.3 | Storage, default value, scope and life of
while(num){ the storage classes
r=num%10; Class Auto Register Static Extern
num=num/10;
sum=sum+r; Storage Memory CPU Memory Memory
} registers
printf("Sum of digits of number: %d",sum); Default Garbage Garbage Zero Zero
return 0; value value
}
Scope Local to Local to Local Global
block block
Variables num, sum are local to the function main(), Life Until Until Persists Till the
and cannot be accessible outside the main function. block block between end of
But the variable r is declared outside the main as a different program
global variable, it can be accessed by any function in function
this program. calls
If local variable and global variable have the same
name, then preference will be given to the value of the
local variable inside the function. 3.2.13 Binding
It is an association of an attribute with a program

Example 3.8 component in a program. For example, the data type of
the value of a variable is an attribute, which is associ-
#include<stdio.h>
ated with the variable name. The binding time for an
/* Global Variable Declaration*/
attribute is the time at which the binding occurs.
int a=10;
int main(){ In C programming, the binding time for a vari-
/* Local Variable Declaration*/ able type is when the program is compiled, but the
int a=5,b=20,sum; value of the variable is not bound until the program
sum= a+b; executes.
printf("Sum of number: %d",sum);
return 0;
} 3.2.13.1 Early Binding
In early binding, the linker copies the referenced module

into the executable image of the program at compila-
In the above example, variable à’ is declared as global tion/link time. The referenced module is a part of the
and local both. And variable `b’ is declared as local to executable image.
the main function. When the sum of à’ and `b’ are cal-
culated by a compiler then priority is given to that à’ 1. Advantages of early binding:
which is declared as local. So the program will print 25 •• It is simple and the best performer.
as the sum of the variables. •• It is at least twice as fast as late binding.
•• The executable is stand alone.
3.2.12.3 Storage Classes
2. Disadvantages of early binding:
A storage class defines the scope and visibility of vari- •• The executable image is large.
ables and functions in C program. The following four •• If the referenced module changes then each
types of storage classes are available: executable must be re-linked.
Chapter 3.indd 118 4/9/2015 9:51:32 AM

3.3 STACK 119
•• If there is more than one executable using the The basic operations on stack are as follows
module at the same time, then multiple copies 1. PUSH: To add an element to the stack.
of it are used in memory. 2. POP: To remove an element from the stack.
3.2.13.2 Late Binding
3.3.1 PUSH Operation on Stack
In late binding, the linker copies only a stub code for the
referenced module into the executable image. It is also When we try to add elements to the stack then
known as dynamic binding. The stub code loads the ref- the o peration is called PUSH operation on stack
erenced module into memory at load/run time. (Fig. 3.7). The element is placed only on the top of
the stack and then the stack position gets increased
1. Advantages of late binding: by 1.
•• The size of the executable image is small.
•• On changing the referenced module, re-link of
the executable is not required.
•• Most operating systems can share the referenced
module. D
2. Disadvantages of late binding: C D C
•• The executable is dependent on the shared B PUSH B
library at runtime. A A
•• Since the module is shared, it must be re-entrant STACK BEFORE STACK AFTER
and thread safe. Figure 3.7 | PUSH operation on stack.
3.2.14 Abstract Data Types

3.3.1.1 Code for PUSH Operation
An abstract data type is an object with a specification
of the components. It is independent of implementation void push(int token)
details. {
Examples: char a;
if(top==MAX-1)
1. Stack: Operations such as push an item onto
{
printf("Stack full"); ®
the stack, pop an item from the stack, ask if the
is Full Condition
stack is empty; implementation may be as array or check
return;
linked list. }
2. Queue: Operations such as add to the end of the do
queue, delete from the beginning of the queue, ask {
if the queue is empty; implementation may be as
printf("\nEnter the token to be inserted:");
array or linked list or heap.
scanf("%d",&token);
3. Search structure: Operations such as insert an top=top+1;
item, ask if an item is in the structure and delete stack[top]=token;
an item; implementation may be as array, linked printf("do you want to continue insertion
list or tree. Y/N");
a=getch();
}
3.3 STACK while(a==’y’);
}
A stack is an ordered collection of items in which the
item is added and removed in order. In stack, the
element is added on the top of the stack. Only one ele-
3.3.2 POP Operation on Stack
ment, which is placed on the top, is removed from the
Whenever we try to remove an element from the stack
top of the stack. The stack is called last-in first-out
then the operation is called POP operation on stack
(LIFO) structure. Stack can be implemented by linked
(Fig. 3.8).
list and array.
Chapter 3.indd 119 4/9/2015 9:51:33 AM

EMPTY
STACK
TOP 15
12 TOP 12
10 10 10 TOP
5 5 5 5 TOP
STACK INITIALLY POP POP POP POP

Figure 3.8 | POP operation on stack.
3.3.2.1 Code for POP Operation how we are accustomed to writing expressions

in standard mathematical notation.
The following is the code of POP function in which we •• Postfix: In postfix notation, the operator imme-
use the return statement to send the result back to the diately follows its operands.
calling function •• Prefix: In prefix notation, operands immediately
follow operator.
int pop()
{
int t; Infix to Postfix Conversion
if(top==-1)
Algorithm for infix to postfix conversion:
{
printf("Stack empty"); ®
Is Empty Condition 1. You create a stack and an output string.
return -1; check 2. Then you read the infix string one token at a time,
} each token is either an identifier or an operator:
t=stack[top];
•• If the token is an identifier (e.g. a variable or a
top=top-1;
constant), then you append it onto the end of
return t; the output string.
} •• If the token is an operator, then you compare its
precedence with the precedence of the operator
3.3.3 Application of Stack on the top of the stack.
Following are the applications on stack: (a) If the operator on the top of the stack has
lower precedence, then push the new opera-
1. Parsing: Stacks are used by compilers to check
tor onto the stack.
the syntax of a program and for generating execut-
(b) Else, pop the operator from the top of the
able code.
stack and append it to the output string.
2. Reversing a list: Stack can be used for reversing
Then push the new operator onto the
a list.
stack.
3. Calling function: When a function is called all
local storage for the function is allocated on the (c) Otherwise,
stack and return address is also stored on the stack. (i) If the token is an open parenthesis, then
4. Recursive function: Stack can be used for push it onto the stack.
implementing recursive functions. (ii) If the token is a close parenthesis, then
5. Expression conversion: pop and append the operators from
•• Infix: An infix expression is one in which opera- the stack until the open parenthesis is
tors are located between their operands. This is popped. Discard both parentheses.
Chapter 3.indd 120 4/9/2015 9:51:33 AM

3.3 STACK 121
Problem 3.4: Convert the given expression into postfix.

Expression = (A + B) * (C − D)
Solution:
Expression Stack Output Comment
(A + B) * (C − D) Empty Initial
A + B) * (C − D) ( Push
+ B ) * (C − D) ( A Print
B) * (C − D) (+ A Push
) * (C − D) (+ AB Print
* (C − D) AB+ Pop until ( and print

(C − D) * AB+ Push
C − D) *( AB+ Push
− D) *( AB + C Print
D) *(− AB + C Push
) *(− AB + CD Print
End * AB + CD− Pop until (
End Empty AB + CD− Pop every element
Infix to Prefix Conversion 3. Apply algorithm infix to postfix on the expression
that comes from step 2.
Algorithm for infix to prefix conversion:
4. And finally reverse the output of step 3.
1. Reverse the given infix expression.
2. Make every `(` as ’)’ and every `)’ as `(’.
Problem 3.5: Convert the given expression into prefix:

Expression = (A + B ∧ C) * D + E ∧ 5
Solution:
Step 1: Reverse the infix expression.
5 ∧E + D*) C ∧B + A (
Step 2: Make every `(` as ’)’ and every `)’ as `(’
5 ∧E + D * (C ∧B + A)
Step 3: Convert expression to postfix form.

∧ ∧
5 E + D * (C B + A) Empty − Initial
∧ ∧
E + D * (C B + A) Empty 5 Print
∧ ∧
E + D * (C B + A) 5 Push
∧ ∧
+D * (C B + A) 5E Push
(Continued)
Chapter 3.indd 121 4/9/2015 9:51:33 AM

Table | Continued
D * (C ∧B + A) + 5E ∧ Pop and push
∧ ∧
*(C B + A) + 5E D Print
(C ∧B + A) +* 5E ∧D Push
C ∧B + A) +*( 5E ∧D Push
∧ ∧
B + A) +*( 5E DC Print
∧ ∧
B + A) +*( 5E DC Push
∧ ∧
+ A) +*( 5E DCB Print
∧ ∧
A) +*(+ 5E DCB Pop and push
∧ ∧
) +*(+ 5E DCB A Print
∧ ∧
End +* 5E DCB A+ Pop until `(’
∧ ∧
End Empty 5E DCB A+*+ Pop every element
Step 4: Reverse the expression.

+*+A ∧BCD ∧E5
3.4 QUEUE printf("\n Queue Overflow ");

}
elseif(front==-1&&rear==-1)
Queue is a special abstract data type storage structure. {
The access to elements in a queue is restricted, unlike front=rear=0;
queue[rear]=x;
arrays. Enqueue and dequeue are two main operations
used for the insertion of element in a queue and removal
}
of an element from a queue, respectively. An item can else
be inserted at the end known as rear of the queue and {
removed from the end known as front of the queue. It is rear++;
also called a first-in first-out (FIFO) list. queue[rear]=x;
It has the following five properties: }
}
1. Capacity: It stands for the maximum number of
elements a queue can hold. 2. Dequeue: Removes the front element from q and
2. Size: It stands for the current size of the queue. returns its value. It can be declared as follows:
3. Elements: It is the array of elements.
voidDequeue(queue)
4. Front: It is the index of the first element (the {
index at which the element has been removed). intx;
5. Rear: It is the index of the last element (the index if(front==-1)
at which the element has been inserted). {
printf("\n Underflow");
3.4.1 Basic Operations on Queue }
x=queue[front];
Queue has the following basic operations: if(front==rear)
{
1. Enqueue: Inserts item x at the rear of the queue q.
front=rear=-1;
Queue can be declared as follows:
}
voidEnqueue(queue, x) else
{ {
if(front==0&& rear==MAX-1) if(front==MAX-1)
{ front=0;
Chapter 3.indd 122 4/9/2015 9:51:33 AM

3.5 LINKED LIST 123
else 3.4.3 Applications of Queues

front++;
pr
intf("\n The deleted element Queues have many applications in computer systems:
is: %d", x);
} 1. Jobs in a single-processor computer
2. Print spooling
} 3. Information packets in computer networks
3.4.2 Types of Queue 3.5 LINKED LIST

Queue is of the following types:
1. Simple queue or linear queue: In this data Linked list is a dynamic data structure whose length
structure, operations, such as, insertion and dele- can be increased or decreased at runtime. Linked list
tion occurs at the rear and front of the list, respec- basically consists of memory blocks that are located at
tively (Fig. 3.9). random memory locations. Node of a linked list consists
of two parts, one is value or data and the other is pointer
Deletion Insertion to the next node.
3.5.1 Basic Operations

Front Rear Basic operations of a singly-linked list are as follows:
Figure 3.9 | Simple queue. 1. Create: Creates an empty node of linked list.
2. Insert: Inserts a new element at the end of the list.
2. Circular queue: A circular queue is a queue
3. Delete: Deletes any node from the list.
where all nodes are treated as circular (Fig. 3.10).
4. Find: Finds any node in the list.
The first node follows the last node.
5. Print: Prints the list.
Functions of these operations are listed as follows:
1. Insert: It takes the start node and data to be
Front Rear
inserted as arguments. The new node is inserted
Figure 3.10 | Circular queue. at the end of the linked list, till it reaches the last
node. Allocate the memory for the new node and
3. Priority queue: It contains items which have put data in it. Store the address in the next field of
some preset priority (Fig. 3.11). When an element the new node as NULL.
has to be removed from a priority queue; the item 2. Delete: It takes the start node (as pointer) and
with the highest priority is removed first. data to be deleted as arguments. Reach the node
for which the node next to it has to be deleted.
Assumption
•• If that node points to NULL (i.e. pointer →
30,40 highest priority
Priority next = NULL) then the element to be deleted is
11, 22,33,next priority
30 40 NULL 88,99,least priority not present in the list.
•• Else, the pointer points to a node, whose next
0
1 11 22 33 NULL node has to be removed, declare a temporary
node (temp), which points to the deleted node.
2 88 99 NULL Store the address of the node next to the tem-
Figure 3.11 | Priority queue.

porary node in the next field of the node pointer
(pointer → next = temp → next).
Break the link of the node which is next to the
4. Dequeue (double-ended queue): In this,
pointer (which is also temp). Function free() will
insertion and deletion occur at both the ends, that
deallocate the memory.
is, front and rear of the queue (Fig. 3.12).
3. Find: It takes the start node (as pointer) and data
Deletion value of the node (key) to be found as arguments.
Insertion
The first node is dummy node, so start with the
Rear Insertion second node. Iterate through the entire linked list
Deletion Front
and search for the key. Until next field of the pointer
Figure 3.12 | Double-ended queue. is equal to NULL, check if pointer → data = key.
Chapter 3.indd 123 4/9/2015 9:51:34 AM

•• If the condition holds, then the key is found. Start=temp->next;

•• Else, move to the next node and search (pointer = free(temp);
pointer → next). return 1;
•• If key is not found return 0, else return 1. }
else
4. Display: Function takes the start node (as pointer) {
as an argument. If pointer = NULL, then there is prev->next=temp->next;
no element in the list. Else, print the data value of free(temp);
the node (pointer → data) and move to the next return 1;
node by recursively calling the print function with }
pointer → next sent as an argument. }
else
3.5.2 Linked List Implementation {
prev=temp;
Linked list is implemented as follows: temp=temp->next;
}
1. Node Structure of Linked List:
}
structnode return 0;
{ }
int data;
structnode *next; 4. Find a Node in Linked List:
}*Start; int find (Struct node *Start,x)
2. Insert Node in Linked List: {
struct node *temp;
void insert(Start,x) temp=start;
{ if(temp==NULL)
structnode *temp,*New_Node; {
Ne
w_Node=(structnode *) printf("List is empty");
malloc(sizeof (structnode)); exit();
New_Node->data=num; }
New_Node->next=NULL;
else
temp=Start; {
if(temp==NULL) while(temp->data!=x && temp->
{ next!=NULL)
Start=New_Node; {
exit(1) temp=temp->next;
} }
else if(temp->data==x)
{ {
while(temp->next!=NULL) printf(" Node is found at address
{ %u",temp);
temp=temp->next; return 1;
} }
Start=New_Node; else
} {
} printf(" Node is not found ");
return 0;
3. Delete Node in Linked List: }
int delete(int num) }}
{
5. Display Linked List:
structnode *temp, *prev;
temp=Start; void display(start)
while(temp!=NULL) {
{ r=start;
if(temp->data==num) if(r==NULL)
{ {
if(temp==Start) return;
{ }
Chapter 3.indd 124 4/9/2015 9:51:34 AM

3.6 TREES 125
while(r!=NULL) 6. Children: It is a set of nodes (S ), which have

{ incoming edges from the same node. In Fig. 3.13, 2
printf("\n%d ",r->data); and 6 are the children of node 7.
r=r->next; 7. Parent: It is the parent of all the nodes that are
} connected to the outgoing edges. In Fig. 3.13, node
printf("\n"); 5 is parent of node 9.
} 8. Sibling: The nodes in the tree having the same
parent are said to be siblings. The nodes 2 and 6
3.6 TREES are siblings in the tree shown in Fig. 3.13.
9. Subtree: It is a set of nodes and edges comprised
of a parent and all the descendants of that parent.
A tree is a collection of nodes and edges. When we con- 10. Leaf node: It is a node that has no children. For
nect each node with the help of edges then it constructs example, node 5 and node 11 are the leaf node in
a tree. The following has the properties of a tree: Fig. 3.13.
1. Starting node is called the root node. 11. Level: The level of a node (n) is the number of
2. Except the root node every other node(n) in a tree is edges on the path from the root node to (n). For
surely connected by an edge from another node(y), example, the level of node 4 in Fig. 3.13 is three.
the other node(y) is called the parent node of n. 12. Height: It is equal to the maximum level of any
There is unique path from start node to end node. node in the tree. The height of the tree in Fig. 3.13
is three.
3.6.1 Binary Tree
A binary tree is a tree data structure. In a binary tree,

3.6.2 Types of Binary Trees
any node in the tree can have at most two children.
Binary trees are of the following types:
Binary tree are used to construct binary search trees
and binary heap. An example of a binary tree is shown 1. Strictly binary tree: Figure 3.14 shows the
in Fig. 3.13. Strictly binary tree. It has the following properties:
•• If in a tree, non-leaf nodes have exactly two
2 children then that tree is called strictly binary
tree.
7 5 •• In strictly binary tree, every leaf nodes have
degree 0 and every non-leaf nodes have degree 2.
•• A strictly binary tree with n leaf nodes has (2n − 1)
2 6 9 total number of nodes. Thus, strictly binary tree
always have odd number of nodes.
5 11 4
Figure 3.13 | Binary tree.

A Root
1. Vocabulary and definitions: We will formally

define a tree and its components. C
B
2. Node: A node is an important part of a tree. It
can have a name, which we call the `key’. It may
also have additional information.
3. Edge: An edge is another important part of a D E F G
tree. It connects two nodes to show a relation-
ship between them. Every node (except the root)
is connected with exactly one incoming edge from H I
another node. Each node may have several outgo-
ing edges. Figure 3.14 | Strictly binary tree.
4. Root: The root of the tree does not have incoming
edges.
5. Path: A path is an ordered list of nodes, which are 2. Complete binary tree: Figure 3.15 shows
connected by edges. For example, 2 → 7 → 6 → the strictly binary tree. It has the following
11 is a path. properties:
Chapter 3.indd 125 4/9/2015 9:51:35 AM

•• A complete binary tree of depth d is strictly a

binary tree having all the leaves at level d. F
•• The total number of nodes in a complete binary
tree of depth d equals (2d + 1 − 1). B G
A D I
A
C E H
C X
F B G A D I C E H
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15
E M S O Figure 3.17 | Storage representation.
Figure 3.15 | Complete binary tree. 3.6.4 Tree Applications
The following are common uses of a tree:

3. Almost complete binary tree: A binary tree 1. Manipulate hierarchical data
of depth d is an almost complete binary tree (see 2. Make information easy to search
Fig. 3.16) if: 3. Manipulate sorted lists of data
•• Each leaf in the tree is at either level d or level 4. As a workflow for compositing digital images for
d − 1. visual effects
•• For any node nd in the tree with a right descen- 5. Router algorithms
dant at level d, all the left descendants of nd
that are leaves are also at level d. 3.6.5 Binary Search Tree
•• An almost complete binary tree with N leaves
that is not strictly binary has 2N nodes. Binary search tree is a node-based binary tree data
structure with the following properties:
1. The left subtree contains the nodes with keys less
A than the node’s key.
2. The right subtree contains the nodes with keys
greater than the node’s key.
C X 3. Both the right and left subtree should also be
binary search tree.
4. There should not be any duplicate nodes.
E M S O
3.6.5.1 Tree Traversal in Binary Search
Tree
K W N
There are three types of traversal techniques for tree:
Figure 3.16 | Almost complete binary tree. 1. Preorder:
•• Traverse the root
•• Traverse all the left external nodes starting with
3.6.3 Array Representation of Binary Trees the left-most subtree
•• Traverse the right subtree starting at the left
An array can be used to store a binary tree using the external node
following mathematical relationships. 2. Postorder
•• Traverse all the left external nodes starting with
If root data is stored at index n, then left child is
the left-most subtree
stored at index 2*n and right child is stored at index
2*n + 1.
external node
Figure 3.17 demonstrates the storage representation. •• Traverse the root
Chapter 3.indd 126 4/9/2015 9:51:36 AM

3.6 TREES 127
3. Inorder V = {a, b, c, d, e} and E = {(a, b), (a, c), (a, d), (b, e),
•• Traverse the left-most subtree starting at the (c, d), (c, e), (d, e)}
left external node
•• Traverse the root
3.6.6.1 Basic Terminology
external node Following are the basic terms used for graphs (see
figure 3.19):
Problem 3.6: Consider the given tree and find the 1. Undirected graphs: It consists of a set V
inorder, preorder and postorder of the tree. of vertices and a set E of edges, each of them
connecting to two different vertices.
9 2. Directed graph: In this, edges provide a
connection from one vertex to another, but not
4 17 necessarily in the opposite direction.
3. Adjacent vertices: Connected by an edge.
4. Degree of a vertex: It is defined as the number
3 6 22 of adjacent vertices.
5. Path: It is defined as a sequence of vertices such
5 7 20 that consecutive vertices are adjacent.
6. Simple path: Path with no repeated vertex.
7. Cycle: A simple path, where the first and last
Solution: vertex are same.
8. Connected graph: When any two vertices are
Inorder Traversal: connected by some path.
The inorder traversal of the above tree will output: 9. Subgraph: When a subset of vertices and edges
form a graph.
3, 4, 5, 6, 7, 9, 17, 20, 22
10. Tree: A connected graph without cycles.
Preorder Traversal: 11. Balanced binary tree: A binary tree in which no
The preorder traversal of the above tree will output: leaf is at much greater depth than any other leaf.
Example of balanced binary search trees are red-
9, 4, 3, 6, 5, 7, 17, 22, 20 black trees, AVL trees, etc. Examples of non-binary
Postorder Traversal: balanced search trees are B trees, B+ trees, etc.
The postorder traversal of the above tree will output:
3, 5, 7, 6, 4, 20, 22, 17, 9 3.6.6.2 Traversing a Graph
There are two types of traversal techniques for graph:

3.6.6 Graphs 1. Breadth first traversal: It is a level-by-level
traversal of an ordered tree. It starts with vertex
A graph G = (V, E) is composed of V (set of vertices) I, then it traverses to neighbours of vertex i, then
and E (set of edges), where edges connect the vertices. neighbours of neighbours of vertex i and so on. It
For example, see Fig. 3.18. uses a queue.
a b a b a b a b a b
c d c d c d c d
d e Connected Path Disconnected Tree
Figure 3.18 | A graph G = (V, E). Figure 3.19 | Examples showing connected, path, disconnected and tree graphs.
Chapter 3.indd 127 4/9/2015 9:51:37 AM

Example 3.9 2
b
1 c
b
a
c e
a h
e
d i
i
g
f
f
(iii)
2
b b
1
a
a c
e
Q = {a}
i
i
3 Q = {f,i,c,e}
f
f
(i) (iv)
b b
1
a
c
a
i
i
Q = {b,f,i} f g
3
(ii)
f
(v)
Chapter 3.indd 128 4/9/2015 9:51:38 AM

3.6 TREES 129
2 2
b b 5
1 c
c
1 a 6 h
8
e
a d
4
e
i 7
g
3
i
g f
3
(ix)
f
Q = {i,c,e,g}
(vi)
b
2
b 5
a c
c
1 h
e
a d
i
e d
g
4
i f
Q = {h}
g
3 (x)
f
(vii)
b
b c a c
e
e h
a d
i d
g f g
i
f Q = {e,g,d} {a,b,f,i,c,e,g,d,h}
(viii) (xi)
2. Depth first traversal: It is a preorder traversal before returning to the current node and then tra-
of an ordered tree. It first traverses one subtree verses another subtree. It uses stack.
Chapter 3.indd 129 4/9/2015 9:51:40 AM

Example 3.10 2
b
1
b
a c
a c e
e h i
i
d
g f
f (iii)
2
b b
1 1 3
a a c
i i
f f
(i) (iv)
2 2
b b
3
1 1 c
a a
e
d
i i
f
f
(ii) (v)
Chapter 3.indd 130 4/9/2015 9:51:41 AM

3.6 TREES 131
2 2
b b 3
3 c
1 1
a c h
a e
4
4
d
e
d
i i 5
g
f
f
(vi) (ix)
2 2
b b 3
3
1 c h
c 1 6 4
a h
a e d
e 4
d 5
i
i g
g
f f
(vii) (x)
2 2
b b 3
c h
1 3 1 6 4
a c a e
h d
4 5
e d
i g
i
5
g
7
f f
(viii) (xi)
Chapter 3.indd 131 4/9/2015 9:51:43 AM

2 Source
b 8 a
3
h 2
c 5
1 6 4 3
a e 6
d e b
5 10
1
i g 4
2
d c
7
2
Figure 3.20 | Graph for finding the shortest path.
f
(xii)
2 To find the shortest path, an approach known as Greedy

method is used. The method is as follows:
b
1 Choose the vertex with smallest weight and mark the
3 path from source to the chosen vertex. Next, mark the
a
c 8 same for other vertices.
9 6 4 h
i e d Example 3.11
5 Source
g a
7
2
f 5
3
(xiii) 6
e b
10
b 1
4
2
a c h d c
2
e d
a
i g S = {a}
5
f 2
{a,b,c,d,g,e,f,h,i}
e b
(ix) d=2 d=5
3
3.6.6.3 Shortest Path
d c
For a given directed graph, where each edge has a non-
negative weight or cost, the shortest path problem is to d=∞ d=3
find a path of least total weight from the source to every
other vertex in the graph (Fig. 3.20). (i)
Chapter 3.indd 132 4/9/2015 9:51:44 AM

3.6 TREES 133
3.6.6.4 Spanning Tree
a
5 S = {a,e} A spanning tree of a graph G is a tree which contains all
2
3 the vertices of the graph G. See Fig. 3.21.
6
d=2 e b d=5 a b a b
10
4
c d e c d e
d c
Figure 3.21 | Spanning tree of a graph.
d=6 d=3
(ii)
Minimum Spanning Tree
a Minimum spanning tree in an undirected connected
S = {a,e,c} weighted graph is a spanning tree with minimum weight.
2 5
3 Prim’s algorithm:
e 1. Begin with all the vertices.
d=2 b d=4
2. Choose and draw any vertex.
3. Find the edge of least weight joining a drawn
4 vertex to a vertex not currently drawn. Draw the
1
weighted edge and the corresponding new vertex.
d c 4. Repeat step 3 until all the vertices are connected,
2
d=3 then STOP.
d=5
(iii)
Example 3.12
a S = {a,e,c,b}
b 10 e
2
4 8 7 6
e 3 b d
d=4
a 9 5 g
d=2 6 1 2 9
8 2
d c c 1 f
2
d=5 d=3 (i)
(iv)
Step 1
a S = {a,e,c,b,d} U = {a}, V − U = {b, c, d, e, f, g}
2
Closest
e b U V−U Low Cost
3 d=4
d=2 b A 4
1 c A 8
d A ∞
d c
2 e A ∞
d=5 d=3
f A ∞
(v) g A ∞
Chapter 3.indd 133 4/9/2015 9:51:46 AM

Select vertex b to include in U. b 10 e
b 10 e 4 8 7 6
4 8 7 6 a 9 g
d 5
d 2 9
a 9 5 g 2
2 9 8
8 2 c 1 f
(iv)
c 1 f
Step 4
(ii)
U = {a, b, c, d}, V − U = {e, f, g}
Step 2 Closest
U = {a, b}, V − U = {c, d, e, f, g}
U V−U Low Cost
e D 7
Closest
f C 1
U V−U Low Cost
g A ∞
C A 8
D B 8 Select vertex f to include in U.
E B 10 b 10 e
F A ∞
4 8 7 6
G A ∞
d
a 9 5 g
Select vertex d to include in U. 2 9
8 2
b 10 e
c 1 f
4 8 7 6
(v)
d
a 9 g Step 5
5
2 9 U = {a, b, c, d, f }, V − U = {e, g}
8 2
Closest
c 1 f
U V−U Low Cost
(iii)
e F 5
Step 3 g F 2
U = {a, b, d}, V − U = {c, e, f, g}

Select vertex g to include in U.
b 10 e
Closest
U V−U Low Cost 4 7 6
8 d
c D 2
a 5 g
e D 7 9 9
2
f D 9 8 2
g A ∞ c 1 f
Select vertex c to include in U. (vi)
Chapter 3.indd 134 4/9/2015 9:51:47 AM

3.6 TREES 135
Step 6 4
U = {a, b, c, d, f, g}, V − U = {e} 44
2 3
Closest 17 78
U V−U Low Cost 1 2 1
e F 5 32 50 88
1 1
48 62
Select vertex e to include in U.
b 10 e
4 8 7 6 Figure 3.22 | AVL tree.

d
a g AVL has the following properties:
9 5
2 9 1. Balance Factor of a Node: It is the difference
8 2 in heights of its two subtrees (hR − hL).
2. Balanced Node: Node with  BF ≤ 1; if  BF > 1,
c 1 f the node is unbalanced.
(vii) 3. Balance Factor of Binary Trees: It corre-
sponds to the balance factor of its root node.
•• Tree is `left-heavy’ if BF ≤ −1
Step 7
•• Tree is èqual-height’ if BF = 0
U = {a, b, c, d, e, f, g}, V − U = {} •• Tree is `right-heavy’ if BF ≥ +1
4. Balance Factor of AVL Trees: BF of each
MST Complete
node in an AVL tree can only ∈ {−1, 0, 1}.
Rotation
3.6.6.5 AVL Tree
It is the restructuring of the tree which maintains the
AVL tree is a binary tree which satisfies the height balance binary search tree property.
property (Fig. 3.22). It has a time complexity of O(log(n)). 1. Types of Rotation:
•• Single-Left rotation: Refer Fig. 3.23.
Z
Original tree
Y Y ZR
X Z Single-Left
X ZL
rotation
XL XR ZL ZR XL XR
XL, X, XR, Y, ZL, Z, ZR XL, X, XR, Y, ZL, Z, ZR

Figure 3.23 | Single-Left rotation.
Chapter 3.indd 135 4/9/2015 9:51:48 AM

•• Single-Right rotation: Refer Fig. 3.24.
Original tree XL Y
Y
Single-Right XR Z
X Z rotation
XL XR ZL ZR
ZL ZR
XL, X, XR, Y, ZL, Z, ZR XL, X, XR, Y, ZL, Z, ZR

Figure 3.24 | Single-Right rotation.
•• Double left-right rotation: Refer Fig. 3.25.
Original tree After Single-Left rotation Double Left-Right rotation

Y Y
Z
X YR Z YR X Y
XL X ZR XL ZL ZR YR
Z
ZL ZR XL ZL
XL, X, ZL, Z, ZR, Y, YR XL, X, ZL, Z, ZR, Y, YR XL, X, ZL, Z, ZR, Y, YR

Figure 3.25 | Double left-right rotation.
•• Double right-left rotation: Refer Fig. 3.26.
Original tree After Single-Right rotation Double Right-Left rotation

Y
Y
Z
YL X YL Z
Y X
Z XR ZL X
YL ZL ZR YR
ZR XR
ZL ZR
YL, Y, ZL, Z, ZR, X, XR YL, Y, ZL, Z, ZR, X, XR YL, Y, ZL, Z, ZR, X, XR
Figure 3.26 | Double right-left rotation.
Chapter 3.indd 136 4/9/2015 9:51:49 AM

IMPORTANT FORMULAS 137
3.6.6.6 Binary Heap or equal to the root element. This binary heap (see
Fig. 3.27) is known as min-heap. In addition, if all the
A complete binary tree is said to be binary heap if all elements are lesser or equal to the root element, this is
the elements below in hierarchy are either greater than called max-heap.
1. The height of a binary heap is O(log n). The
runtime of deleteMin, insert and remove is
6
O(log n).
2. The root node is the highest priority element. To
remove the highest/lowest priority object (root),
7 12 heap data structure is preferred. Priority queue is
the best example to be implemented.
3. The running time of heap sort is O(nlog n).
10 15 17
Figure 3.27 | Binary heap.
IMPORTANT FORMULAS
1. Insertion/Deletion in array: O(n) where lb is the lower bound of array and ub is the
upper bound of array.
2. Split/Merge in array: O(n)
Suppose we want to calculate the ith element
3. Time complexity of Push(): O(1) address, then
4. Time complexity of Pop(): O(1) Address (arr[i]) = BA + (i − lb) * c
5. Time complexity of Enqueue(): O(1)
where BA is the base address of array and lb is
6. Time complexity of Dequeue(): O(1) the lower bound of array and c is the size of each
7. Number of elements in queue = element
 rear − front + 1, if rear = front 15. Two-dimensional arrays


rear − front + n, otherwise In two dimensions, an array À’ is declared as:
8. Number of nodes in full binary tree: 2h+1 −1 A[lb1 … … ub1][lb2 … … ub2]
[h: levels] where lb1 is the lower bound for row, lb2 is the
9. Number of leaf nodes in full binary tree: 2h lower bound for column, ub1 is the upper bound for
row and ub2 is the upper bound for column.
10. Number of waste pointers in complete binary tree
of n nodes: n + 1 Row major order
Address (arr[i][j]) = BA + [i − lb1) * Nc +
11. Complexity of inorder, preorder and postorder tree
(j − lb2)] * c
traversal: O(n)
12. Minimum number of moves for tower of Hanoi:
2n − 1 Column major order
2n
13. Number of unique binary trees: n
C Address (arr[i][j]) = BA + [j − lb2) * Nr +
n +1 (i − lb1)] * c
14. One-dimensional arrays where BA is the base address, Nr is the number of
In one dimension, an array À’ is declared as: rows = (lb2 − lb1 + 1), and Nc is the number of
A[lb … … ub] columns = (ub2 − ub1 +1).
Chapter 3.indd 137 4/9/2015 9:51:50 AM

SOLVED EXAMPLES
1. Variables of function call are allocated in (a) 70 (b) 50

(c) 46 (d) 69
(a) registers and stack. (b) cache and heap.
(c) stack and heap. (d) registers and heap. Solution: m=2*f(3,g(4,5));
The function g(4,5) will return 20 by performing
Solution: The variables of function call can be
multiplication operations. Then function f (3,20)
stored in stack and heap.
will return 23 by performing addition operation.
Ans. (c)
In the last output of function f (3,20) is multiplied
2. Predict the output of: with 2. So, the final answer will be 46.
Ans. (c)
void main()
{ 5. A program to find the factorial of a given number
float x=1.1; is written (a) using recursion and (b) without using
double y= 1.1; recursion. Which of these will result in stack over-
if(x==y)
flow for a given number?
printf("I see You");
else (a) Only first program with recursion
printf("I hate You"); (b) Both may result for the same number
} (c) Only second program
(a) I see You (d) None of these
(b) I hate You
Solution: When the factorial program is written
(c) Cannot compare double and float
using recursion only then will it use stack, so stack
(d) Run time error
overflow will occur in recursive program only.
Solution: The stored value may or may not be Ans. (a)
exact, the precision of the value depends upon
6. If two strings are identical then strcmp() returns:
the number of bytes. Double stores value 1.1 with
more precision than float because double takes (a) True (b) 1
8 bytes, whereas float takes 4 bytes. (c) −1 (d) 0
Ans. (b)
Solution: strcmp() is a pre-defined function
3. Global variable conflicts due to multiple file occur- of C library which returns value by calculating
rence is resolved during value=(second string length — first string length).
Hence, if both the strings are identical then it will
(a) load time.
return 0.
(b) execution time.
Ans. (d)
(c) link time.
(d) parsing phase of compiler. 7. In tree construction, which one will be suitable and
efficient data structure?
Solution: A global variable provides access from
multiple files. Due to merging of one file code to (a) Queue (b) Linked list
another code, conflict occurs in the variable names (c) Heap (d) String
and this cause an error at link time.
Ans. (c) Solution: Linked list is the best suitable and effi-
cient data structure for constructing a tree because
4. #define f(a,b) a+b an item can be inserted and deleted from linked list
#define g(a,b)a*b with less cost.
main() Ans. (b)
{
8. Which of the following is not true about spanning
int m;
m=2*f(3,g(4,5)); tree?
printf("\n m is %d", m); (a) It is a tree derived from a graph.
} (b) All the nodes of a network appear on the tree
What is the value of m? only once.
Chapter 3.indd 138 4/9/2015 9:51:50 AM

SOLVED EXAMPLES 139
(c) Spanning tree cannot have at most two edges (a) Associativity
repeated. (b) Commutativity
(d) Spanning tree cannot be minimum and (c) Mathematical induction
maximum. (d) Factorial
Solution: A spanning tree is a subgraph of a given Solution: Mathematical induction is the method
graph G, which covers all the vertices of G and through which correctness of recursive function can
should be a tree. Spanning tree could be minimum be proved.
or maximum. Ans. (c)
Ans. (d)
11. In a doubly linked list, the number of pointers
9. Which of the following has compilation error in C? affected for an insertion operation will be:
(a) int n = 17; (a) 4
(b) char ch = 99; (b) 0
(c) float f = (float) 99.32; (c) 1
(d) #include<stdio.h> (d) Depends upon the nodes of the doubly linked
list.
Solution: The syntax for inclusion of header file
is #include <stdio.h>. There should be space Solution: Let a node X be inserted after node 1,
between #include and <stdio.h>. then both pointer of node X(Prv and Nxt) will be
Ans. (d) affected and node 1’s Nxt pointer will point to node
X and node 2’s Prv pointer will also point to node
10. is often used to prove the correctness X. So, total four pointers will be affected.
of a recursive function.
Node 1 Node 2 Node 3

Prv data Nxt Prv data Nxt Prv data Nxt
Ans. (a)
12. The queue data structure is to realised by using Solution: Let there are 6 elements (1, 2, 3, 4, 5
stack. The number of stacks needed would be: and 6). First we will push them in stack 1 and then
pop from stack 1 and push in stack2. The same
(a) It cannot be implemented.
behaviour will be observed when we will pop from
(b) 2 stacks.
stack2
(c) 4 stacks.
(d) 1 stacks.
Element6 Element1
Element5 Element2
Element4 Element3
Element3 Element4
Element2 Element5
Element1 Element6
Stack1 Stack2
So, only 2 stacks are required. So, option (b) is correct.

Ans. (b)
Chapter 3.indd 139 4/9/2015 9:51:51 AM

13. Postfix expression for the infix expression

(A*B − (C*D))/(F/D + E)
(a) AB*CD* − FD/E+/ (b) AB*CD − *FDE//+
(c) AB*CD − FD*E//+ (d) AB*CD − *FDE/+/
Solution:
Infix Expression Postfix Expression Stack
(A*B − (C *D))/(F/D + E)
*B − (C *D))/(F/D + E) A (
B − (C *D))/(F/D + E) A (*
− (C *D))/(F/D + E) AB (*
(C *D))/(F/D + E) AB* (−
*D))/(F/D + E) AB*C (−(
D))/(F/D + E) AB*C (−(*
/(F/D + E) AB*CD (−(*))
/(F/D + E) AB*CD*−
(F/D + E) AB*CD*− /
/D + E) AB*CD*−F /
D + E) AB*CD*−F /(/
+ E) AB*CD*−FD /(/
E) AB*CD*−FD/ /(+
AB*CD*−FD/E /(+)
AB*CD*−FD/E+/
Ans. (a)
14. There are six nodes. The different types of trees Solution: Passing an element of an array by call
that can be realised are . by name behaves similar to call by value. Therefore,
if array elements are passed as parameter then they
(a) 10 (b) 64
can produce different results for call-by-reference
(c) 58 (d) 65
and call-by-name parameter passing.
Solution: 58 trees can be realized. Ans. (d)
Ans. (c)
16. A variant record in Pascal is defined by
15. In which of the following case(s) is it possible to
obtain different results for call-by-reference and typevarirec = record
number: integer;
call-by-name parameter passing
case (var1, var2) of
(a) Passing an expression as a parameter var1; (x, y : integer)
(b) Passing an array as a parameter var2: (p,q: real)
(c) Passing a pointer as a parameter end
(d) Passing an array elements as a parameter end
Chapter 3.indd 140 4/9/2015 9:51:51 AM

SOLVED EXAMPLES 141
Suppose an array of 100 records was declared on a {float y (4 bytes) and long z(8 bytes)} 12 bytes.
machine which uses 6 bytes for an integer and 10 So, total size will be 22 bytes (10 + 12).
bytes for a real. How much space would the com- Ans. (a)
piler have to reserve for the array
(a) 3800 (b) 3200 19. In a compact single-dimensional array represen-
(c) 2800 (d) 4000 tation for lower triangular matrices (i.e. all the
elements above the diagonal are zero) of size n × n,
Solution: There are 100 records non-zero elements (i.e. elements of the lower
So 600 B for `record no’: [100 × 6] triangle) of each row are stored one after another,
starting from the first row, the index of the (i,j)th
1200 B for var(x, y): [so, (6 + 6)*100] element of the lower triangular matrix in this new
2000 B for var(p, q) : [so, (10 + 10)*100] representation is
Ans. (a) (a) i + j (b) i + j −1
17. What is the result of the following program? i −1 j ( j − 1)
(c) ( j − 1) + (d) i +
Program side-effect (input, output) 2 2
var x, result: integer;
Function f (var x: integer): integer; Solution: To find location in lower triangular
begin matrix formula is given as
x: x+1;
f:=x; i (i − 1)
LOC (i,j ) = ( j − 1) +
end 2
begin
x:=5; Ans. (c)
result:=f(x)*f(x);
writeln (result) 20. The following sequence of operation is performed
end on a stack:
(a) 5 (b) 25 PUSH(10), PUSH(20), POP, PUSH(10),
(c) 36 (d) 42 PUSH(20), POP, POP, POP, PUSH(20), POP
Solution: Function `f ’ uses `x ’ as local variable. The sequence of values popped out is
So value of `x’ does not change globally and both
(a) 20, 10, 20, 10, 20 (b) 20, 20, 10, 10, 20
the time x = 5 is passed to the function `f ’.
(c) 10, 20, 20, 10, 20 (d) 20, 20, 10, 20, 10
Ans. (c)
Solution:
18. Consider the following C declaration
struct { Operation Stack Popped
short s [5] Elements List
union{
float y; Push 10
long z; Push 10, 20
}u;
Pop 10 20
}t;
Push 10, 10 20
Assume that objects of the type short, float and
Push 10, 10, 20 20
long occupy 2 bytes, 4 bytes and 8 bytes, respec-
tively. The memory requirement for variable t, Pop 10, 10 20, 20
ignoring alignment considerations, is Pop 10 20, 20, 10
(a) 22 bytes (b) 18 bytes Pop 20, 20, 10, 10
(c) 14 bytes (d) 10 bytes Push 20 20, 20, 10, 10
Solution: The structure is created with total Pop 20, 20, 10, 10, 20
size of `short and union’ short have size of 2 byte,
as it is an array of five elements so it will take So, option (b) is correct.
10 bytes. Union will consider the maximum of Ans. (b)
Chapter 3.indd 141 4/9/2015 9:51:52 AM

21. The postfix expression for the infix expression Postfix conversion steps
A + B * (C + D)/F + D * E is A + B * CD + F/+ DE*
(a) AB + CD +* F/D + E* A + BCD +*F/DE*+
(b) ABCD + *F/DE *++ ABCD +*F/DE*++
(c) A * B + CD/F * DE++
(d) A + * BCD/F * DE++ So, option (b) is correct.
Ans. (b)
Solution: Given expression A + B * (C + D)/F
+D*E
1. Consider the following C function. Solution: B[1] is address of second one dimen-
float f(float x, int y) { sional array. Address can not be assigned to simple
float p, s; int i; variable. So it will produce compile time error.
for (s=1,p=1,i=1; i<y; i++) { Ans. (a)
p *= x/i;
3. Let T(n) be the number of different binary
s+=p;
search trees on n distinct elements. Then T(n) =
} n
return s; ∑T (k − 1) T (x), where x is
} k =1
For large values of y, the return value of the func- (a) n − k + 1 (b) n − k
tion f best approximates (c) n − k − 1 (d) n − k − 2
(a) xy (b) ex (c) ln(1 + x) (d) xx (GATE 2003: 1 Mark)
(GATE 2003: 1 Mark)
Solution:
Solution: n
Applying iterative method: T(n) = ∑ T (k − 1) T (x)
k =0
i=1 p = x/1 s=1+x T(n) = T(x) [T(0) + … + T(n−1)]
i =2 p = x * x/2 s = 1 + x + x2/2 x=n−k+1
i=3 p = x3/ 6 s = 1 + x + x2/2 + x3/ 6 Ans. (a)
from above iterations: s = 1 + x + x2/2 + x3/ 6 + ….. 4. Suppose the numbers 7, 5, 1, 8, 3, 6, 0, 9, 4, 2

s = 1 + x + x2/2! + x3/ 3! + x4/ 4! + …. = ex
are inserted in that order into an initially empty
binary search tree. The binary search tree uses
Ans. (b) the usual ordering on natural numbers. What is
the inorder traversal sequence of the resultant
2. Assume the following C variable declaration tree?
int * A[10], B[10][10];
(a) 7 5 1 0 3 2 4 6 8 9 (b) 0 2 4 3 1 6 5 9 8 7
Of the following expressions (c) 0 1 2 3 4 5 6 7 8 9 (d) 9 8 6 4 2 3 0 1 5 7
I. A[2] II. A[2][3] (GATE 2003: 1 Mark)
III. B[1] IV. B[2][3]
Solution: In-order traversal of a binary search
which will not give compile-time errors if used as
tree arrange elements in increasing order.
left-hand sides of assignment statements in a C
Ans. (c)
program?
(a) I, II and IV only (b) II, III and IV only 5. A data structure is required for storing a set of inte-
(c) II and IV only (d) IV only gers such that each of the following operations can
be done in O(log n) time, where n is the number of
(GATE 2003: 1 Mark) elements in the set.
Chapter 3.indd 142 4/9/2015 9:51:52 AM

I. Deletion of the smallest element to the start of the pop operation that removes m
II. Insertion of an element if it is not already from S. The average stack-life of an element of this
present in the set stack is
Which of the following data structures can be used (a) n(X + Y). (b) 3Y + 2X.
for this purpose? (c) n(X + Y) −X. (d) Y + 2X
(a) A heap can be used but not a balanced binary (GATE 2003: 2 Marks)
search tree.
(b) A balanced binary search tree can be used but
Solution : Time taken by Pop and Push
not a heap.
opé rations : X + X = 2X
(c) Both balanced binary search tree and heap can
be used. Time delay between two stack operations = Y
(d) Neither balanced binary search tree nor heap Total time taken = 2X + Y
can be used. Ans. (d)
7. Consider the following 2−3−4 tree (i.e. B-tree with
a minimum degree of two) in which each data item
Solution: Time taken by heap to insert and
is a letter. The usual alphabetical ordering of let-
delete n elements is O(n). Only balanced binary
ters is used in constructing the tree.
search tree takes O(log n) time for insertion and
deletion.
Ans. (b)
6. Let S be a stack of size n ≥ 1. Starting with the

L P U
empty stack, suppose we push the first n natu-
ral numbers in sequence, and then perform n pop
operations. Assume that Push and Pop operations VXZ
BHI N QT
take X seconds each, and Y seconds elapse between
the end of one such stack operation and the start of
the next operation. For m = 1, define the stack-life
of m as the time elapsed from the end of Push(m) What is the result of inserting G in the above tree?
(a) P (b) P
GL U HL U
B HI N QT VXZ BG I N QT VXZ
(a) (b)
(c) (d) None of the above

I P U
BGH L N Q T VXZ
(c) (GATE 2003: 2 Marks)
Chapter 3.indd 143 4/9/2015 9:51:53 AM

Solution: 2-3-4 B tree has minimum degree 2 and maximum 4. According to max degree, atmost 3 elements can
be there in node.
L P U L P
G
VXZ VXZ
BHI BGH
N QT N Q
This node will split due to 4 elements

Result will be
I P U
BGH L N Q T VXZ
Ans. (c)
Common Data Questions 8 and 9: The following Solution: In dynamic scoping, the variables and
program fragment is written in a programming values are resolved at run time.
language that allows global variables and does not P(100 + 5) = P(105)
allow nested declarations of functions. First Print x + 10 will print 115
global int i = 100, j = 5; Second Print x will update x = 200 + 20 = 220 and
void P(x) { will print 220
int i = 10; Ans. (a)
print(x + 10);
i = 200; 10. In the following C program fragment, j, k, n and
j = 20; TwoLog_n are integer variables, and A is an array
print (x); of integers.
}
The variable n is initialised to an integer ≥ 3,
main() {P(i + j);}
and TwoLog_n is initialised to the value of n
8. If the programming language uses static scoping 2*log2(n)
and call by need parameter passing mechanism, for (k=3; k <= n; k++)
the values printed by the above program are A[k] = 0;
(a) 115, 220. (b) 25, 220. for (k=2; k <= TwoLog_n; k++)
(c) 25, 15. (d) 115, 105. for (j=k+1; j <= n; j++)
A[j] = A[j] || (j%k);
(GATE 2003: 2 Marks) for (j=3; j <= n; j++)
if (!A[j]) printf("% d", j);
Solution: In static scoping, the variables and
values are resolved at compile time. The set of numbers printed by this program frag-
P(100 + 5) = P(105) ment is
Print x + 10 will print 115 (a) {m m ≤ n, (∃i) [m = i!]}
Print x will print 105 (b) {m m ≤ n, (∃i) [m = i2]}
Ans. (d) (c) {m m ≤ n, m is prime}
(d) {}.
9. If the programming language uses dynamic scoping (GATE 2003: 2 Marks)
and call-by-name parameter-passing mechanism,
the values printed by the above program are Solution: The condition ! A[j] is equivalent to A[j] =
(a) 115, 220. (b) 25, 220. = 0. But in array A no element is zero. So it will
(c) 25, 15. (d) 115, 105. print nothing.
(GATE 2003: 2 Marks) Ans. (d)
Chapter 3.indd 144 4/9/2015 9:51:54 AM

11. Consider the C program shown below. (a) The list is empty or has exactly one element.
(b) The elements in the list are sorted in non-
#include <stdio.h>
decreasing order of data value.
#define print(x) printf("% d —, x)
int x; (c) The elements in the list are sorted in non-
void Q(int z) { increasing order of data value.
z += x; print(z); (d) Not all elements in the list have the same data
} value.
void P(int *y) { (GATE 2003: 2 Marks)
int x = *y+2;
Q(x); *y = x-1; Solution: The given function returns 1 if the
print(x); elements are unique and are sorted increasing
} order.
main(void) {
x = 5; Ans. (b)
P(&x)
print(x); 13. Consider the following class definitions in a
} hypothetical object oriented language that sup-
ports inheritance and uses dynamic binding. The
The output of this program is language should not be assumed to be either Java
(a) 12 7 6 (b) 22 12 11 or C++, though the syntax is similar.
(c) 14 6 6 (d) 7 6 6
class P Class Q subclass of
(GATE 2003: 2 Marks) { P
void f(int i) {
Solution: { void f(int i)
P(&x) where x = 5 print(i); {
} print(2*i);
x = *y + 2 = 5 + 2 = 7 } }
Q(x) = Q (7) }
→ Here z = 7 Now consider the following program fragment:
→ z = z + x = 7 + 5 = 12 Px = new Q();
→ Print 12 Qy = new Q();
*y = x − 1 = 7 − 1 = 6 Pz = new Q();
print(x) = print 7 x.f(1); ((P)y).f(1); z.f(1);
Value returned by function P is 6. So main will where ((P)y) denotes a typecast of y to P. The
print 6. output produced by executing the above program
So, answer is 12 7 6 fragment will be
Ans. (a) (a) 1 2 1 (b) 2 1 1 (c) 2 1 2 (d) 2 2 2

12. Consider the function f defined below:
Solution:
struct item {
int data; Px = newQ();
struct item * next; Qy = newQ();
}; Pz = newQ();
int f(struct item *p) {
x : f(1); print 2 * i = 2
re
turn ((p == NULL) || (p ->next ==
NULL) || ((P)y) : f(1);
((
p->data <= p -> next -> data) && z : f(1) print 2 * i = 2
f(p-> next)));
} It will print 2 because typecast to parent class
cannot prevent overriding. So, function f(1) of class
For a given linked list p, the function f returns 1 if Q will be called not f(1) of class P.
and only if Ans. (d)
Chapter 3.indd 145 4/9/2015 9:51:54 AM

14. Which of the following is NOT an advantage ends of the array. Variables top 1 and top 2 (top
of using shared, dynamically linked libraries as 1 < top 2) point to the location of the topmost
opposed to using statically linked libraries? element in each of the stacks. If the space is to be
used efficiently, the condition for `stack full’ is
(a) Smaller sizes of executable files
(b) Lesser overall page fault rate in the system (a) ( top 1 = MAXSIZE/2) and (top 2 =
(c) Faster program startup MAXSIZE/2 + 1)
(d) Existing programs need not be re-linked to take (b) top 1 + top 2 = MAXSIZE
advantage of newer versions of libraries (c) (top 1 = MAXSIZE/2) or (top 2 = MAXSIZE)
(GATE 2003: 2 Marks) (d) top 1 = top 2 − 1
(GATE 2004: 1 Mark)
Solution: The shared, dynamic linked libraries
use smaller sizes of executable files, lesser overall Solution: If we are to use space efficiently then
page fault rate and faster startup of programs. size of any stack can be more than MAXSIZE/2.
Ans. (d) Both stacks will grow from both ends and if any
of the stack top reaches near to the other top then
15. The goal of structured programming is to
stacks are full. So the condition will be top1 =
(a) Have well-indented programs. top2 −1 (given that top1 < top2)
(b) Be able to infer the flow of control from the Ans. (d)
compiled code.
18. The following numbers are inserted into an empty
(c) Be able to infer the flow of control from the
binary search tree in the given order: 10, 1, 3, 5,
program text.
15, 12, 16. What is the height of the binary search
(d) Avoid the use of GOTO statements.
tree (the height is the maximum distance of a leaf
(GATE 2004: 1 Mark) node from the root)?
Solution: The goal of structured programming is (a) 2 (b) 3
that the user can control the flow according his (c) 4 (d) 6
requirement. (GATE 2004: 1 Mark)
Ans. (c)
Solution:
16. Consider the following C function
10
void swap (int a, int b)
{ 1 15
int temp;
temp = a ;
3 12 16
a=b;
b= temp ;
} 5
Ans. (b)
In order to exchange the values of two variables x
and y: 19. Given the following input (4322, 1334, 1471, 9679,
1989, 6171, 6173, 4199) and the hash function x mod
(a) Call swap(x, y)
10, which of the following statements are true?
(b) Call swap(&x, &y)
(c) swap(x, y) cannot be used as it does not return I. 9679, 1989, 4199 hash to the same value.
any value. II. 1471, 6171 hash to the same value.
(d) swap(x, y) cannot be used as the parameters III. All elements hash to the same value.
are passed by value. IV. Each element hashes to a different value.
(GATE 2004: 1 Mark) (a) I only (b) II only
(c) I and II only (d) III or IV
Solution: The value will not be interchanged. The
(GATE 2004: 1 Mark)
values are passed by value, not by reference.
Ans. (d) Solution:
17. A single array A[1..MAXSIZE] is used to implement Mod 10 of all the elements in the list is given below:
two stacks. The two stacks grow from opposite
Chapter 3.indd 146 4/9/2015 9:51:54 AM

4322 mod 10 = 2 23. Consider the following program fragment for

1334 mod 10 = 4 reversing the digits in a given integer to obtain a
1471 mod 10 = 1 new integer. Let n = d1 d2 … dm.
9679 mod 10 = 9
int n, rev;
1989 mod 10 = 9 rev = 0;
6171 mod 10 = 1 while (n > 0) {
6173 mod 10 = 3 rev = rev * 10 + n % 10;
4199 mod 10 = 9 n = n / 10 ;
From the list, it is clearly seen that 9679, 1989 and }
4199 have same hash values 9. But, 1471 and 6171
The loop invariant condition at the end of the i
have hash value of 1. So, options I and II are correct.
iteration is:
Ans. (c)
20. The best data structure to check whether an arith- (a) n = d1d2…dm−1 and rev = dmdm−1…dm−1+1
metic expression has balanced parentheses is a (b) n = dm+1…dm−1dm or rev = dm−1…d2d1
(c) n ≠ rev
(d) n = d1d2…1dm or rev = dm…d2d1
(a) queue. (b) stack.
(c) tree. (d) list.
(GATE 2004: 1 Mark) (GATE 2004: 2 Marks)
Solution: Stack is used to check balancing of Solution: A loop invariant is a condition that must
parenthesis. be true before and after each iteration of a loop. So
Ans. (b) n = d1d2…dm−1 and rev = dmdm−1…dm−1+1 pro-
21. Level order traversal of a rooted tree can be done vides the correct loop invariant condition.
by starting from the root and performing Ans. (a)
(a) Preorder traversal. (b) Inorder traversal. 24. Consider the following C program segment:
(c) Depth first search. (d) Breadth first search. char p [20]
(GATE 2004: 1 Mark) char * s = "string";
int length = strlen (s);
Solution: Breath first search gives the level order for (i = 0 ; i < length; i++)
traversal. p[i] = s[length - i];
Ans. (d) print f("%",p);
22. Consider the following C function: The output of the program is
int f(int n) (a) gnirts. (b) String.
{ staticint i = 1 ; (c) gnirt. (d) No output is printed.
if (n >= 5) return n;
n = n+i;
i++; Solution: In C value of pointer variable cannot be
return f(n); assigned to normal character variable.
} Ans. (d)
The value returned by f(1) is
25. A circularly linked list is used to represent a queue.
(a) 5. (b) 6. A single variable p is used to access the queue.
(c) 7. (d) 8. To which node should p point such that both the
(GATE 2004: 2 Marks) operations enqueue and dequeue can be performed
Solution: in constant time?
Iteration n i
0 1 1
1 1+1=2 2
2 2+2=4 3 Front Rear
3 4+3=7 4
4 7 ≥ 5 Return 7
P ?
Ans. (c)
Chapter 3.indd 147 4/9/2015 9:51:55 AM

(a) Rear node The postfix expression corresponding to the infix

(b) Front node expression a + b × c − d ∧ e ∧ f is
(a) abc × + def ∧ ∧− (b) abc × + de ∧ f ∧−
(c) Not possible with a single pointer
(c) ab + c × d − e f (d) − + a × bc ∧ ∧ def
∧ ∧
(d) Node next to front
Solution: Required condition is not possible with
Solution:
single pointer.
∧ ∧
Ans. (c) a + b × c−d e f
∧ ∧
26. The elements 32, 15, 20, 30, 12, 25, 16 are inserted a+b×c−d ef
∧ ∧
one by one in the given order into a maxHeap. The a+b×c−def
∧ ∧
resultant maxHeap is a+bc×−def
∧ ∧
abc×+−def
32 32 ∧ ∧
abc×+def −
Ans. (a)
30 25 25 30 28. Consider the following C program:
main ( )
{ int x, y, m, n;
15 12 20 16 12 15 20 16 scanf ("%d %d", &x, &y);
/ *Assume x > 0 and y > 0 */
(a) (b) m = x; n = y;
while (m! = n)
{ if (m > n)
32 32
m = m - n;
else
n = n - m;
30 25 25 30 }
printf("%d",n);
}
15 12 16 20 12 15 16 20 The program computes
(c) (d) (a) x + y using repeated subtraction.

(b) x mod y using repeated subtraction.
(GATE 2004: 2 Marks) (c) The greatest common divisor of x and y.
(d) The least common multiple of x and y.
Solution: Root is greater than both left and right
children.
Solution: The given program finds greatest
common divisor of x and y.
32 Ans. (c)
29. What does the following algorithm approximate?
(Assume m > 1, ε > 0)
30 25
x = m;
y = 1;
while (x - y > ε)
15 12 16 { x = (x + y)/2;
20
y = m/x;
}
Ans. (a) print(x);
27. Assume that the operators +, −, × are left asso- (a) log m (b) m2
ciative and ∧ is right associative. The order of pre- (c) m1/2 (d) m1/3
cedence (from highest to lowest) is ∧, × , +, −. (GATE 2004: 2 Marks)
Chapter 3.indd 148 4/9/2015 9:51:55 AM

Solution: The given program finds the square

Group 1 Group 2
root of program.
Ans. (c) III. Object-oriented C. Side-effect free,
declarative, expression
30. Consider the following C program segment: evaluation
struct CellNode{ IV. Imperative D. Declarative, clausal
structCellNode representation, theorem
*leftChild; proving
int element;
(a) I-B, II-C, III-D, IV-A
structCellNode
*rightChild; (b) I-D, II-C, III-B, IV-A
};
(c) I-C, II-D, III-A, IV-B
intDosomething (structCellNode *ptr) (d) I-C, II-D, III-B, IV-A
{ (GATE 2004: 2 Marks)
int value = 0;
Solution: Functional languages have no side
if (ptr ! = NULL)
effects and follow the declarative approach. Logic
{
if (ptr -> leftChild ! = NULL)
languages are declarative and used for theorem
value = 1 + DoSomething (ptr - proving. Object-oriented languages use abstract
>leftChild);
data types. Imperative languages are command
if (ptr -> rightChild ! = NULL)
based and procedural.
value = max(value,1 + Ans. (d)
DoSomething (ptr->rightChild));
} 32. What does the following C-statement declare?
return (value); int (*f) (int * );
}
(a) A function that takes an integer pointer as
The value returned by the function DoSomething argument and returns an integer.
when a pointer to the root of a non-empty tree is (b) A function that takes an integer as argument
passed as argument is and returns an integer pointer.
(c) A pointer to a function that takes an integer
(a) The number of leaf nodes in the tree. pointer as argument and returns an integer.
(b) The number of nodes in the tree. (d) A function that takes an integer pointer as
(c) The number of internal nodes in the tree. argument and returns a function pointer.
(d) The height of the tree.
(GATE 2005: 1 Mark)
Solution:
Solution: DoSomething( ) returns max(height of ⇒ int ( * f) (int * )
left child + 1, height of left child + 1). So given that ⇒ return type is int, (*f) means pointer to function
pointer to root of tree is passed to DoSomething( ), with argument as integer pointer.
it will return height of the tree. Note that this Ans. (c)
implementation follows the convention where
height of a single node is 0. 33. An abstract data type (ADT) is:
Ans. (d) (a) Same as an abstract class.
31. Choose the best matching between the program- (b) A data type that cannot be instantiated.
ming styles in Group 1 and their characteristics in (c) A data type for which only the operations
Group 2. defined on it can be used, but none else.
(d) All of the above.
Group 1 Group 2 (GATE 2005: 1 Mark)
I. Functional A. Command-based, Solution: ADT is user-defined data type that

procedural specifies the set of operations.
Ans. (c)
II. Logic B. Imperative, abstract
data types 34. A common property of logic programming lan-
guages and functional languages is:
Chapter 3.indd 149 4/9/2015 9:51:55 AM

(a) Both are procedural languages. (a) 8, 4, 0, 2, 14 (b) 8, 4, 0, 2, 0

(b) Both are based on l calculus. (c) 2, 0, 4, 8, 14 (d) 2, 0, 4, 8, 0
(c) Both are declarative. (GATE 2005: 2 Marks)
(d) Both use Horn-clauses.
(GATE 2005: 1 Mark) Solution:
Solution: Both are declarative because we declare Iteration k j foo(n, sum)
variable before use.
Ans. (c) 1, n = 2048, 8 204 foo(204,8)
Sum = 0
35. Which one of the following are essential features of
2, n = 204, 4 20 foo(20, 12)
an object-oriented programming language?
Sum = 8
I Abstraction and encapsulation 3, n= 20, 0 2 foo(2,12)
II Strictly typedness Sum = 12
III Type-safe property coupled with sub-type rule 4, n = 2, 2 0 foo(0,14), for
IV Polymorphism in the presence of inheritance Sum = 12 this call function
(a) I and II only (b) I and IV only will return.
(c) I, II and IV only (d) I, III and IV only
For recursive function calls, stack is used to store
(GATE 2005: 1 Mark)
values. So the values will be printed as 2 0 4 8. And
Solution: Common property of OOP is abstrac- value of sum is 0. Because sum is passed by value
tion and encapsulation and polymorphism in the not by reference. Changes are not reflected back to
presence of inheritance. the main program.
Ans. (b) Ans. (d)
36. A program P reads in 500 integers in the range 38. Consider the following C program:
[0, 100] representing the scores of 500 students. It double foo(double); /* Line 1 */
then prints the frequency of each score above 50. int main() {
What would be the best way for P to store the double da, db;
frequencies? // input da
(a) An array of 50 numbers db = foo(da);
}
(b) An array of 100 numbers
double foo(double a) {
(c) An array of 500 numbers
return a;
(d) A dynamically allocated array of 550 numbers
}
(GATE 2005: 1 Mark)
The above code compiled without any error or
Solution: We need to know the frequency of num- warning. If line 1 is deleted, the above code will
bers above 50. So, array of size 50 will be sufficient. show:
Ans. (a)
(a) No compile warning or error.
37. Consider the following C program: (b) Some compiler warnings not leading to unin-
tended results.
void foo (int n, int sum 0) {
(c) Some compiler warnings due to type-mismatch
int k = 0, j = 0;
if (n==0) return;
eventually leading to unintended results.
k = n % 10; j = n / 10; (d) Compiler errors.
sum = sum + k; (GATE 2005: 2 Marks)
foo (j, sum);
printf ("%d,", k); Solution: Compilation error will be thrown
} by deleting Line1. Function is called before it is
int main () { declared anywhere.
int a = 2048, sum = 0; Ans. (c)
foo (a, sum);
printf("%d\n", sum); 39. Postorder traversal of a given binary search tree T
} produces the following sequence of keys
What does the above program print? 10, 9, 23, 22, 27, 25, 15, 50, 95, 60, 40, 29
Chapter 3.indd 150 4/9/2015 9:51:56 AM

Which one of the following sequences of keys can 43. Consider line number 3 of the following C program:
be the result of an inorder traversal of the tree T ?
int min ( ) { /* Line 1 */
(a) 9, 10, 15, 22, 23, 25, 27, 29, 40, 50, 60, 95 int I, N; /* Line 2 */
(b) 9, 10, 15, 22, 40, 50, 60, 95, 23, 25, 27, 29 fro (I =0, I<N, I++); /* Line 3 */
(c) 29, 15, 9, 10, 25, 22, 23, 27, 40, 60, 50, 95 }
(d) 95, 50, 60, 40, 27, 23, 22, 25, 10, 9, 15, 29 Identify the compiler’s response about this line
(GATE 2005: 2 Marks) while creating the object-module:
Solution: Inorder traversal is always sorted in (a) No compilation error

increasing order. (b) Only a lexical error
Ans. (a) (c) Only syntactic errors
(d) Both lexical and syntactic errors
40. A priority queue is implemented as a max-heap. (GATE 2005: 2 Marks)
Initially, it has five elements. The level-order tra-
versal of the heap is given below: Solution: Misspellings of int keyword in line 2 and
there should be `for’ instead of `fro’ in line 3.
10, 8, 5, 3, 2 Ans. (c)
Two new elements `1’ and `7’ are inserted in the Linked Answer Questions 44 and 45: Let s and t
heap in that order. The level-order traversal of the be two vertices in a undirected graph G = (V, E)
heap after the insertion of the elements is: having distinct positive edge weights. Let [X, Y ]
(a) 10, 8, 7, 5, 3, 2, 1 (b) 10, 8, 7, 2, 3, 1, 5 be a partition of V such that s ∈ X and T ∈ Y.
(c) 10, 8, 7, 1, 2, 3, 5 (d) 10, 8, 7, 3, 2, 1, 5 Consider the edge e having the minimum weight
amongst all those edges that have one vertex in X
and one vertex in Y.
Solution: 44. The edge e must definitely belong to:
10 10 (a) the minimum weighted spanning tree of G
10 10
/ / /
/
/ (b) the weighted shortest path from s to t

/
8 5 5 8 (c) each path from s to t

8 8 5 7
(d) the weighted longest path from s to t
/ / / / / / /
/
3 2 3 2 1 3 2 1 7 3 2 1 5 (GATE 2005: 2 Marks)
Level order traversal is 10, 8, 7, 3, 2, 1, 5 Solution: Minimum weight edge is included in

Ans. (d) minimum spanning tree.
Ans. (a)
41. How many distinct binary search trees can be
created out of 4 distinct keys? 45. Let the weight of an edge e denote the congestion
on that edge. The congestion on a path is defined
(a) 5 (b) 14 (c) 24 (d) 42 to be the maximum of the congestions on the edges
(GATE 2005: 2 Marks) of the path. We wish to find the path from s to
t having minimum congestion. Which one of the
Solution: Number of distinct binary trees = following paths is always such a path of minimum
1/1+n (2nCn) = 14 congestion?
Ans. (b)
(a) a path from s to t in the minimum weighted
42. In a complete k-ary tree, every internal node has spanning tree
exactly k children. The number of leaves in such a (b) a weighted shortest path from s to t
tree with n internal nodes is: (c) an Euler walk from s to t
(a) nk (b) (n − 1)k + 1 (d) a Hamiltonian path from s to t
(c) n(k − 1) + 1 (d) n(k − 1) (GATE 2005: 2 Marks)
Solution: Path from s to t in the minimum
Solution: If there are n internal nodes, then weighted spanning tree is the path of minimum
number of leave nodes will be n(k−1) + 1 congestion.
Ans. (c) Ans. (a)
Chapter 3.indd 151 4/9/2015 9:51:56 AM

46. In a binary max-heap containing n numbers, the Solution: Best algorithm to find leader is pass the
smallest element can be found in time array from right side. With each traversal of node,
maximum will be updated. Hence leader can be
(a) O(n) (b) O(log n)
updated in O(n). So, option (b) is correct.
(c) O(log log n) (d) O(1)
Ans. (b)
(GATE 2006: 1 Mark)
50. An implementation of a queue Q, using two stacks
Solution: In max heap smallest element will be S1 and S2, is given below:
present in the leaf node. To get it, n nodes need to
void insert (Q, x) {
be traversed.
push (S1, x);
Ans. (b)
}
47. Let T be a depth first search tree in an undirected void delete (Q) {
graph G. Vertices u and v are leaves of this tree T. if (stack-empty(S2)) then
The degrees of both u and v in G are at least 2. if (stack-empty(S1)) then {
Which one of the following statements is true? print("Q is empty");
return;
(a) There must exist a vertex w adjacent to both }
u and v in G else while (!(stack-empty(S1))){
(b) There must exist a vertex w whose removal x=pop(S1);
disconnects u and v in G push(S2,x);
(c) There must exist a cycle in G containing u• }
and v x=pop(S2);
(d) There must exist a cycle in G containing u and }
all its neighbours in G. Let n insert and m (≤ n) delete operations be
(GATE 2006: 1 Mark) performed in an arbitrary order on an empty
queue Q. Let x and y be the number of push and
Solution: There must exist a vertex w adjacent to pop operations performed, respectively, in the pro-
both u and v in G. So, option (a) is true. cess. Which one of the following is true for all m
Ans. (a) and n?
48. A scheme for storing binary trees in an array X is (a) n + m ≤ x < 2n and 2m ≤ y ≤ n + m
as follows. Indexing of X starts at 1 instead of 0. (b) n + m ≤ x < 2n and 2m ≤ y ≤ 2n
The root is stored at X[1]. For a node stored at (c) 2m ≤ x < 2n and 2m ≤ y ≤ n + m
X[i], the left child, if any, is stored in X[2i] and the (d) 2m ≤ x ≤ 2n and 2m ≤ y ≤ 2n
right child, if any, in X[2i + 1]. To be able to store (GATE 2006: 2 Marks)
any binary tree on n vertices, the minimum size of
Solution:
X should be
Number of push operation = insert operation +
(a) log2 n (b) n delete operation
(c) 2n + 1 (d) 2n − 1 ⇒n + m
(GATE 2006: 1 Mark) ⇒n + m ≤ x
⇒ n + m ≤ x < 2n and 2m ≤ y ≤ n + m
Solution: In case of right skewed tree, the size of Ans. (a)
array will be 2n − 1.
Ans. (d) 51. Consider the following C function in which a[n] and
b[m] are two sorted integer arrays and c[n + m]
49. An element in an array X is called a leader if it is be another integer array.
greater than all elements to the right of it in X.
The best algorithm to find all leaders in an array void xyz(int a[], int b [], int c []){
int i,j,k;
(a) Solves it in linear time using a left to right pass i=j=k=0;
of the array while ((i<n) && (j<m))
(b) Solves it in linear time using a right to left pass if(a[i] < b[j])
of the array c[k++] = a[i++];
(c) Solves it using divide and conquer in time else c[k++] = b[j++];
Θ(n log n) }
(d) Solves it in time Θ(n2) Which of the following condition(s) hold(s) after
(GATE 2006: 1 Mark) the termination of the while loop?
Chapter 3.indd 152 4/9/2015 9:51:56 AM

(i) j < m, k = n + j − 1 and a[n − 1] < b[j] if i = n (c) S1 is true and S2 is false
(ii) i < n, k = m + i − 1 and b[m − 1] ≤ a[i] if j = m (d) S1 is true and S2 is true
(a) Only (i) (GATE 2006: 2 Marks)
(b) Only (ii)
(c) Either (i) or (ii) but not both Solution:
(d) Neither (i) nor (ii) S1: True, both work1 and work2 perform the same
task
S2: True, work2 will improve the performance
Solution: because it reduces calculations using temporary
x[i] ∧ ~y[i]) = only 1’s in x where corresponding variables to store results.
entry in y is 0 → x − y S1 and S2 are correct statements.
~x[i] ∧ y[i] = only 1s in y where corresponding Ans. (d)
entry in x is 0 → y − x. 54. Consider the following code written in a pass-by-
The operator “V” results in union of the above two sets. reference language such as FORTRAN and these
Ans. (c) statements about the code.
52. Given two arrays of numbers a1, ..., an and subroutine swap(ix,iy)
b1, ..., bn where each number is 0 or 1, the fastest it = ix
algorithm to find the largest span (i, j) such that L1 : ix = iy
ai + ai+1 + … + aj = bi + bi+1 + … + bj, or report L2 :iy = it
end
that there is not such span,
ia = 3
(a) Takes O(3n) and Ω(2n) time if hashing is permitted. ib = 8
(b) Takes O(n3) and Ω(n2.5) time in the key call swap (ia, 1b+5)
comparison model. print *, ia, ib
(c) Takes Θ(n) time and space. end
(d) Takes O(√n) time only if the sum of the 2n S1: The compiler will generate code to allocate a
elements is an even number. temporary nameless cell, initialise it to 13 and pass
(GATE 2006: 2 Marks) the address of the cell swap
S2: On execution, the code will generate a runtime
Solution: To find the largest span (i, j), average error on line L1
Θ(n) time and space is required. S3: On execution, the code will generate a runtime
Ans. (c) error on line L2
53. Consider these two functions and two statements S4: The program will print 13 and 8
S1 and S2 about them: S5: The program will print 13 and −2
Exactly the following set of statement(s) is correct:
int work1(int *a, int work2(int *a, int (a) S1 and S2 (b) S1 and S4
int i, int j) i, int j)
(c) S3 (d) S1 and S5
{ {
int x = a[i+2]; int t1 = i+2; (GATE 2006: 2 Marks)
a[j] = x+1; int t2 = a[t1];
return a[i+2] a[j] = t2+1; Solution: S1 and S2 are correct statements.
– 3; return t2 – 3; Ans. (a)
} }
55. Consider this C code to swap two integers and
S1: The transformation from work 1 to work 2 is these five statements: the code
valid, that is, for any program state and input argu- void swap (int *px, int *py) {
ments, work 2 will compute the same output and *px = *px - *py;
have the same effect on program state as work 1. *py = *px + *py;
S2: All the transformations applied to work 1 to get *px = *py - *px;
work 2 will always improve the performance (i.e }
reduce CPU time) of work 2 compared to work 1.
S1: Will generate a compilation error
(a) S1 is false and S2 is false S2: May generate a segmentation fault at runtime
(b) S1 is false and S2 is true depending on the arguments passed
Chapter 3.indd 153 4/9/2015 9:51:56 AM

3: Correctly implements the swap procedure for

S (c) 10, 9, 4, 5, 7, 6, 8, 2, 1, 3
all input pointers referring to integers stored in (d) 10, 8, 6, 9, 7, 2, 3, 4, 1, 5
memory locations accessible to the process
S4: Implements the swap procedure correctly for
some but not all valid input pointers Solution: After inserting 7, 2, 10, 4, the heap
S5: May add or subtract integers and pointers sequence will be 10, 7, 9, 8, 3, 1, 5, 2, 6, 4.
(a) S1 (b) S2 and S3 Ans. (a)
(c) S2 and S4 (d) S2 and S5
58. Consider the following segment of C code:
int j, n;
Solution: j = 1;
S2: May generate segmentation fault if value at while (j <=n)
pointers px or py is constant or px or py points to
j = j*2;
a memory location that is invalid.
S4: May not work for all inputs as arithmetic over- The number of comparisons made in the execution
flow can occur. of the loop for any n > 0 is:
Ans. (c)
(a) [log2n] + 1 (b) n
Common Data Questions 56 and 57: A 3-ary max- (c) [log2n] (d) [log2n] + 1
heap is like a binary max-heap, but instead of 2
(GATE 2007: 1 Mark)
children, nodes have 3 children. A 3-ary heap can
be represented by an array as follows: The root Solution: Check using values n = 3,4,5….
is stored in the first location, a[0], nodes in the Ans. (d)
next level, from left to right, are stored from a[1]
to a[3]. The nodes from the second level of the 59. The maximum number of binary trees that can be
tree from left to right are stored from a[4] location formed with three unlabeled nodes is:
onward. An item x can be inserted into a 3-ary
(a) 1 (b) 5
heap containing n items by placing x in the loca-
(c) 4 (d) 3
tion a[n] and pushing it up the tree to satisfy the
heap property. (GATE 2007: 1 Mark)
56. Which one of the following is a valid sequence of Solution: Maximum number of binary trees = 2n
elements in an array representing 3-ary max-heap? − n = 23 − 3 = 5
(a) 1, 3, 5, 6, 8, 9 (b) 9, 6, 3, 1, 8, 5 Ans. (b)
(c) 9, 3, 6, 8, 5, 1 (d) 9, 5, 6, 8, 3, 1 60. The height of a binary tree is the maximum number
(GATE 2006: 2 Marks) of edges in any root to leaf path. The maximum
number of nodes in a binary tree of height h is:
Solution: The 3-ary max heap can be formed by (GATE 2007: 1 Mark)
(a) 2h− 1 (b) 2h−1−1 (c) 2h+1− 1 (d) 2h+1.
9, 5, 6, 8, 3, 1.
9
Solution: Maximum number of node with height
h = 1 + 2 + 4 + ….. + 2h = 2h+1 − 1
56 8 Ans. (c)
3 1 61. The following postfix expression with single-digit
Ans. (d) operands is evaluated using a stack:
∧
823 /23*+51*−
57. Suppose the elements 7, 2, 10 and 4 are inserted, in
∧
that order, into the valid 3-ary max-heap found in Note that is the exponentiation operator. The
the above question Q.56. Which one of the follow- top two elements of the stack after the first * is
ing is the sequence of items in the array represent- evaluated are:
ing the resultant heap?
(a) 6, 1 (b) 5, 7
(a) 10, 7, 9, 8, 3, 1, 5, 2, 6, 4 (c) 3, 2 (d) 1, 5
(b) 10, 9, 8, 7, 6, 5, 4, 3, 2, 1 (GATE 2007: 2 Marks)
Chapter 3.indd 154 4/9/2015 9:51:57 AM

Solution: 64. Consider the following C function:

Expression TOS int f(int n)
{static int r = 0;
8 8
if (n <= 0) return 1;
2 8,2 if (n > 3)
3 8,2,3 {r = n;
∧ return f(n-2)+2;
8,8 }
/ 1 return f(n-1)+r;
2 1,2 }
3 1,2,3 What is the value of f (5)?

* 1,6 (a) 5 (b) 7 (c) 9 (d) 18
Top of stack contain 1,6. (GATE 2007: 2 Marks)
Ans. (a) Solution:

62. The inorder and preorder traversal of a binary tree f(5) r = 5 , 2 + f(3)
are d b e a f c g and a b d e c f g, respectively. The
postorder traversal of the binary tree is: f(3) 2 + 5 + f(2)
f(2) 7 + 5 + f(1)
(a) d e b f g c a (b) e d b g f c a
(c) e d b f g c a (d) d e f g b c a f(1) 12 + 5 + f(0)
(GATE 2007: 2 Marks) f(0) 17 + 1 = 18
Ans. (d)
Solution: Construct equivalent tree from inorder
and preorder traversal. 65. A complete n-ary tree is a tree in which each node
a has n children or no children. Let I be the number
/ / of internal nodes and L be the number of leaves in
b c a complete n-ary tree. If L = 41, and I = 10, what
/ / / / is the value of n?
d e f g
(a) 3 (b) 4 (c) 5 (d) 6
Postorder traversal is: d e b f g c a
Ans. (a) (GATE 2007: 2 Marks)
63. Consider a hash table of size seven, with starting Solution:

index zero, and a hash function (3x + 4) mod 7. I [n − 1] + 1 = L
Assuming the hash table is initially empty, which N = (L − 1)/I + 1 = 40/10 + 1= 5
of the following is the contents of the table when Ans. (c)
the sequence 1, 3, 8, 10 is inserted into the table
using closed hashing? Note that - denotes an empty 66. Consider the following C program segment where
location in the table. CellNode represents a node in a binary tree:
(a) 8, - , - , - , - , - , 10 (b) 1, 8, 10, - , - , - , 3 structCellNode {
(c) 1, - , - , - , - , - , 3 (d) 1, 10, 8, - , - , - , 3 structCellNode *leftChild;
(GATE 2007: 2 Marks) int element;
structCellNode *rightChild;
Solution: };
Hash function is: (3x + 4) mod 7 intGetValue (structCellNode *ptr) {
X = 1 7 mod 7 = 0 int value = 0;
X = 3 13 mod 7 = 6 if (ptr != NULL) {
X = 8 28 mod 7 = 0 since 0 occupied, it will be if((ptr->leftChild ==
on 1st position NULL) &&
X = 10 34 mod 7 = 6 since 6,0,1 is occupied it
(ptr->rightChild
will at 2nd position == NULL))
So 1 8 10 - - - 3 value = 1;
else
Ans. (b)
Chapter 3.indd 155 4/9/2015 9:51:57 AM

va
lue = value + GetValue 69. Which combination of the integer variables x, y
(ptr->leftChild) + GetValue and z makes the variable a get the value 4 in the
(ptr->rightChild); following expression?
}
return(value); a = (x > y) ? ((x > z) ? x :z):((y > z) ? y :z)
} (a) x = 3, y = 4, z = 2
The value returned by GetValue when a pointer (b) x = 6, y = 5, z = 3
to the root of a binary tree is passed as its argu- (c) x = 6, y = 3, z = 5
ment is: (d) x = 5, y = 4, z = 5
(GATE 2008: 1 Mark)
(a) The number of nodes in the tree
Solution: In option A, 3 > 4 is false so it moves
(b) The number of internal nodes in the tree
to else part. In else condition, 4 > 2 is true and 4
(c) The number of leaf nodes in the tree
(d) The height of the tree
will be printed.
(GATE 2007: 2 Marks) Ans. (a)
Solution: As we can see in else part of given 70. The breadth first search algorithm has been
program, left and right subtrees are traversed implemented using the queue data structure. One
recursively and value is incremented if they have possible order of visiting the nodes of the following
leaf nodes. graph is
Ans. (c)
67. Consider the process of inserting an element into M N O

a max-heap, where the max-heap is represented
by an array. Suppose we perform a binary search
on the path from the new leaf to the root to find
the position for the newly inserted element, the
number of comparisons performed is:
R Q P
(a) Θ (log2n) (b) Θ (log2log2n)
(c) Θ (n) (d) Θ (nlog2n)
(a) MNOPQR (b) NQMPOR
(GATE 2007: 2 Marks) (c) QMNPRO (d) QMNPOR
Solution: Inserting an element in binary tree (GATE 2008: 1 Mark)

takes Θ(1) and binary search takes Θ(logn).
Solution: In BFS, all the children of node will be
Ans. (a) traversed first.
68. An array of n numbers is given, where n is an even Ans. (a)
number. The maximum as well as the minimum of 71. You are given the postorder traversal, P, of a
these n numbers need to be determined. Which of binary search tree on the n elements 1, 2, …, n.
the following is TRUE about the number of com- You have to determine the unique binary search
parisons needed? tree that has P as its postorder traversal. What is
(a) At least 2n − c comparisons, for some constant, the time complexity of the most efficient algorithm
c are needed. for doing this?
(b) At most 1.5 − 2 comparisons are needed. (a) Θ (logn)
(c) At least nlog2n comparisons are needed. (b) Θ (n)
(d) None of the above. (c) Θ (n log n)
(GATE 2007: 2 Marks) (d) None of the above, as the tree cannot be
uniquely determined
Solution: Using divide and conquer technique
problem is divided into two parts of n/2. (GATE 2008: 2 Marks)
T(n) = 2 T(n/2) + 2 Solution:

Solution of this equation is 3/2 n − 2 comparisons Step 1: The last element in the postorder traversal
Ans. (b) P will act as the root element.
Chapter 3.indd 156 4/9/2015 9:51:58 AM

Step 2: Now in P the elements that are smaller main () {

than root will form the left subtree pr
int f ("Enter Text");
Step 3: The elements in P that are greater than print f ("\ n");
root will form the right subtree
reverse ();
Step 4: Binary search could be used to find the left print f ("\ n");
and right tree elements }
Ans. (c) (a) ? 1 is (getchar ( )! = ‘n’)
? 2 isgetchar (c);
72. We have a binary heap on n elements and wish to
insert n more elements (not necessarily one after (b) ? 1 is (c = getchar ( ) )! = ‘n’)
another) into this heap. The total time required for ? 2 isgetchar (c);
this is (c) ? 1 is (c ! = ‘n’)
(a) Θ (log n) (b) Θ (n) ? 2 isputchar (c);
(c) Θ (nlog n) (d) Θ (n2) (d) ? 1 is ((c = getchar ( ) )! = ‘n’)
? 2 isputchar (c);
Solution: Time require to re-heapify is O(n).
Ans. (c) Solution: ?1 checks the end of string, if not end of
string it calls reverse function.
73. What is printed by the following C program? ?2 prints the reverse string.
int f(int x, int *py, int **ppz) Ans. (d)
{
int y, z; 75. The following C function takes a single-linked
**ppz + = 1; z = *ppz; list of integers as a parameter and rearranges the
*py + = 2; y = *py; elements of the list. The function is called with
x + = 3; the list containing the integers 1,2,3,4,5,6,7 in the
return x + y + z: given order. What will be the contents of the list
} after the function completes execution?
void main()
{ struct node {
int c, *b, **a; int value;
c = 4; b = &c; a = &b; struct node * next;
print f("%d", f(c, b, a)); };
} Void rearrange (struct node * list) {
struct node * p, * q;
(a) 18 (b) 19 int temp;
(c) 21 (d) 22 if (!list || !list ® next) return;
(GATE 2008: 2 Marks) p = list; q = list ®> next;
while q {
Solution: mp = p ® value; p ®> value
te
** ppz = 4 +1 = 5 = q ®value;
z=5 q ®value = temp; p = q ® next
y=5+2=7 q = p? p ® next : 0;
x = 7 + 3 = 10 }
x + y + z = 10 + 7 + 5 = 22 }
Ans. (d) (a) 1,2,3,4,5,6,7 (b) 2,1,4,3,6,5,7
74. Choose the correct option to fill ? 1 and ? 2 so that (c) 1,3,2,5,4,7,6 (d) 2,3,4,5,6,7,1
the program below prints an input string in reverse
order. Assume that the input string is terminated
by a newline character. Solution: The program is simply swapping two
void reverse (void) { consecutive numbers.
int c; Here, input: 1,2,3,4,5,6,7
if (?1) reverse (); Output: 2,1,4,3,6,5,7
?2
} Ans. (b)
Chapter 3.indd 157 4/9/2015 9:51:58 AM

76. P and Q are two propositions. Which of the follow- 8*12*30

ing logical expressions are equivalent?
I. P ∨ ~ Q 5 10 15*20 40
II. ~ (~ P ∧ Q)
III. (P ∧ Q) ∨ (P ∧ ~ Q) ∨ (~ P ∧ ~ Q)
Insertion of 9th key
IV. (P ∧ Q) ∨ (P ∧ ~ Q) ∨ (~ P ∧ Q)
To maximize splits, let us insert a value in a node
that has key in access. Let us insert 17
(a) Only I and II (b) Only I, II and III
(c) Only I, II and IV (d) All of I, II III and IV 8*12*30
/ / /
(GATE 2008: 2 Marks) 5 10 15*17*20 40
Solution: IV represent P ∨ Q. Assuming values of
Insertion of 10th key (4th and 5th Splits)
P and Q can provide us the solution.
Ans. (b)
77. A B-tree of order 4 is built from scratch by 10 suc-
12
cessive insertions. What is the maximum number
/ /
of node splitting operations that may take place?
8 15*30
(GATE 2008: 2 Marks) / / / //
5 1 0 1 3 1 7 * 2 0 40
(a) 3 (b) 4 (c) 5 (d) 6
Ans. (a)
Solution: Insertion of 3 keys
10 20 30 following C functions:
Insertion of 4th key (1st split)
int f1 (int n)
30 {
/ / if (n == 0 | | n == 1)
10*20 40 return n;
else
Insertion of 5th key no split return (2* f1 (n - 1) + 3 * f1 (n - 2))
To maximize splits, let us insert a value in a node }
that has key in access. Let us insert 5 int f2 (int n)
{
30 int i:
/
int X[N], Y[N], Z[N];

5*10*20 40
X[0] = Y[0] = Z[0] = 0;
Insertion of 6th key (2nd Split) X[1] = 1; Y[1] = 2; Z[1] = 3;
To maximize splits, let us insert a value in a node for (i = 2; i <= n; i ++) {
that has key in access. Let us insert 6 X[i] = Y[i - 1] + Z[i - 2];
Y[i] = 2 * X[i];
8 *30 Z[i] = 3 * X[i];
}
5 10 * 20 40 return X[n];
}
Insertion of 7th key 78. The running time of f1(n) and f2(n) are
(a) Θ(n) and Θ(n) (b) Θ(2n) and Θ(n)
(c) Θ(n) and Θ(2n) (d) Θ(2n) and Θ(2n)
8*30 (GATE 2008: 2 Marks)
5 10 *15* 20 40 Solution: For f1 recurrence relation is,

T (n) = T(n −1) + T(n −2)
Insertion of 8th key (3rd Split) This type of relations has non-polynomial
To maximize splits, let us insert a value in a node complexity.
that has key in access. Let us insert 12 Ans. (b)
Chapter 3.indd 158 4/9/2015 9:52:00 AM

79. f1(8) and f2(8) return the values 82. Consider the program below:
(a) 1661 and 1640 (b) 59 and 59 # include <stdio.h>
(c) 1640 and 1640 (d) 1640 and 1661 int fun(int n, int * f_p) {
int t, f;
if (n <= 1) {
Solution: Implementing the code will give the * f_p =1;
output 1640, 1640. return 1;
}
Ans. (c) t = fun (n - 1, f_p);
Common Data for Questions 80 and 81: Consider f = t+ * f_p;
the following C program that attempts to locate an * f_p = t;
return f;
element x in an array Y[ ] using binary search. The
}
program is erroneous.
int main() {
1. f(int Y[10], int x) { int x = 15;
2. int u, j, k; printf (" % d\ n", fun(5, & x));
3. i = 0; j = 9; return 0;
4. do { }
5. k = (i + j) / 2; The value printed is
6. if(Y[k] < x) i = k;
else j = k; (a) 6 (b) 8 (c) 14 (d) 15
7. } while ((Y[k]! = x) && (GATE 2009: 1 Mark)
(i < j));
8. if(Y [k] == x) print f ("x is in Solution:
the array");
Iteration n *f_p t f
9. else print f ("x is not in the
array"); 5 15 ® 3 5 8 ® Ans
10. }
4 15 ® 2 3 5
80. On which of the following contents of Y and x does 3 15 ® 1 2 3
the program fail? 2 15 ® 1 1 2
(a) Y is [1 2 3 4 5 6 7 8 9 10] and x < 10. 1 15
(b) Y is [1 3 5 7 9 11 13 15 17 19] and x < 1.
(c) Y is [2 2 2 2 2 2 2 2 2 2] and x > 2. The function will return 8.
(d) Y is [2 4 6 8 10 12 14 16 18 20] and 2 < x < 20 Ans. (b)
and x is even. 83. What is the maximum height of any AVL tree with
(GATE 2008: 2 Marks) 7 nodes? Assume that the height of a tree with a
Solution: Binary search fails for Y [2 2 2 2 2 2 2 single node is 0.
2 2 2] and x > 2 (a) 2 (b) 3 (c) 4 (d) 5
Ans. (b)
81. The correction needed in the program to make it
work properly is Solution: The height of an AVL tree storing n
items is O(log(n)).
(a) Change line 6 to: if (Y[k] < x) i = k + 1; else Ans. (b)
j = k − 1;
Linked Answer Questions 84 and 85: Consider a
(b) Change line 6 to: if (Y[k] < x) i = k − 1; else
binary max-heap implemented using an array.
j = k + 1;
(c) Change line 6 to: if (Y[k] <= x) i = k; else 84. Which one of the following array represents a
j = k; binary max-heap?
(d) Change line 7 to:} while ((Y[k] == x) && (i < j));
(a) {25,12,16,13,10,8,14}
(GATE 2008: 2 Marks) (b) {25,14,13, 16,10,8,12}
Solution: Change line 6 according to option A to (c) {25,14,16,13,10,8,12}
correct the program. (d) {25,14,12,13,10,8,16}
Ans. (c) (GATE 2009: 2 Marks)
Chapter 3.indd 159 4/9/2015 9:52:00 AM

Solution: Solution: i and j are passed by address, any modi-

Conditions for max heap: fication by function will be reflected back.
•• Data at every node should be greater than its When p = q means address of q is changed.
children nodes *p = 2 will modify value of q to 2.
•• Left child is at 2i+1 and right at 2i+2 Ans. (d)
Ans. (c)
88. Which languages necessarily need heap allocation
85. What is the content of the array after two delete oper- in the runtime environment?
ations on the correct answer to the previous question?
(a) Those that support recursion
(a) {14, 13,12,10, 8} (b) {14,12,13, 8,10} (b) Those that use dynamic scoping
(c) {14,13, 8, 12,10} (d) {14,13,12, 8,10} (c) Those that allow dynamic data structures
(GATE 2009: 2 Marks) (d) Those that use global variables
Solution: (GATE 2010: 1 Mark)

Deletion can be done by following steps:
•• Replace the root with last node Solution: Runtime environment require dynamic
•• Re-heapify the tree memory allocation. Heap data structure is used for
dynamic memory allocation.
14
/ Ans. (a)
/
/
/
89. What is the value printed by the following C

13 12 program?
/ #include <stdio.h>
/
/ int f(int * a, int n)

/
{
8 10
if (n<= 0)return 0;
Ans. (d) el
se if(*a% 2= = 0)
86. In a binary tree with n nodes, every node has an return * a + f(a
odd number of descendants. Every node is consid- +1,n - 1);
ered to be its own descendant. What is the number el
se return * a -
of nodes in the tree that has exactly one child? f(a + 1, n - 1);
}
(a) 0 (b) 1 int main( )
(c) (n − 1)/2 (d) n − 1 {
(GATE 2010: 1 Mark) in
t a[ ] = {12, 7,
13, 4, 11, 6};
Solution: The given tree is called full binary printf("%d",
f(a,6));
tree which has 0 or 2 children. No node will have return 0;
exactly one child node. }
Ans. (a)
(a) −9 (b) 5 (c) 15 (d) 19
87. What does the following program print?
#include <stdio.h>
void f (int *p, int * g) { Solution:
p = q;
*p = 2; f(a,6) 12 + f(7,5)
}
int i = 0, j =1; f(7,5) 12 + 7 − f(13,4)
int main ( ){
f(13,4) 12 + 7 −(13 − f(4,3))
f(&i, & j);
print f ("%d%d / n", i, j ; f(4,3) 12 + 7 −(13 − ( 4 + f(11,2)))
return 0;
} f(11,2) 12 + 7 − (13 − (4 + 11 − f(6,1)))
(a) 2 2 (b) 2 1 (c) 0 1 (d) 0 2 f(6,1) 12 + 7 − (13 − (4 + 11 − 6 )) = 15

(GATE 2010: 1 Mark) Ans. (c)
Chapter 3.indd 160 4/9/2015 9:52:00 AM

90. The following C function takes a simply linked Solution:

list as input argument. It modifies the list by Let base address is 100
moving the last element to the front of the list So, 100 + E − A = 104
and returns the modified list. Some part of the
code is left blank. 100 101 102 103 104 105 106 107
typedefstruct node { G A T E 2 0 1 1
int value;
struct node *next;
} Node; 2011 will be printed.
Node *move_to_front(Node *head) { Ans. (c)
Node *p, *q;
if
((head = = NULL: || (head->next =
= NULL)) return head; 92. A max-heap is a heap where the value of each
q = NULL; p = head; parent is greater than or equal to the value
while (p-> next !=NULL) { of its children. Which of the following is a
q=P; max-heap?
p=p->next;
} (a)
10
_______________________________
return head;
}
8 6
Choose the correct alternative to replace the blank
line.
(a) q = NULL; p → next = head; head = p;
(b) q → next = NULL; head = p; p → next = head;
4 5 2
(c) head = p; p → next = q; q → next = NULL;
(d) q → next = NULL; p → next = head; head = p; 1
Solution: When the while loop ends, q contains

(b)
address of second last node and p contains address 10
of last node. So, we need to do the following things
after while loop.
(i) Set next of q as NULL (q->next = NULL). 8 6

(ii) Set next of p as head (p->next = head).
(iii) Make head as p ( head = p)
Step (ii) must be performed before step (iii). If we
4 5 1 2
change head first, then we lose track of head node
in the original linked list.
Ans. (d)
(c) 10
91. What does the following fragment of C program
print?
char c [] = “GATE2011”;
char *p = c; 5 6
printf(”%s”, p + p[3] — p[1]):
(a) GATE2011 (b) E2011

(c) 2011 (d) 011 4 8 2 1
(GATE 2011: 1 Mark)
Chapter 3.indd 161 4/9/2015 9:52:01 AM

(d) (c)
5
2 8
2
3
1 4 6 10
3 4
(GATE 2011: 1 Mark)
Solution: Heap is also complete binary tree. (d)
Ans. (b) 2
93. Consider two binary operators ” and `↓’ with the

precedence of operator ↓ being lower than that of 3
the operator . Operator is right associative while
operator ↓ is left associative. Which one of the
4
following represents the parse tree for expression
(7↓343↓2)?
7 3
(a) (GATE 2011: 2 Marks)
Solution:
Given expression is:
7
Precedence of ↑ is higher, so first the sub-expres-
sion (3↑ 4 ↑3) will be solved. In this 4 ↑ 3 will be
solved first because ↑ is right to left associative. So,
the expression will be solved as ((7 ↓(3↑(4 ↑3))).
3
↓ operator is left to right associative, so ((7 ↓(3↑(4
↑3))) will be solved first. Build the in order expres-
4 sion considering operator precedence.
((7 ↓(3↑(4 ↑3)))↓2
3 2 Ans. (b)
following recursive C function that takes two
(b) arguments:
unsigned int foo (unsigned int n,
2 unsigned int r) {
if (n>0) return ((n% r) + foo (n/r, r) );
else return 0;
7 }
94. What is the return value of the function foo when

3 it is called foo(345, 10)?
(a) 345 (b) 12 (c) 5 (d) 3
4 3 (GATE 2011: 2 Marks)
Chapter 3.indd 162 4/9/2015 9:52:03 AM

Solution: index variables, respectively. Initially, REAR =

FRONT = 0. The conditions to detect queue full
n = 345 and r = 10 345 % 10 + foo(34,10) = and queue empty are
5 + ….
(a) full: (REAR+1) mod n == FRONT
n = 34, r = 10 5 + 4 + foo(3,10) empty: REAR == FRONT
n = 3, r = 10 9 + 3 + foo (0,10) (b) full: (REAR+1) mod n == FRONT
n = 0, r = 10 12 + 0 = 12 empty: (FRONT+1) mod n == REAR
Ans. (b) (c) full: REAR == FRONT
empty: (REAR+1) mod n == FRONT
95. What is the return value of the function foo when
(d) full: (FRONT+1) mod n == REAR
it is called foo(513, 2)?
empty: REAR == FRONT
(a) 9 (b) 8 (c) 5 (d) 2 (GATE 2012: 2 Marks)
Solution:
Solution: Condition for queue full:
(Rear + 1) mod n = = front
n = 513 and r = 2 513 % 2 + foo(256, 2) = 1
Condition for queue free:
+ ….
Rear = = front
n = 256, r = 2 1 + 0+ foo(128, 2) Ans. (a)
n = 128, r = 2 1 + foo (64, 2) 98. The height of a tree is defined as the number of
n = 64, r = 2 1 + foo(32, 2) edges on the longest path in the tree. The func-
….. n = 2, r = 2 …. 1 + foo(1, 2)
tion shown in the pseudocode below is invoked as
height(root) to compute the height of a binary tree
n = 1, r = 2 1+1+0=2 rooted at the tree pointer root.
Ans. (d) int height (treeptr n)
{ if (n == NULL) return -1;
96. What will be the output of the following C pro- if (n ® left == NULL)
gram segment? if (n ® right == NULL) return 0;
char inChar = ‘A’ ;
switch(inChar) { else return B1 ; // Box 1
case ‘A’ : printf("Choice A\ n") ;
case ‘B’ : else { h1 = height (n ® left);
case ‘C’ : printf("Choice B") ; if(n ® right == NULL)
case ‘D’ : return (1+h1);
case ‘E’ : else { h2 = height (n ® right);

default :printf("No Choice") ; } return B2 ; // Box 2
}
(a) No choice }
(b) Choice A }
(c) Choice A choice B no choice
(d) Program gives no output as it is erroneous The appropriate expressions for the two boxes B1
(GATE 2012: 1 Mark) and B2 are
(a) B1: (1 + height (n → right))
Solution: No break statement is available, it will B2: (1 + max (h1, h2))
execute all the cases.
Choice A (b) B1: (height (n → right))
Choice B No Choice B2: (1 + max (h1,h2))
Ans. (c) (c) B1: height (n → right)
97. Suppose a circular queue of capacity (n − 1)
B2: max (h1, h2)
elements is implemented with an array of n elements. (d) B1: (1 + height (n → right))
Assume that the insertion and deletion operations B2: max (h1, h2)
are carried out using REAR and FRONT as array (GATE 2012: 2 Marks)
Chapter 3.indd 163 4/9/2015 9:52:03 AM

Solution: The above pseudocode is traversing tree 100. What output will be generated by the given code
to calculate the tree height. segment if:
Box 1: it should be replaced with the code that will
Line 1 is replaced by autoint a = 1;
traverse the right sub tree, so
Line 2 is replaced by registerint a = 2;
B1: (1 + height(n → right))
Box2: Now both left and right sub tree height is (a) (b)
calculated. 3 1 42
So the maximum out of both will be chosen.
4 1 61
B2: (1 + max(h1, h2)).
Ans. (a) 4 2 61
Common Data Questions 99 and 100: Consider the (c) (d)

following C code segment: 4 2 42
int a, b, c = 0; 6 2 42
void prtFun(void);
2 0 20
main( )
{ staticint a = 1; /*Line 1*/ (GATE 2012: 2 Marks)
prtFun( );
a += 1; Solution: Line 2 is replaced with register storage
prtFun( ); class, the values will not be preserved.
printf(" \n %d %d ", a, b); So it will print
} 4 2
4 2
voidprtFun(void) 2 0
{ staticint a = 2; /*Line 2*/ Ans. (d)
int b = 1;
101. Consider an undirected random graph of eight ver-
a += ++b;
printf(" \n %d %d ", a, b);
tices. The probability that there is an edge between
a pair of vertices is 1/2. What is the expected
} number of unordered cycles of length three?
99. What output will be generated by the given code (a) 1/8 (b) 1 (c) 7 (d) 8
segment? (GATE 2013: 1 Mark)
(a) (b) Solution: P(e) = (1/2)
31 42 Number of ways vertices can be chosen of cycle
three = 8C3 × (1/2)3 = 7
41 61
Ans. (c)
42 61
102. Which of the following statements is/are TRUE
(c) (d) for undirected graphs?
42 31
P : Number of odd degree vertices is even.
62 52
Q: Sum of degrees of all vertices is even.
20 52 (GATE 2013: 1 Mark)
(GATE 2012: 2 Marks) (a) P only (b) Q only
(c) Both P and Q (d) Neither P nor Q
Solution: A is static variable, its value will be
preserved between different function calls. Solution: Both statements P and Q are correct.
First call to prtFun ( ): a = 2 and b = 1
a = 2 +2 = 4 Ans. (c)
Print 4 2 103. What is the return value of f(p, p), if the value of
Second call to prtFun ( ): a = 4 and b = 1 p is initialised to 5 before the call? Note that the
a=4+2=6 first parameter is passed by reference whereas the
Print 6 2 second parameter is passed by value.
Main printf function prints : 2 0 int f (int&x, int c) {
Ans. (c) c = c - 1;
Chapter 3.indd 164 4/9/2015 9:52:03 AM

if(c==0) return 1; (a) Q(n) (b) Q(n + k)

x = x + 1; (c) Q(nk) (d) Q(n2)
return f(x,c) * x;
} (GATE 2013: 2 Marks)
(a) 3024 (b) 6561
Solution: Any combination of n Enqueue and
(c) 55440 (d) 161051
Dequeue operations will take Q(n)
Ans. (a)
Solution: For p = 5,
f(6, 4) * 6 = 9 * 8 * 7 * 6 = 3024 106. Suppose depth first search is executed on the
graph below starting at some unknown vertex.
Ans. (b) Assume that a recursive call to visit a vertex is
104. The preorder traversal sequence of a binary search made only after first checking that the vertex
tree is 30, 20, 10, 15, 25, 23, 39, 35, 42. Which one has not been visited earlier. Then the maximum
of the following is the postorder traversal sequence possible recursion depth (including the initial call)
of the same tree? is .
(a) 10, 20, 15, 23, 25, 35, 42, 39, 30

(b) 15, 10, 25, 23, 20, 42, 35, 39, 30
(c) 15, 20, 10, 23, 25, 42, 35, 39, 30
(d) 15, 10, 23, 25, 20, 35, 42, 39, 30
Solution: Postorder traversal: Left Right Root

15, 10, 23, 25, 20, 35, 42, 39, 30 is the correct order.
Ans. (d) (a) 17 (b) 18
105. Consider the following operation along with (c) 19 (d) 20
enqueue and dequeue operations on queues, where
k is a global parameter. Solution: By following the given path in graph,
maximum possible depth is 19.
MultiDequeue(Q){
m = k;
while (Q is not empty) and (m > 0) {
Dequeue(Q);
m = m – 1;
}
}
What is the worst case time complexity of a

sequence of n queue operations on an initially
empty queue? Ans. (c)
PRACTICE EXERCISES
Set 1
1. Predict the output of: (a) 5 (b) 6

(c) Compile time error (d) Runtime error
void main()
{ 2. Declaration of a pointer is given as (*ptr)[10]. Pick
int const*p=5; the correct answer.
printf("%d",++(*p));
(a) ptr is a pointer to an integer quantity.
}
(b) ptr is a pointer to a 10-element integer array.
Chapter 3.indd 165 4/9/2015 9:52:03 AM

(c) ptr is a 10-element array of pointers to integer 7. char func(intval)

quantities.
{
(d) ptr is a pointer to an 11-element integer array.
char ch=val;
3. struct a return ch;
{ }
int a; void main()
char b; {
int c; float a=115.76;
}; printf("%d",func(a));
Union b }
{
The output will be:
char a;
int b; (a) 0 (b) 115.76
int c; (c) 116 (d) Runtime error
};
8. staticint func(intval)
Which of these is correct?
{
(a) Size of a is always different from size of b.
staticint sum;
(b) Size of a is always same as size of b. sum+=val;
(c) a and b cannot be same because of non- return sum;
homogeneity. }
(d) It might result in runtime error.
void main()
4. char a[5]=“hello” which of the following is {
correct? int i, n=9;
(a) In array we cannot do the operation for(i=1;i<n--;i++)
(b) Size of a is too small func(i*2);
printf("%d", func(0));
(c) Size of a is too large
}
(d) Runtime error
5. void f(int i) What is the output?
{ (a) 18 (b) 20
int j; (c) 10 (d) 22
for(j=0; j<16; j++)
{ 9. A variable P is called a pointer if
if(i & (0x8000>>j))
printf("1"); (a) P contains the address of an element in DATA.
else (b) P points to the address of first element in
printf("0"); DATA.
} (c) P points to the address of last element in
} DATA.
(d) P contains the DATA and the address of
What is the purpose of this program?
DATA.
(a) Its output is hex representation of i.
(b) Its output is octal representation of i. 10. Size of (int) is
(c) Its output is decimal representation of i. (a) always 2 bytes.
(d) Its output is binary representation of i. (b) depends on compiler being used.
6 . What is the meaning of this pointer declaration: (c) depends on computer hardware being used.
char ptr(int(*a)[ ]); (d) always 32 bits.
(a) Invalid statement 11. Predict the output of:

(b) ptr is a function that accepts an argument main()
which is a pointer to an integer array and {
returns a character quantity. char s[]="man";
(c) ptr is a function that accepts an argument int i;
which is an array of pointers to integers and for(i=0;s[i];i++)
returns a character of integer representation. printf("\n%c%c%c%c",*(s+i),i[s]);
(d) Compile time error }
Chapter 3.indd 166 4/9/2015 9:52:04 AM

(a) Will execute successfully (a) ascending order. (b) descending order.
(b) Incorrect syntax (c) random order. (d) none of these.
(c) May execute on some compiler
(d) Will result in execution error 22. Which of the following is not an application of
binary search trees?
12. Global variables in different files are checked
(a) Finding a particular alphabet in a given string
(a) at load time. (b) Sorting of characters in a string
(b) at assembly time. (c) Faster searching as compared to linear search
(c) at linking time. (d) Minimising the length of a string
(d) during execution time.
23. Which of the following is known as two-way list?
13. Indirect addressing uses:
(a) Grounded header list
(a) Arrays (b) Pointers (b) Circular header list
(c) Structures (d) Array with pointers (c) Linked list with header and trailer
14. typedef char *ch; (d) None of the above
const ch P; 24. Which of the following operators cannot be over-
In this code, what is P? loaded in C++?
(a) P is a constant. (a) + (b) ?
(b) P is a character constant. (c) = = (d) : :
(c) P is a character type.
(d) P is a pointer of character type. 25. The correct way to round off a floating number x
to an integer value is
15. Which of the following will have efficient space
complexity and lesser time complexities? (a) y = (int) (x + 0.5)
(b) y = int (x − 0.5)
(a) Incomplete binary tree
(c) y = int (x + 0.5)
(b) Complete binary tree
(d) y = (int) ((int)x + 0.5)
(c) B+ tree
(d) Left-inclined binary tree 26. Given two sorted list of size `m’ and `n’, respec-
16. Binary tree can be represented as tively. The number of comparison needed in the
worst case by the merge sort algorithm will be
(a) Linked list and graph. (b) Array and tree.
(c) Array and linked list. (d) Stack and queue. (a) m * log n (b) min (m, n)
(c) avg (m, n) (d) m + n − 1
17. The worst case height of AVL tree with n nodes is
27. Consider a forest of t-trees having n vertices. The
(a) 1.44 log(n) (b) 1.44 log(n + 1)
number of edges will be:
(c) 1.44 log(n + 2) (d) 1.44 nlogn
(a) nlog t (b) n − t
18. Which of the following sorting algorithm has almost (c) n * t (d) (n * t)/2
the same worst case and best case complexity?
28. Consider a binary search tree, traversal
(a) Quick sort (b) Merge sort
will result in nodes arranged in ascending order?
(c) Heap sort (d) Shell sort
(a) Inorder (b) Polish
19. Which of the following is max-heap property if A is
(c) Reverse polish (d) Depth first search
an array and i is the index of the array?
(a) A[Parent(i) ≥ A[i]]
29. The balance factor of height balance trees is:
(b) A[A[i]>Parent(i)]
(c) A[Parent(i) ≤ A[i]] (d) A[A[i]<Parent(i)] (a) −1, 0, 1 (b) 1, 0, 1
(c) −1, 2, −1 (d) 1, 2, 1
20. Complete binary tree can be implement by making
use of 30. The term `push’ and `pop’ is related to the
(a) array. (b) dequeue. (a) array. (b) linked list.
(c) priority queue. (d) stack. (c) queue. (d) stack.
21. If a binary search tree traversed in inorder then the 31. A digraph when represented in an adjacency
number of nodes will be printed in matrix, the matrix is a
Chapter 3.indd 167 4/9/2015 9:52:04 AM

(a) weight matrix. (b) asymmetric matrix. (a) A (b) B

(c) symmetric matrix. (d) unit matrix. (c) H (d) G
32. The merge sort algorithm is based upon 40. If Inorder = EDFBGAJHCI, and Preorder =
(a) divide and conquer technique. ABDEFGCHJI, the root is
(b) 0/1 knapsack technique. (a) A (b) I
(c) optimisation technique. (c) J (d) B
(d) simulated annealing technique.
41. A graph having n vertices and k components can
33. A linked list is a kind of data structure.
have at most:
(a) Static
(a) n + k/2 edges
(b) Dynamic
(b) n*k/2 edges
(c) Static as well as dynamic
(c) (n − k)(n − k + 1)/2 edges
(d) Neither dynamic nor static
(d) (n − k)(n − k + 2)/2 edges
34. When the stack is empty, the value of the top of
stack(TOS) pointer is 42. A graph in which all nodes have the same degree is
called
(a) 1
(b) 0 (a) Euler graph. (b) Complete graph.
(c) −1 (c) Regular graph. (d) Colour graph.
(d) depends on the memory address
43. The time complexity of hashing is
35. In an empty circular queue, the front and rear are?
(a) O(n) (b) O(n log (n − 1))
(a) −1, −1 (b) 0, 0 (c) O(n3 − 1) (d) O(n2log (n − 1))
(c) 0, 1 (d) 1, 1
44. The number of fields required in a doubly linked is
36. Which of the following data structures is non- .
linear type?
(a) 7 (b) 2
(a) Queue (b) Lists (c) 3 (d) 1
(c) String (d) Graph
45. The runtime complexity of building max-heap is
37. A full binary tree has a height of 3, then the
number of nodes of such a tree is . (a) O(n) (b) O(log n)
(c) O(nn − 1) (d) O(n2log n/2)
38. For the following tree, give the postorder traversal
46. A binary tree is of level 7, the maximum number of
(a) 2465371 (b) 2456371 nodes can be
(c) 4257631 (d) 7654321
(a) 129 (b) 128
(c) 255 (d) 256
1
Set 2
2 3 1. What does the following code do?

Var a, b, c; integer
begin
6
4 5 a: = a+b;
b: =a-b;
c: =a-b
7 end
(a) Exchanges (a) and (b)
(b) Doubles (a) and stores in (b)
39. Find the root in the given preorder traversal (c) Doubles (b) and stores in (a)
ABDFGECH (d) Leaves (a) and (b) unchanged
Chapter 3.indd 168 4/9/2015 9:52:04 AM

2. FORTRAN implementation do not permit recur- (a) An array, each element of which is a pointer to
sion because a structure of type node.
(b) A structure of 2 fields, each field being a pointer
(a) They use static allocation for variables
to an array of 10 elements.
(b) They use dynamic allocation for variables
(c) A structure of 3 fields: an integer, a float and
(c) Stacks are not available on all machines
an array of 10 elements.
(d) It is not possible to implement recursion on all
(d) An array, each element of which is a structure
machines
of type node.
3. An unrestricted use of the `goto’ statement is 7. Aliasing in the context of programming language
harmful because refers to
(a) It makes it more difficult to verify programs. (a) multiple variables having the same memory
(b) It increases the running time of the programs. location.
(c) It increases the memory required for the (b) multiple variables having the same value.
programs. (c) multiple variables having the same identifier.
(d) It results in the compiler generating longer (d) multiple uses of the same variable.
machine code.
8. Consider the following three C functions:
4. What is the value of X printed by the following
[P1] int* g(void)
program? {
Program COMPUTE (input, output); int x = 10;
Var X: integer return (&x);
Procedure FIND (X: real) }
Begin X: sqrt(X); end [P2] int* px;
begin {
X:=2 int *px;
Find (X) *px= 10
Writeln(X) return px;
end }
[P3] int* g(void)
(a) 2 (b) 4
{
(c) Runtime error (d) None of the above
int *px;
px = (int*)malloc(size of (int));
5. Consider the following C function definition *px =10;
int Trial (int a, int b, int c) return px;
{ }
if ((a>=b)&& (c <b) return b;
el
se if (a >=b) Which of the above three functions is likely to
return Trial (a, c, b); cause problems with pointers?
else return Trial (b, a, c); (a) Only P3 (b) Only P1 and P3
} (c) Only P1 and P2 (d) P1, P2 and P3
The function trial: 9. The value of j at the end of the execution of the
(a) Finds the maximum of a, b and c. following C program
(b) Finds the minimum of a, b and c. int incr (int i)
(c) Finds the middle number of a, b and c. {
(d) None of these. staticint count = 0;
count = count + i
6. The following C declarations return (count);
}
struct node {
main (){
int i;
int i, j;
float j;
for (i=0; i<=4; i++)
};
j =incr(i)
struct node *s[10];
}
Define s to be is
Chapter 3.indd 169 4/9/2015 9:52:04 AM

(a) 10 (b) 4 14. Which of the following is essential for convert-

(c) 6 (d) 7 ing an infix expression to the post-fix form
efficiently?
10. What is printed by the print statements in the
program P1, assuming call-by-reference parameter (a) An operator stack
passing? (b) An operand stack
(c) An operand stack and an operator stack
Program P1()
(d) A parse tree
{
x = 10; 15. A priority queue Q is used to implement a stack
y = 3; that stores characters. PUSH(C) is implemented
func1 (y, x, x); INSERT(Q, C, K) where K is an appropriate inte-
print x;
ger key chosen by the implementation. POP is
print y;
implemented as DELETEMIN(Q). For a sequence
}
Func1(x, y, z)
of operations, the keys chosen are in
{ (a) non-increasing order.
y = y+4; (b) non-decreasing order.
z = x + y +z (c) strictly increasing order.
} (d) strictly decreasing order.
(a) 10, 3 (b) 27, 7
16. What value would the following function return for
(c) 31, 3 (d) None of these
the input x = 95?
11. Let A be a two-dimensional array declared as Function fun (x: integer): integer;
follows: Begin
A: array [1…10] [1..15] of integer if x > 100 then fun = x - 10
Assuming that each integer takes one memory loca- else fun: = : fun (fun(x+11))
End
tion, the array is stored in row-major order and the
first element of the array is stored at location 100. (a) 89 (b) 90
What is the address of the element A[i][j]? (c) 91 (d) 92
(a) 15i + j + 84 (b) 15j + i + 84 17. In a circular-linked list organisation, insertion of a
(c) 10i + j + 89 (d) 10j + i + 89 record involves modification of
12. Suppose you are given an array s[1 … n] and a (a) one pointer. (b) two pointer.
procedure reverse (s, i, j) which reverses the order (c) multiple pointer. (d) no pointer.
of elements in between positions i and j (both inclu-
sive). What does the following sequence do, where 18. Linked lists are not suitable data structures of
1≤k≤n which one of the following problems?
reverse (s, 1, k); (a) Insertion sort
reverse (s, k + 1, n); (b) Binary search
(c) Radix sort
reverse (s, 1, n)
(d) Polynomial manipulation
(a) Rotates s left by k position
(b) Leaves s unchanged 19. In the worst case, the number of comparisons
(c) Reverses all elements of s needed to search singly linked list of length n for a
(d) None of the above given element is
(a) log2 n (b) n/2
13. Which of the following permutations can be
(c) log2 n − 1 (d) n
obtained in the output (in the same order) using a
stack assuming that the input is the sequence 1, 2, 20. A binary tree T has n leaf nodes. The number of
3, 4, 5 in that order? nodes of degree 2 in T is
(a) 3, 4, 5, 1, 2 (b) 3, 4, 5, 2, 1 (a) log2 n (b) n − 1
(c) 1, 5, 2, 3, 4 (d) 5, 4, 3, 1, 2 (c) n (d) 2n
Chapter 3.indd 170 4/9/2015 9:52:04 AM

21. In the balanced binary tree in the figure given of a complete n-ary tree, the number of leaves in it
below, how many nodes will become unbalanced is given by
when a node is inserted as a child of the node `g’?
(a) x(n − 1) + 1 (b) xn − 1
(c) xn +1 (d) x(n + 1)
a
27. Let LASTPOST, LASTIN and LASTPRE denote
the latest vertex visited in a postorder, inorder
b e and preorder traversal, respectively, of a complete
binary tree. Which of the following is always true?
c d (a) LASTIN = LASTPOST
f (b) LASTIN = LASTPRE
(c) LASTPRE = LASTPOST
g (d) None of these
(a) 1 (c) 7 28. The number of leaf nodes in a rooted tree of n

(a) 1 (b) 3 nodes with each node having 0 to 4 children is
(b) 3
(c) 7 (d)(d)
8 8
(a) n/2 (b) (2n − 1)/4
22. A binary search tree is generated by inserting in (c) (n + 1)/4 (d) (2n + 1)/4
order of the following integers: 50, 15, 62, 5, 20,
58, 91, 3, 8, 37, 60, 24. The number of nodes in the 29. Consider the following program:
left sub-tree and right sub-tree of the root, respec- #define funct(x) x*x+x
tively, are int main()
(a) (4, 7) (b) (7, 4) {
(c) (8, 3) (d) (3, 8) int x;
x=23+funct(4)*funct(3);
23. A binary search tree is used to locate the number printf(“%d”,x);
43. Which one of the following sequences is not return 0;
possible? }
(a) 61 52 14 17 40 43 What will be the output of the above program?

(b) 2 3 50 40 60 43 (a) 126 (b) 196
(c) 10 65 31 48 37 43 (c) 149 (d) 132
(d) 81 61 52 14 41 43
Output of the following program will be
24. A binary search tree contains the values 1, 2, 3, int main
4, 5, 6, 7, 8. The tree is traversed in preorder and {
the values are printed out. Which of the following int i, a[6]= 000, 001, 010, 011, 100,
sequences is a valid output? 101, 110, 111;
(a) 53124786 (b) 53126487 for i=0;i<6;i++
(c) 53241678 (d) 53124768 printf"%d", a[i];
return0;
25. Which of the following statements is false? }
(a) A tree with n nodes has (n − 1) edges. (a) 0, 1, 2, 3, 4, 5
(b) A labelled rooted binary tree can be uniquely (b) 0, 1, 10, 11, 100, 101
constructed given its postorder and preorder (c) 0, 1, 8, 9, 100, 101
traversal results. (d) None of these
(c) A complete binary tree with n internal nodes
has (n + 1) leaves. 30. The minimum and maximum number of keys
(d) The maximum number of nodes in a binary tree in the internal nodes of B tree with order 4 is,
of height h is (2h+1 − 1) respectively
26. A complete n-ary tree is one in which every node (a) 1, 3 (b) 2, 4
has 0 or n sons. If x is the number of internal nodes (c) 1, 4 (d) 2, 3
Chapter 3.indd 171 4/9/2015 9:52:05 AM

31. Consider the following code segment. What is the final value of x in both call by value
void foo(int x, int y) and call by reference, respectively?
{ (a) 5 and 16 (b) 5 and 12
x+=y;
(c) 5 and 20 (d) 12 and 20
y+=x;
} 32. What is the maximum height of any AVL tree with
main() 7 nodes? Assume that the height of a tree with a
{ single node is 0.
int x=5.5;
foo(x,x);
}
Set 1
1. (c) 11. (a) 21. (a) 31. (b) 41. (d)

2. (b) 12. (c) 22. (d) 32. (a) 42. (c)
3. (a) 13. (b) 23. (d) 33. (b) 43. (a)
4. (b) 14. (a) 24. (d) 34. (c) 44. (c)
5. (a) 15. (b) 25. (a) 35. (a) 45. (b)
6. (b) 16. (c) 26. (d) 36. (d) 46. (b)
7. (a) 17. (c) 27. (b) 37. (3)
8. (b) 18. (c) 28. (a) 38. (c)
9. (a) 19. (a) 29. (a) 39. (a)
10. (b) 20. (a) 30. (d) 40. (a)
Set 2
1. (a) The given program is for swapping values of 6. (a) S is an array, each element of which is a pointer
two variables. to a structure of type node.
2. (a) In recursion, stack is used to store the latest 7. (a) Aliasing describes a situation in which a data
value of variables, means value of variables change location in memory can be accessed through differ-
with the recursive call. But FORTRAN uses ent symbolic names in the program.
static allocation for variables which does not allow 8. (c) In P1, there is no parameter passing and type
changes in every different function call. casting so it is wrong.
3. (a) A goto statement does not increase the running In P2, pointer is defined as pointer function which
time or memory requirement of the program. It is wrong.
also does not contribute to longer machine code. 9. (a) Count is declared as static variable in incr( ).
But the use of goto can result in unstructured So, statement static int count = 0 will assign count
code with multiple entry and exit points which can to 0 only in the first incr call. Remaining calls to
cause an issue for program verification. incr function will use old values of count.
4. (b) Global variable value not changed because of at first incr(0) call count will become 0
static scoping. after incr(1) call count will become 1
after incr(2) call count will become 3
5. (d) Just take a = 21, b = 8 and c = 14 and verify after incr(3) call count will become 6
the options.
after incr(4) call count will become 10
Chapter 3.indd 172 4/9/2015 9:52:05 AM

10. (c) Initial value of x and y will be 10, 3 and address fun (101)
are 1000, 2000, respectively. Then after first state- fun = 101−10 = 91
ment execution in the function, value at address
1000 will be {(10 as old value) + (4) = 14} because Return value of fun is 91. So, option (c) is correct.
y at the called function refers to address of x. After
16. (b) If a record is inserted in the circular-linked list
second statement execution in the function, value
at address 1000 will be {3 + 14 + 14 = 31} because
then two pointers have to be modified. Let a record
k be inserted between the node m and n, then
`x’ at called function refers to y in the calling func-
tion and `z’ at the called function refers to address (a) A
ddress field of node m contains address of
of x. So final values will be node x.
x(address = 1000) = 31 (b) Address field of node x contains address of
node n.
y(address = 2000) = 3
11. (a) To find location of an element in a two- 17. (b) With linked list binary search will take O(n)
dimensional array by row-major order is: time to sort the element list. Thus, binary search
performs worst with linked list.
A[i][j] = Base address + (i − 1)*r + (j − 1), where
r = no. of rows. 18. (d) On linked list we can perform only linear
Given base address = 100 and number of rows = 15 search. The number of comparisons required using
= 100 + (i − 1) × 15 + (j − 1) linear search is O(n).
= 100 + 15i − 15 + j − 1
= 100 + 15i + j − 16 19. (b) In a binary tree, two nodes are connected by a
= 84 + 5i + j single node to having degree 2. If a tree has n leaf
nodes, then the number of nodes with degree 2 will
11. (a) Let an array S[1 … 8] always be less than n. So option (b) is the strong
answer.
1 2 3 4 5 6 7 8
20. (b)
S = 7, k = 3 1
a
reverse (S, 1, 3) we get [3, 2, 1, 4, 5, 6, 7, 8]
reverse (S, 4, 8) we get [3, 2, 1, 8, 7, 6, 5, 4] 1
b e 1
reverse (S, 1, 8) we get [4, 5, 6, 7, 8, 1, 2, 3]
after performing above three reverse, we can see it d
1 c
is equivalent to the option (a). f
0
12. (b) Take a stack and perform push and pop opera- 0
0 g
tion to verify the given options.
13. (a) During the conversion of infix to postfix expres-

sion, operands do not change their position. So, After a node h is added as a child of g
only operator stack is needed.
2 a
14. (d) Stack works on concept of first-in last-out.
But according to the question, it should work like
queue means first-in first-out. For that, we give the Unbalanced 2 b e 1
higher priority to the first element. So, all elements nodes
should form strictly decreasing order. 2 d
c
f
15. (c) fun (98) 0
0
fun fun(109) 1 g
fun (99)
fun (fun (200)) h
0
fun (100)
fun (fun(111)) Nodes a, b and c become unbalanced.
Chapter 3.indd 173 4/9/2015 9:52:06 AM

21. (b) The binary search tree is as follows: In postorder traversal, root node is traversed in the
end. So we can remove option (a) and (c).
50 Let us take an example for option (b).
15 62
4
5 20 58 91
5 6
3 8
37 60
7 8 9
24
Inorder: 7 5 8 4 9 6
In the above binary tree, number of nodes in the
left sub-tree of root is 7 and in the right sub-tree Preorder: 4 5 7 8 6 9
of the root is 4. When a tree is a complete binary tree with full
at last level, then only last-visited node will
22. (b) Not possible always be same for inorder traversal and preorder
(b) 2 3 50 40 0 43 traversal.
23. (d) Inorder sequence: 53124768 Ans. (d)
5 27. (d) Total number of nodes in n-ary tree = n(internal

nodes) + 1
3 Given n = 4i + 1
7
Total nodes = internal nodes (i) + leaf nodes (l)
n=i+l
i=n−l
1 4
6 8
n = 4(n − l) + 1
2 n = 4n − 4l + 1
4l = (2n + 1)
24. (b) A labelled rooted binary tree cannot be con- l = (2n + 1)/4
structed uniquely when inorder traversal is given
along with postorder or preorder traversal. 28. (c) According to program,
F(n) = n + 2F(n − 1) for n ≥ 2
25. (a) Total number of nodes in n-ary tree = ni + 1,
= 0 for n = 1
where i is internal nodes 1
F(4) = 4 + 2(3 + 2(2 + 2(0)))) = 18
Total nodes = internal nodes (i) + leaf nodes (l ) F(3) = 3 + 2(2 + 2(0)))) = 7
2 Ans. (c)
From 1 and 2
29. (c) 0, 1, 8, 9, 100, 101, 110 is an octal number
i + l = ni + 1
format.
l = (n − 1)I + 1
Replace I with x, because in question internal node 30. (b) B tree of order k, every internal node must
is given as x. have k/2 to k children and number of keys will be
So number of leaf nodes in n-ary tree with x inter- one less than the children.
nal nodes is (n − 1)x + 1.
31. (c) In call by value, there is no change in the value
26. (d) In the question, it is given that the tree is a com- of x because function foo( ) operates on copy of x.
plete binary tree. For a complete binary tree, the In call by reference, first value of x changes to 5 + 5
last visited node will always be the same for inorder and then it changes to 10 + 10 = 20.
traversal and preorder traversal because in both the
traversal right child is traversed in the last. 32. (3) Draw AVL tree to find the height.
Chapter 3.indd 174 4/9/2015 9:52:07 AM

UNIT IV: ALGORITHMS
9
8
Number of questions
7
6
Marks 1
5
Marks 2
3
2
1
0
2015 2014 2013 2012 2011 2010

Year Concept
2015 Algorithms, Unsorted array, BFS, Spanning tree, Recurrence relation, Job sequence, Planar graph, Hash
table, Complexity, Asymptotic notations, Polynomial time reducible, SAT Problem, Connected graphs,
Algorithm design paradigms, Graphs
2014 Arithmetic expressions, Rooted tree, Graph, Depth-first search, Sorting algorithms and their complexity,
Balanced binary, NP-complete problem, Hashing
2013 Sorting algorithms, Time complexity, NP problems, Bellman—Ford algorithm, Graphs, Heap sort, queue
2012 Time complexity, Recurrence relation, Balanced binary search tree, NP problem, Dijkstra’s algorithm,
Merge sort
2011 Arrays, Binary heap, Dynamic programming, Time complexity, Hashing, Minimum spanning tree (MST)
2010 Undirected graphs, Spanning tree, Hashing
Chapter 4.indd 175 4/9/2015 9:54:02 AM

Chapter 4.indd 176 4/9/2015 9:54:02 AM
CHAPTER 4
ALGORITHMS
Syllabus: Algorithms: Analysis, Asymptotic notation, Notions of space and time complexity, Worst- and average-case
analysis; Design: Greedy approach, Dynamic programming, Divide-and-conquer; Tree and graph traversals, Connected
components, Spanning trees, Shortest paths; Hashing, Sorting, Searching. Asymptotic analysis (best, worst, average
cases) of time and space, Upper and lower bounds, Basic concepts of complexity classes — P, NP, NP-hard, NP-complete.
4.1 INTRODUCTION 4.2.1 Properties of Algorithm
There are certain properties algorithms have:

In computer science, an algorithm provides a set of step- 1. An algorithm has zero or more inputs.
by-step instructions for solving various problems. The 2. It should produce at least one output.
set of instructions should be unambiguous, finite and 3. It should terminate within the finite time.
provide a clear termination point. An algorithm accepts 4. It should be deterministic (ambiguous).
some input values and produces outputs. The algorithm 5. It is generic to all programming languages.
can be built in simple English language or in program-
ming language. This subject helps in solving and analys- 4.2.2 Steps to Solve a Problem
ing complex problems through logical thinking.
1. Define the problem, which means what is input
and what should be the output.
4.2 ALGORITHM 2. Identify the constraints which are to be satisfied.
3. Design the algorithm for a given problem.
4. Draw the flowchart.
An algorithm is a well-defined procedure which takes a 5. Verify the process.
set of values as an input and produces a set of values as 6. Implement the coding.
the output. 7. Perform time and space analysis.
Chapter 4.indd 177 4/9/2015 9:54:03 AM

178 Chapter 4: Algorithms
4.2.3 Algorithm Analysis Example 4.1

If a problem has more than one solution, then the tech- Let us consider an example of a search problem
nique used to decide which solution is best is known as (searching a given item) in a sorted array. One way to
algorithm analysis. search is linear search (the order of growth is linear)
and the other way is binary search. Suppose we have
4.2.3.1 A Posteriori and A Priori Analysis two machines with different processing speeds. We run
linear search on the fast machine and binary search on
Analysis is done on two parameters: time complexity and slow machine. For small input size, the linear search
space complexity. Time complexity means time taken will perform better. But as we increase the size of
by an algorithm to solve a problem. Space complexity input, binary search will perform better. The reason
means memory space required by an algorithm to solve behind this is,the order of growth of binary search is
a problem. There are two types of analysis: logarithmic and that of linear search is linear. So, by
1. A Posteriori Analysis: It is a software- and using this technique the issues of machine dependency
hardware-dependent analysis. It keeps on chang- are resolved.
ing from one system to other system. This type of
analysis tells the exact time and space complexity, 4.2.4 Asymptotic Notation
meaning how much an algorithm takes time and
space for solving a problem on the given system. Asymptotic notation is based on two assumptions, which
2. A Priori Analysis: It is a software- and hard- hold in most of the cases and have their importance.
ware-independent analysis. It is constant for all the It is very much important to understand the assump-
systems. This type of analysis tells the approximate tions and limitations of asymptotic notations:
time and space complexity. It is the determination 1. Input size: It is interesting to know how the run-
of the order of magnitude of a statement, meaning ning time of an algorithm grows for a large input
how many times a statement is executing. size n.
2. Constant: The running time of an algorithm also
4.2.3.2 Asymptotic Analysis depends on various constants. But for a large input
n, the constant factors would be ignored.
Suppose there are given two algorithms for a task, how
do we find out which one is better? For representing best-case, average-case, and worst-case
This can be done by actually implementing both the complexity, there are three different notations as given
algorithms and running them on the same computer with in the following sub-sections.
different input values. Then the comparison is made in
terms of time. Algorithm that takes less time is considered 4.2.4.1 Theta Notation (Q)
better. But there are so many problems with this approach.
1. It might be possible that for certain inputs first Theta notation shows the tightest upper bound and the
algorithms performs better than second and for tightest lower bound to the given function (Fig. 4.1). It is
other set of inputs second algorithm performs used to define the average-case running time. For a given
better. So,the decision of choosing best among function f(n), we denote Q(g(n)) as Q(g(n)) = {f(n):
them becomes difficult. there exists positive constants c1 > 0, c2 > 0 and n0 > 0
2. With the change in operating machine, performance such that 0 < c1(g(n)) ≤ f (n) ≤ c2 (g(n)) ∀ n ≥ n0}
of algorithm varies. On one machine first algorithm
can work better and on another machine, the second
works better. This is another problem associated c1g(n)
with this approach.
f(n)
Above issues can be handled using asymptotic analy-
f(n) c2g(n)
sis techniques for analysing algorithms. In asymptotic
analysis, analysis of algorithm in done in terms of input g(n)
size. Actual running time is not computed, variation of
time with input size is calculated.
Now, understand that how the issues with above n
n0
Figure 4.1 | Theta notation (Q)
technique are resolved by asymptotic analysis technique
using the below example:
Chapter 4.indd 178 4/9/2015 9:54:04 AM

4.2 ALGORITHM 179
Problem 4.1: Consider the following functions: Problem 4.2: Consider the following functions:
f (n) = n f (n) = n
n g(n) = n
g(n) =
8
For f(n) = O(g(n)), find the values of c and n0.
For f(n) = Q(g(n)), find the values of c1, c2 and n0.
Solution: From the definition of Big-O notation, we
Solution: From the definition of theta (Q) notation, have
f (n) ≤ c(g(n)) ∀n ≥ n0
we have
0 < c1(g(n)) ≤ f (n) ≤ c2 (g(n)) ∀n ≥ n0 By putting c = 1 and n = 1, we get

By putting c1 = 1, 2, 3 to satisfy the above condition, n ≤ c ⋅ n ∀ n ≥ n0
we get
c1(g(n)) ≤ f (n) The above condition is satisfied. So, c = 1 and n0 = 1.
n
1   ≤ n
 8 
4.2.4.3 Omega Notation (W)
c1 = 1
Omega notation shows the lower bound to the given
By putting c2 = 1, 2, 3 to satisfy the above condition, function (Fig. 4.3). It is used to define the best-case run-
we get ning time. For a given function f(n), we denote W(g(n))
f (n) ≤ c2 (g(n)) as W(g(n)) = {f(n): there exist positive constants c > 0
and n0 > 0 such that 0 < c(g(n)) ≤ f (n) ∀ n ≥ n0 }.
n
8   ≤ n
 8 
c2 = 8 f(n)
cg(n)
By putting n0 = 1, 2, 3, …, to satisfy the above condi-
tion, we get n0 = 1. f(n)
g(n)
4.2.4.2 Big-O Notation (O)
Big-O notation shows the upper bound to the given func- n0

tion (Fig. 4.2). It is used to define the worst-case running Figure 4.3 | Omega notation (W).
time. For a given function f(n), we denote O(g(n)) as
O(g(n)) = {f(n): there exist positive constants c > 0 and
n0 > 0 such that 0 < f (n) ≤ c(g(n))∀n ≥ n0} Problem 4.3: Consider the following functions:
f (n) = n2
c(g(n)
g(n) = n
f(n)
For f(n) = W(g(n)), find values of c and n0.
Solution: From the definition of Omega notation, we
f(n) have
g(n) c(g(n)) ≤ f (n) ∀ n ≥ n0
By putting c = 1 and n = 1, we get
n0 c ⋅ n ≤ n2 ∀ n ≥ n0
Figure 4.2 | Big-O notation (O). The above condition is satisfied. So, c = 1 and n0 = 1.
Chapter 4.indd 179 4/9/2015 9:54:09 AM

O(nn)
O(an)
O(nk)
O(n2)
O(n)
Time
O(√n)
Olog(n)
O(log(log n)
O(1)
Complexities
Figure 4.4 | Time complexities in increasing order.
4.2.4.4 Small-o Notation (o) 4.2.4.7 Types of Complexities

1. Constant time complexity: O(1)
Small-o represents an upper bound which is not a tight
2. Logarithmic time complexity: O(log(log n)),
upper bound. It is also used to define the worst-case run-
O( log n ) and O(log n)
ning time. For a given function f(n), we denote o(g(n))
as o(g(n)) = {f(n): there exist positive constants c > 0
3. Linear time complexity: O( n ) and O(n)
4. Polynomial time complexity: O(nk ), where k is a
and n0 > 0 such that 0 < f (n) < c(g(n)) ∀ n ≥ n0}.
constant and is >1
5. Exponential time complexity: O(an ), where a > 1
4.2.4.5 Small Omega Notation (w)
The complexities in increasing order are represented in
Small omega (w) represents a lower bound which is not a Fig. 4.4.
tight lower bound. It is used to define the best-case run-
ning time. For a given function f(n), we denote w(g(n)) 4.2.5 Recurrence Relations
as w(g(n)) = {f(n): there exist positive constants c > 0 Recurrence relations are recursive definitions of
and n0 > 0 such that 0 < c(g(n)) < f (n) ∀ n ≥ n0 }. mathematical functions or sequences. For example,
the recurrence relation defines the famous Fibonacci
4.2.4.6 Properties of Asymptotic Notations sequence 1, 1, 2, 3, 5, 8, 13, …:
f (n) = f (n - 1) + f (n - 2)
1. Transitivity: All asymptotic notations show
transitivity. f(1) = 1
If f(n) = Q(g(n)) and g(n) = Q(h(n)), then f(n) = f(0) = 1

Q(h(n)). Given a function defined by a recurrence relation, we
If f(n) = O(g(n)) and g(n) = O(h(n)), then f(n) = want to find a `closed form’ of the function. In other
O(h(n)). words, we would like to eliminate recursion from the
If f(n) = W(g(n)) and g(n) = W(h(n)), then f(n) = function definition. There are several techniques for solv-
W(h(n)). ing recurrence relations, and these are discussed in the
If f(n) = o(g(n)) and g(n) = o(h(n)), then f(n) = following sub-sections.
o(h(n)).
If f(n) = w(g(n)) and g(n) = w(h(n)), then f(n) = 4.2.5.1 Iteration Method
w(h(n)).
This method is accurate but involves a lot of algebraic
2. Reflexivity: Only Theta, Big-O, and Omega nota- expressions which are difficult to keep track of. It can get
tions show reflexivity. Small-o and small-omega do very challenging for further complicated recurrence
not show reflexive property. relations.
f (n) = Q(f (n)) f (n) = O(f (n)) f (n) = W(f (n))
Problem 4.4: We have T (n) = 1 if n = 1
3. Symmetric: Only Theta notation is symmetric.
n
f(n) = Q(g(n)) if and only if g(n) = Q(f(n)). T (n) = 4T   + n if n > 1
 2 
4. Transpose symmetry:
f(n) = O(g(n)) if and only if g(n) = W(f(n)) n  n n
T (n) = 4T   + n = 4 4T   +  + n
f(n) = o(g(n)) if and only if g(n) = w(f(n))  2    4  2 
Chapter 4.indd 180 4/9/2015 9:54:13 AM

4.2 ALGORITHM 181
How deep does the tree go?

  n n n
= 4 4 4T   +  +  + n We stop at the leaf, and we know we are at a leaf when
   8  4  2  we have a problem of size 1:
n  n 2
= 64T   + 4n + 2n + n 1 =  i  ⇒ n2 = 22i ⇒ n = 2i ⇒ i = log n
 8   2 
n
= n + 2n + 4n + 64T  
The amount of work done is given as
 8  log n
ni
n
= n + 2n + 4n + + 2 j n + + 4i T  i 
∑ = Qn2
 2  i=0 2
This is geometrically decreasing in size, so it would not
Find the time complexity when (n/2i) = 1. get any bigger than n2.
Solution:
We have i = log n . So, the time complexity of n can 4.2.5.3 Master Theorem
be represented as
Master theorem provides a method to solve various
T (n) = n + 2n + 4n + 8n + 2i n + 4log n O(1)
recurrence relations. Although all the relations can’t
log(n -1)  be solved using this method, but it is one of the useful

We get n  ∑ 2i  + 4log n O(1) method of solving complex relations of the form:
 i = 0 
n
T (n) = aT   + f (n)
 b 
m
xm +1 - 1
We Know that ∑ x =
k
where a and b are constants. f is a function of n. this
k =0 x -1
recursive algorithm breaks up the input into sub-problems
 log(n -1)+1 - 1 of size n/b each.
So, n  2  + 4log n O(1)
 2 -1  Three cases of master theorem are as follows:

1. Case 1: If f(n) is O(n logb a-e ) then T(n) is Q(n logb a ) .
Solving, we get
The rate of growth at leaves is faster than f, more
n2log n - n + 4log n O(1) ⇒ n2 - n + n log 4O(1) work is done on leaves.
2. Case 2: If f(n) is Q(n logb a ) then T(n) as
⇒ 2n2O(1) - n ⇒ O(n2 ) Q(n logb a log n) . The rate of growth of leaves is same
as f. So, the same amount of work is done at every
So, the running time complexity is given as: T(n) = O(n2). level of tree. The leaves grow at the same rate as
f, so the same order of work is done at every level
of the tree. The tree has O(log n) times the work
done on one level, yielding T(n) as Q(n logb a log n) .
4.2.5.2 Recursion Tree 3. Case 3: f(n) is W(n logb a+e ). Here f grows faster than
the number of leaves, which means that asymptoti-
A recursion tree is useful for visualizing what happens
cally the total amount of work is dominated by the
when a recurrence is iterated. It represents a tree of
work done at the root node. For the upper bound, we
recursive calls and the amount of work done at each call
also need an extra smoothness condition on f in this
case, namely, that af(n/b) ≤ cf(n) for some constant
is shown with the help of a diagram.
c < 1 and large n. In this case, T(n) is Q(f(n)).
Example 4.2
Example 4.3
Consider the recurrence T (n) = 2T (n/2) + n2.
Consider the following recurrence:
n2
n
T (n) = 4T   + n
 2 
T(n/2) T(n/2)
n2 In the given relation, a = 4 and b = 2.
n2
For this recurrence, there are a = 4 sub-problems, each
(n/2)2 (n/2)2 1/2n2 dividing the input by b = 2, and the work done on each
call is f(n) = n. Thus, n logb a is n2 and f(n) is O(n2-e) for
e = 1, and Case 1 applies. Thus, T(n) is Q(n2).
(n/4)2 (n/4)2 (n/4)2 (n/4)2 1/4n2
Chapter 4.indd 181 4/9/2015 9:54:19 AM

Example 4.4 is placed. It is important to note that hash function

returns the same index value every time for same key.
n
T (n) = 4T   + n2 4.3.3 Collisions
 2 
For this recurrence, there are again a = 4 sub-problems, Collision occurs when one key value stores more than one
each dividing the input by b = 2; but now the work done value. Hash function producesthe same position for two
on each call is f(n) = n2. Again, n logb a is n2 and f(n) is values. Hash table cannot ensure to have single value
thus Q(n2), so Case 2 applies. Thus, T(n) is Q(n2 log n). corresponding to a key. This is because a hash table has
Note that increasing the work on each recursive call from fixed number of keys. Any new entry in hash table, after
linear to quadratic has increased the overall asymptotic all the slots are filled, causes collision. However, good
running time by only a logarithmic factor. hash tables use some mechanism to avoid the situations
of collisions or to handle them efficiently.
Example 4.5 4.3.3.1 Separate Chaining

1. If more than one item is assigned the same hash
n
T (n) = 4T   + n3 index, “chain” them together.
 2  2. In hash table, each position act as bucket that can
For this recurrence, there are again a = 4 sub-problems, store multiple data items.
each dividing the input by b = 2, but now the work done 3. There are the following two implementations:
on each call is f(n) = n3. Again, n logb a is n2 and f(n) is • Each bucket is itself an array: The disadvan-
thus W(n2+e ) for e = 1. Moreover, 4(n/2)3 ≤ kn3 for k = tages of bucket as array are
1/2, so Case 3 applies. Thus, T(n) is Q(n3). (a) Memory wastage
(b) Problem of overflow occurs when we try to
add new item to a full bucket.
• Each bucket is a linked list: The disadvantage of
4.3 HASHING
bucket as linked list is
Pointers require more memory.
Hashing is technique that transforms a variable length 0 arrow “ape”
input to a fixed length output. This technique is used for
1 null null
indexing, which helps in constant time search for inser-
tion, deletion and information retrieval. 2 “cat”
... ... null
4.3.1 Hash Table
A hash table is a data structure used for hashing. It is 4.3.3.2 Open Addressing

an associative array that contains keys with values. In
an array, indices are required to be integer, but a hash 1. Collision occurs if the hash function produces the
table allows floating point numbers, strings, array or same position occupied by another item. In this
structures as keys. Hash tables provide efficient lookup case, another open position is searched.
operation, that is, whenever a key is given, hash table
Example: Hash code produced for `White’ is 22.
provides associated value in almost constant time com-
But position 22 is already occupied, so it is placed
plexity. This transformation is done using hash function.
at position 23.
2. There are three ways of finding an open position.
4.3.2 Hashing Functions
This process is known as probing. To search any
A hashing function is a transformation function, coded item in hash table, probing may be required.
to return a value based on key. If h is a hash function Example: When item “white” is searched, first we
that takes key k as parameter, will compute the index at look for it at position 22. If it is not found there, we
which value is placed. If h is a hash function that takes move to next location. The search will be stopped
key as a parameter, it will compute index where value only when an empty position is encountered.
Chapter 4.indd 182 4/9/2015 9:54:21 AM

4.3 HASHING 183
4.3.3.4 Quadratic Probing
0 Arrow
The probe sequence is
1
h(key), h(key) + 1, h(key) + 4, h(key) + 9, …,
2 Cat wrapping around as necessary
3 The offsets are perfect squares: h + 12, h + 22, h + 32, ….
Examples:
4 Eye
a. Àpe’ (h = 0): try 0, 0 + 1 − open!
5 b. `Ball’ (h = 1): try 1, 1 + 1, 1 + 4, 1 + 9 − open!
... ... ... Advantage: It reduces clustering.
22 When 0 Arrow
23 White 1 Ape
24 Yellow 2 Cat
25 Zero 3
4 Eye
4.3.3.3 Linear Probing 5 Ball
In case of linear probing the probe sequence is h(key), ... ... ...
h(key) + 1, h(key) + 2, …, uptill required. 22 When
Examples: The item Àpe’ (h = 0) will be placed at
23 White
position 1, because position 0 is already occupied. then
`Ball’ (h = 1): try to place it at position 1, 1+1, 1+2. 24 Yellow
Position 3 is open so it will be placed here.
25 Zero
Advantage: If open position is available, linear probing
will definitely find it.
Disadvantage: It may fail to find an existing open
Disadvantage: If there are clusters of filled positions,
position. For example, let table size = 10, and x denote
then length of subsequent probes increases. Probe length
the occupied blocks. Trying to insert a key with h(key) =
is the number of positions considered during a probe.
0 offsets the probe sequence shown in italic.
0 Arrow
0 x 5 x 25
1 Ape
2 Cat 1 x 1 81 6 x 16 36
3 Ball 2 7
4 Eye 3 8
5 x
4 x 4 64 9 9 49
... ... ...
22 When 4.3.3.5 Double Hashing

23 White
It uses the following two hash functions:
24 Yellow 1. h1 computes the hash code.
25 Zero 2. h2 computes the increment for probing.
The probe sequence is: h1, h1 + h2, h1 + 2 ×h2, ….
Chapter 4.indd 183 4/9/2015 9:54:22 AM

Examples:
0 Arrow
1. h1 = our previous h
2. h2 = number of characters in the string 1 1
3. Àpe’ (h1 = 0, h2 = 3): try 0, 0 + 3 − open!
4. `Ball’ (h1 = 1, h2 = 4): try 1 − open!
2 Cat

3 Ball
Advantages:
4 Eye
1. It combines the good features of linear and qua-
dratic probing: 5
2. It reduces clustering
3. It will find an open position if there is one, pro- ... ... ...
vided the table size is a prime number 22 When
23 White
0 Arrow 24 Yellow
1 Ball 25 Zero
2 Cat
3 Ape
4.4 BINARY HEAP
4 Eye
5 A binary heap is a binary tree with two additional constraints:
... ... ... 1. Binary heap should satisfy the property of com-
22 When plete binary tree.
2. It should satisfy the heap ordering property, that
23 White is, all the nodes should be either greater than or
equal to or less than or equal to its children.
24 Yellow
According to the heap ordering property, binary heaps
25 Zero
are of two types:
1. Min-Heap Tree: In Min heap ordering, the

value of each node is less or equal to its child node.
4.3.3.6 Removing Items Under Open Minimum element of the tree will be the root ele-
Addressing ment. The largest element exists in the last level.
The tree should be an almost complete binary tree.
Consider the following scenario:
Using linear probing:
(a) Insert Àpe’ (h = 0): Try 0, 0 + 1 − open! 7
(b) Insert `Ball’ (h = 1): Try 1, 1 + 1, 1 + 2 − open!
(c) Remove Àpe’.
(d) Search for Àpe’: Try 0, 0 + 1 − no item. 10 17
(e) Search for `Ball’: Try 1 − no item, but `Ball’ is
further down in the table.
1. On removing an item from a position, leave a spe- 14 23 21 42
cial value in that position to indicate that an item
was removed.
2. There are three types of positions: occupied, empty 2. Max-Heap Tree: In max heap, each node in the
and removed. tree will be greater or equal to its child node. The
3. Stop probing on encountering an empty position, maximum element of tree will be the root element.
but not when encountering a removed position. The smallest element exists in the last level. The
4. Insert items in either empty or removed positions. tree should be an almost complete binary tree.
Chapter 4.indd 184 4/9/2015 9:54:23 AM

4.4 BINARY HEAP 185
4.4.1 Insertion in Min-Heap Tree

10
The following are the steps to follow for insertion in the
min-heap tree:
8 9 1. Add the new element to the last level of heap.
2. Compare the new element with its parent node.
3. If the order is correct then stop, otherwise swap the
4 1 2 6 element with its parent.
4. Perform steps 2 and 3 until a new element is at position.
Example 4.6
1. Insert the given value (5) in the given min-heap tree. 4. To satisfy the property of min-heap, swap (5, 7).
2. Element 5 is inserted at position of left child of node 5. Now, element 5 is in its right position. This is the
14. To satisfy the property of min-heap, swap (5, 14). final tree.
3. To satisfy the property of min-heap, swap (5, 10).
7 7
7
10 17 10 17
10 17
14 23 21 42 5 23 21 42
14 23 21 42
5 14
1 2 3
7 5
5 17 7 17
10 23 21 42 10 23 21 42
14 14
4 5
4.4.2 Deletion in Min-Heap Tree Example 4.7

The following are the steps to follow for deletion in the 1. Deletion always occurs at the root level, so element
min-heap tree: 7 will be deleted. Before deletion of element 7 at
the root, the element (42) is swapped to the root
1. Root of the heap is replaced with the element at node from the extreme right-hand side of the last
the last level. level and then 7 is deleted.
2. New root element is compared with its children; if 2. To maintain the min-heap property of the tree, swap
it maintains ordering then stop. 42 and min(left-child, right-child), which means
3. Otherwise, swap the root with one of its chil- swap 42 with 10.
dren and return to previous step. (Smaller ele- 3. To maintain the min-heap property of the tree, swap
ment is swapped for min-heap and larger child for 42 and min(left-child, right-child), which means
max-heap) swap 42 with 14.
Chapter 4.indd 185 4/9/2015 9:54:24 AM

4. This represents the final tree, which is in the min- a set of numbers in an ascending or descending
heap property. order.
3. In internal sorting, the data which is to be sorted fits
7 42 into the main memory. Internal sorting has an advan-
tage of the random access nature of the main memory.
10 17 10 17 4. In external sorting, large amount of data is
sorted. This large amount of data cannot be fit
14 23 21 42 14 23 21 in main memory, so the data is placed in auxil-
iary memory. When sorting has to perform, small
1 2
chunks of data are brought into main memory,
10 10 sorted and placed back to a temporary file. At last
all the sub-files are merged in one large file.
42 17 14 17 5. In stable sorting, the order of equal elements is
preserved in a sorted list. Some of the sorting algo-
rithms that show characteristics of stable sort are
14 23 21 42 23 21
insertion sort, bubble sort, merge sort, etc.; whereas
3 4 heap sort, quick sort, etc., are not stable sorts.
6. In-place sorting: In this, the input is transformed
using a data structure with a small, constant amount
4.4.3 Time Complexity of extra storage space. The input is usually overwrit-
ten by the output as the algorithm executes. Quicksort
1. The worst-case running time for insertion is is commonly described as an in-place algorithm.
O(log n) because at most (log n) swaps are required 7. Measurement of time complexity: The time
to stable the min-heap property of the tree. complexity of a sorting algorithm is measured
2. The worst-case running time for deletion is O(log n) by the number of critical operations performed.
because at most (log n) swaps are required to stable Examples of critical operations are as follows:
• Key comparisons
the min-heap property of the tree.
Some important points of binary heap are as follows: • Movements of records
1. Because of its structure, a heap with height k will • Interchanges of two records
have between 2k and 2k+1 - 1 elements. Therefore, 8. We use asymptotic analysis to denote the time
a heap with n elements will have height = log2n. efficiency of an algorithm. Efficiency of an algorithm
2. In heap data structure, element with highest or lowest also depends on the type of data for input, so we
priority is stored as root element. Heap is not consid- define best-, worst- and average-case efficiencies.
ered as sorted structure, rather it is partially ordered.
There is no relation among nodes at a given level. 4.5.1 Linear Search
Heap is very useful data structure for implement
ing priority queues. Element with highest priority Linear search is a sequential way of finding an element in
can be easily removed. a given list. Linear search is a special case of brute force
3. Since a binary heap is a complete binary tree, it search. Its worst case is proportional to the number of
can be easily represented as an array, and an array- elements in list.
based representation is space efficient. If the parent
node is stored at index I, the left child can be cal-
4.5.1.1 Pseudocode for Linear Search
culated by 2 ×I + 1 and right child by 2 ×I + 2.
LINEAR_SEARCH (A[ ], x)
4.5 SEARCHING AND SORTING 1. i = 1
2. while (A[i] = = x and i < n)
3. i = i + 1
4. if (A[i] = = x)
Searching and sorting are the most common activities
5. return i
performed by the computer. These operations are the 6. else return 0
most important part for managing the data.
1. Searching is the algorithmic process of finding an 4.5.1.2 Time Complexity
item in a collection of items. The search result is
either true or false depending on whether or not From the analysis of above pseudocode, it can be
the item is present or not in the list. observed that with the growth of input list, average and
2. Sorting is ordering a list of objects according to a worst case complexities also grow linearly. In general, if
given condition or property. For example, ordering there are n elements in the list, the worst-case requires n
Chapter 4.indd 186 4/9/2015 9:54:24 AM

4.5 SEARCHING AND SORTING 187
comparisons. So, the complexity of linear search can be 7, 5, 2, 4, 3, 9

expressed in terms of some linear function. 5, 7, 2, 4, 3, 9
So, running time T(n) = O(n).
5, 2, 7, 4, 3, 9
5, 2, 4, 7, 3, 9
4.5.2 Binary Search
5, 2, 4, 3, 7, 9
Binary search relies on a divide-and-conquer strategy to After the first pass, the first largest element (9) is on the
find an element within an already sorted list. It compares top of array. So we have
the element to be searched with the middle element of
array. If the element is equal to the middle element, the 5, 2, 4, 3, 7, 9
search stops here and the position of element is returned. After the second pass, we get
If it is less or greater than the middle element, then 2, 4, 3, 5, 7, 9
the values of sub-array are adjusted accordingly. The
algorithm returns a unique value, that is, position of the After the third pass, we get
element in the list, if the element is present. 2, 3, 4, 5, 7, 9
After the fourth pass, we get
4.5.2.1 Pseudocode for Binary Search
2, 3, 4, 5, 7, 9
BINARY_SEARCH (A, value, left, right) After the fifth pass, we get
1. if right < left
2. return ‘not found’ 2, 3, 4, 5, 7, 9
3. mid = floor((right-left)/2)+left The sorted list is 2, 3, 4, 5, 7, 9
4. if A[mid] = value
5. return mid
6. if value < A[mid] 4.5.3.2 Time Complexity
7. return BINARY_SEARCH(A, value, left, mid-1)
8. else
9. return BINARY_SEARCH(A, value, mid+1, right) It can be seen from the above example that in the bubble
sort, each element passes one at a time and is placed in
its correct position.
4.5.2.2 Time Complexity So, the number of passes = (number of elements − 1)
= (n − 1).
The running time of the binary search is Q(log n).
Cost per pass = O(n).
Total cost = O(n)(n − 1).
4.5.3 Bubble Sort
So, the worst-case runtime complexity is O(n2).
Bubble sort algorithm is a comparison-based algorithm. The average-case and best-case runtime complexity is
It compares each element in the list with the next ele- O(n2).
ment. Elements are swapped if required. In every pass,
one element comes to its position. This process repeats
until a pass is made without disturbing any of the ele- 4.5.4 Selection Sort
ments from their position.
The selection sort first selects the smallest element from
4.5.3.1 Pseudocode for Bubble sort the unsorted list. Then that element is swapped with the
first element in the list. In the next step, the size of the
BUBBLE_SORT(A) list is reduced by one because one element is present at
1. for i 1 to n-1 its position. Next, the smallest element is swapped with
2. for j 1 to n-1 the second element in list, and so on.
3. If(A(j) > A(J+1))
4. Temp = A(j)
5. A(j) = A(j+1) 4.5.4.1 Pseudocode for Selection Sort
6. A(j+1) = Temp
SELECTION_SORT(A)
1. for j 1 to n-1
2. smallest j
Example 4.8
3. for i j + 1 to n
Sort the given elements (7, 5, 2, 4, 3 and 9) by using 4. if A[ i ] < A[ smallest]
bubble sort. 5. smallest i
6. Exchange A[ j ] A[ smallest]
After the first pass, we get
Chapter 4.indd 187 4/9/2015 9:54:25 AM

Example 4.9 Example 4.10

Sort the given elements (29, 64, 73, 34 and 20) by using Sorted part is bold in text and unsorted part is non-bold.
selection sort. Suppose 30, 10, 75, 33, 65 needs to be sorted using inser-
We have tion sort.
29, 64, 73, 34, 20 After the first pass, we get
After the first pass, we get 10, 30, 75, 33, 65
20, 64, 73, 34, 29 After the second pass, we get
After the second pass, we get 10,30,75, 33, 65
20, 29, 73, 34, 64 After the third pass, we get
After the third pass, we get 10, 30, 35, 75, 65
20, 29, 34, 73, 64 After the fourth pass, we get
After the fourth pass, we get 10, 30, 35, 65, 75
20, 29, 34, 64, 73 The sorted list is 10, 30, 35, 65, 75.
After the fifth pass, we get
20, 29, 34, 64, 73 4.5.5.2 Recurrence Relation of Insertion Sort
The sorted list is 20, 29, 34, 64, 73.
The recurrence relation for insertion sort is
 Q(1), n = 1
T (n) =  
4.5.4.2 Time Complexity T (n - 1) + Q(n), n > 1
Number of passes = (Number of elements − 1)
4.5.5.3 Time Complexity
Cost per pass = Swap + Comparisons = 1 + O(n)
Total cost = (n − 1)[1 + O(n)] In sorting, the most expensive part is the comparison of
So, the worst-case runtime complexity is O(n2). two elements. Zero comparisons are required to insert
first element, one for second element, two comparisons
The average-case and best-case runtime complexity is O(n2). for third element, and so on. Atmost N − 1 comparisons
are required for the last element in the list.
Complexity is: 1 + 2 + 3 + 4 … + (N − 1) = O(n2)
4.5.5 Insertion Sort
The insertion sort inserts each element into its proper The worst-case running time is O(n2)
position. It chooses one element, inserts it to its posi- The best-case running time is O(n)
tion and shifts the rest of the list by one location. This
process is repeated until no input element remains. Step- 4.5.6 Heap Sort
by-step procedure is given as follows.
1. Take an element from unsorted list. Heap sort is one of the comparison based sort that uses
2. Compare that element with sorted list from right heap data structure. It is a special case of selection sort.
to left. Its typical implementation is not stable, but can be made
3. Shift the list. stable. Heap can be built by the following two types:
Here, it is assumed that first element is already sorted. 1. Min-Heap: The parent node will be less than or
equal to its children nodes. So, the minimum ele-
ment will be at root.
4.5.5.1 Pseudocode for Insertion Sort 2. Max-Heap: The parent node will be greater than
or equal to its children nodes. The maximum ele-
INSERTION_SORT(A)
1. for j = 2 to n ment will be at the root.
2. key A [j]
3. // Insert A[j] into the sorted sequence A[1..j-1] 4.5.6.1 Heap Sort Algorithm for Sorting in
4. j i– 1 Increasing Order
5. while i > 0 and A[i] > key
6. A[i+1] A[i] 1. First build a max-heap from the input List.
7. i i– 1 2. In max heap, the largest element is at the root. So,
8. A[j+1] key swap the root with last element. Re-heapify the
Chapter 4.indd 188 4/9/2015 9:54:26 AM

tree. Repeat the above steps until the size of the 4.5.6.3 Pseudocode for BuildHeap Function
heap is greater than 1.
•• Heapify function: Heapify is an important BUILD_HEAP(A)
subroutine for manipulating max-heaps. Its 1. heapsize <- length(A)
2. for i <- floor( length/2 ) down to 1
inputs are an array A and an index i into the
3. Heapify(A, i)
array. When Heapify is called, it is assumed
that the binary tree rooted at left(i) and right(i) 4.5.6.4 Pseudocode for Heap Sort
is max_heap, but that A[i] may be smaller
than its children, thus violating the max-heap HEAP_SORT(A)
property. 1. BuildHeap(A)
•• Build Heap function: Heapify procedure 2. for i <- length(A) down to 2
is used in a bottom-up to convert an array A 3. exchange A[1] <-> A[i]
4. heapsize <- heapsize -1
into a max-heap. The elements in the subarray
5. Heapify(A, 1)
A[(n/2 + 1)⋅⋅⋅n] are all leaves of the tree, and
so each is a one-element heap to begin with. The
procedure BuildHeap goes through the remain- 4.5.6.5 Recurrence Relation of Heapify
ing nodes of the tree and runs Heapify on each Function
node. The recurrence relation of Heapify() is
  2n  
4.5.6.2 Pseudocode for Heapify Function T (n) ≤ T   + Q(1)
  3  
HEAPIFY(A, i)
1. le <- left(i) 4.5.6.6 Time Complexity
2. ri <- right(i)
3. if (le<=heapsize) and (A[le]>A[i]) The time complexity of Heapify() is O(log n). The time
4. largest <- le
5. else
complexity of BuildHeap() is O(n log n), and the overall
6. largest <- i time complexity of the heap sort is O(n log n).
7. if (ri<=heapsize) and (A[ri]>A[largest])
8. largest <- ri 4.5.6.7 Applications of Heap Sort
9. if (largest != i)
10. exchange A[i] <-> A[largest] 1. It sorts a nearly sorted (or k-sorted) array.
11. HEAPIFY(A, largest) 2. It sorts k largest (or smallest) elements in an array.
Example 4.11
Let {6, 5, 3, 1, 8, 7, 2, 4} be the list that we want to sort recursively checked if another swap is needed, to keep
from the smallest to the largest. (Note: For `building larger numbers above smaller numbers on the heap
the heap’ step, larger nodes do not stay below smaller binary tree.)
node parents. They are swapped with parents, and then 1. Build the heap:
Heap Newly Added Element Swap Elements

Nil 16
16 20
20 16 16 20
20 16 11 11
20 16 11 8 8
20 16 11 8 18
20 16 11 8 18 16 18
20 18 11 8 16 14
20 18 11 8 16 14 11 14
20 18 14 8 16 11
Chapter 4.indd 189 4/9/2015 9:54:28 AM

2. Sort the list:

Heap Swap Elements Delete Element Sorted Array Details
20 18 14 8 16 11 20, 11 Swap 20 and 11 in order to
delete 20 from the heap.
11 18 14 8 16 20 20 Delete 20 from the heap and
add to the sorted array.
11, 18, 14, 8, 16, 20 11, 18 20 Swap 11 and 18 as they are
not in order in the heap.
18 11 14 8 16 20 11 16 20 Swap 11 and 16 as they are not
in order in the heap.
17, 15, 16, 7, 10, 19 18 11 20 Swap 18 and 11 in order to
18 16 14 8 11 20 delete 17 from the heap.
11 16 14 8 18 20 18 20 18 Delete 18 from the heap and
11 16 14 8 18 20 11 14 20 18 Swap 11 and 14 as they are not
11 14 16 8 18 20 14 16 20 18 Swap 14 and 16 in order to
16 14 11 8 18 20 20 18 16 Delete 16 from the heap and
8 14 11 16 18 20 8 14 20 18 16 Swap 14 and 8 as they are not
14 8 11 16 18 20 14 11 20 18 16 Swap 14 and 11 in order to
10, 7, 15, 16, 17, 19 14 20 18 16 14 Delete 14 from the heap and
11 8 14 16 18 20 add to the sorted array.
11 8 14 16 18 20 11 8 20 18 16 14 Swap 11 and 8 in order to
8, 11, 15, 16, 17, 11 20 18 16 14 11 Delete 11 from the heap and
19 add to the sorted array.
7, 10, 15, 16, 17, 19 8 20 18 16 14 11 8 Delete 8 from the heap and add
8 11 14 16 18 20 to the sorted array.
MERGE_SORT
20 18 16 (A,
14 11p, r)Completed
1. if p < r
2. Then q = floor[(p + r)/2]
4.5.7 Merge Sort
3. MERGE (A, p, q)
Merge sort is a divide-and-conquer algorithm. It divides 4. MERGE (A, q + 1, r)
the input array in two subarrays, calls itself for the two 5. MERGE (A, p, q, r)
sub-arrays and then merges the two sorted arrays. The Pseudocode for MERGE function is given as follows:
MERGE(A, p, q, r) function is used for merging the two
arrays. The MERGE(A, p, q, r) is a key process that MERGE (A, p, q, r)
assumes that array[l⋅⋅⋅m] and array[m + 1 ⋅⋅⋅ r] are sorted 1. n1 ¬ q − p + 1
and merges the two sorted sub-arrays into one. Merge sort 2. n2 ¬ r − q
is a stable sort. It is not an in-place sorting technique. 3. create arrays L[1…n1 + 1] and R[1…n2 + 1]
4.5.7.1 Pseudocode for Merge Sort 4. For i ¬ 1 to n1

5. doL[i] ¬ A[p + i − 1]
MERGE_SORT (A, p, r)
6. For j ¬ 1 to n2
1. if p < r
2. Then q = floor[(p + r)/2] 7. Do R[j] ¬ A[q + j]
3. MERGE (A, p, q) 8. L[n1 + 1] ¬ ∞
4. MERGE (A, q + 1, r) 9. R[n2 + 1] ¬ ∞
5. MERGE (A, p, q, r)
10. i ¬ 1
Chapter 4.indd 190 4/9/2015 9:54:29 AM
11. j ¬ 1
2. n2 ¬ r − q
3. create arrays L[1…n1 + 1] and R[1…n2 + 1]
4. For i ¬ 1 to n1
5. doL[i] ¬ A[p + i − 1]
6. For j ¬ 1 to n2
7. Do R[j] ¬ A[q + j]
8. L[n1 + 1] ¬ ∞ 4.5.7.3 Time Complexity
9. R[n2 + 1] ¬ ∞
The above recurrence can be solved using either recur-
10. i ¬ 1 rence tree method or master theorem. It falls under case
11. j ¬ 1 2 of the master theorem, and the solution of the recur-
rence is O(n log n). The time complexity of merge sort
12. For k ¬ p to r
in all three cases (worst, average and best) is the same,
13. Do if L[i] R[j] as merge sort always divides the array in two sub-arrays
14. then A[k] ¬ L[i] and takes linear time to merge the two arrays.
15. i ¬ i + 1
16. Else A[k] ¬ R[j] 4.5.7.4 Applications of Merge Sort
17. j ¬ j + 1
1. Merge sort is useful for sorting linked lists in O(n
log n) time. Other n log n algorithms such as heap
4.5.7.2 Recurrence Relation of Merge Sort
sort and quick sort (average case n log n) cannot
The recurrence relation for merge sort is be applied to linked lists.
2. It is used in inversion count problem.
 Q(1), n = 1
 3. It is used in external sorting.
T (n) =   n  
2T   + Q(n), n > 1 Note: Merge sort is not preferable for smaller-size
  2   element array.
Example 4.12
Sort the given list of elements: (54, 26, 93, 17, 77, 31, 44, Step 2: After dividing the array, merge sort calls the
55 and 20). MERGE function to put the subarray into a sorted order.
Step 1: A recursive merge sort is called on the array Array A:
of elements. Merge sort is a recursive function which
divides an array into two parts; and calls itself recur- 54 26 93 17 77 31 44 55 20
sively to divide the array into sub-parts until each part
is having one element.
p r
Merge sort (A, 0, 8) 54 26 93 17 77 31 44 55 20
0 1 2 3 4 5 6 7 8
54 26 93 17 77 31 44 55 20
54 26 93 17 77 31 44 55 20
54 26 93 17 77 31 44 55 20
54 26
Merge (A, p, q, r)
54 26 93 17 77 31 44 55 20
26 54 93 17 77 31 44 20 55
26 54 93 17 77 31 44 20 55
17 26 54 77 93 20 31 44 55
Completed 17 20 26 31 44 54 55 77 93
Chapter 4.indd 191 4/9/2015 9:54:30 AM

PARTITION(S, p, r)
1. x <- S[r]
2. i <- p-1
192 Chapter 4: Algorithms 3. for j <- p to r-1
4. do if S[j] <= x
4.5.8 Quick Sort 5. then i <- i+1
6. swap S[i] <-> S[j]
Quicksort is also divide-and-conquer strategy of sorting a list 7. swap S[i+1] <-> S[r]
of elements like merge sort. This divides array of elements 8. return i+1
into sub-arrayS[p…r]. The concept is described as follows:
4.5.8.3 Recurrence Relation of Quick Sort
1. Divide: Partition S[p…r] into two sub-arrays
S[p…q − 1] and S[q + 1…r] such that each element 1. For worst case: The worst case of quick sort
of S[p…q − 1] is less than or equal to S[q], which occurs when the pivot we picked turns out to be
is, in turn, less than or equal to each element of the least element of the array to be sorted, in every
S[q + 1…r]. Compute the index q as part of this step (i.e. in every recursive call). A similar situa-
partitioning procedure. tion will also occur if the pivot happens to be the
2. Conquer: Sort the two sub-arrays S[p…q − 1] and largest element of the array to be sorted. Then
S[q + 1…r] by recursive calls to quicksort. recurrence relation of quick sort is
3. Combine: Since the sub-arrays are sorted in T (n) = T (1) + T (n - 1) + Θ(n)
place, no work is needed to combine them; the 2. For best case: The best case of quicksort occurs
entire array S is now sorted. when the pivot we picked happens to divide the
array into two almost equal parts, in every step.
4.5.8.1 Pseudocode for Quick Sort Thus, we have k = n/2 and n − k = n/2 for the
QUICKSORT(S, P, r) original array of size n. Then the recurrence rela-
1. If p < r tion of quick sort is given as
2. then q <- PARTITION(S, p, r)  n
T (n) = 2T   + Θ(n)
3. QUICKSORT(S, p, q-1)  2
4. QUICKSORT(S, q+1, r)
4.5.8.2 Pseudocode for Partition
The best-case and average-case running time of quick
PARTITION(S, p, r) sort = O(n log n).
1. x <- S[r]
2. i <- p-1 The worst-case running time of quick sort = O(n2).
3. for j <- p to r-1 Note: The worst case of quick sort occurs when the
4. do if S[j] <= x given elements are already sorted or almost sorted.
5. then i <- i+1
6. swap S[i] <-> S[j]
Example
7. 4.13
swap S[i+1] <-> S[r]
8. return i+1
Sort the given list of elements: (38, 81, 22, 48, 13, 69, 93, (a) Elements at the left of the pivot are less than or
14, 45, 58, 79, 72) equal to that of pivot.
Step 1: Choose the pivot element which is at position (b) Elements at the right of pivot are greater than that
[(left + right)/2]. of pivot.
Step 2: Now in partitioning phase: During the parti- The tree of recursive call is given as follows:
tioning process:
Quicksort (A,1,12)
38 81 22 48 13 69 93 14 45 58 79 72
Pivot = 69
14 58 22 48 13 38 45 69 93 81 79 72
Quicksort (A,1,7) Quicksort (A,9,12)

14 58 12 48 13 38 45 Pivot = 48 93 81 79 72 Pivot = 81
38 45 22 14 13 45 58 79 72 81 93

14 45 22 48 13 Pivot = 22 79 72 Pivot = 79
14 13 22 38 45

14 13 38 45 Pivot = 38
Pivot = 14
13 14 38 45
Chapter 4.indd 192 4/9/2015 9:54:31 AM

COUNTING_SORT(A[], B[], k)
1. for i = 1 to k do
2. C[i] = 0
3. for j = 1 to length(A) do AND SORTING 193
4.5 SEARCHING
4. C[A[j]] = C[A[j]] + 1
4.5.9 Randomized Quick Sort 5. for 2 = 1 to k do
6. C[i] = C[i] + C[i-1]
In the randomized version of quick sort, we impose a 7. for j = 1 to length(A) do
distribution on the input. This does not improve the B[C[A[j]]] = A[j]
C[A[j]] = C[A[j]] - 1
worst-case running time, but it improves the probability
of getting the average-case running time. Although this may look complicated, it is actually a
In this version, we choose a random key for the pivot. very simple and clever algorithm.
Assume that the procedure Random(a, b) returns a 1. An array A[ ] stores the initial data to be sorted.
random integer in the range [a, b); there are b − a + 1 2. An array C[ ] is used to count the occurrences of
integers in the range, and the procedure is equally likely the data values.
to return one of them. The new partition procedure 3. An array B[ ] is used to store the final, sorted, list.
simply implements the swap before actually partitioning. 4. The first for loop initializes C[ ] to zero.
5. The second for loop increments the values in C[ ],
4.5.9.1 Pseudocode for Randomized according to their frequencies in the data.
Quick Sort 6. The third for loop adds all the previous values,
making C[ ] contain a cumulative total.
RANDOMIZED_PARTITION (A, p, r)
7. The fourth for loop writes out the sorted data
1. i RANDOM (p, r) into the array B[ ].
2. Exchange A[p] A[i]
3. return PARTITION (A, p, r)
The pseudocode for randomized quick sort has the O(n + k), where n is the number of elements in the input
same structure as quick sort, except that it calls the ran- array and k is the range of input.
domized version of the partition procedure.
Note:
RANDOMIZED_QUICKSORT (A, p, r) 1. Counting sort is efficient if the range of input data
is not significantly greater than the number of
1. If p<r then objects to be sorted.
2. Q RANDOMIZED_PARTITION (A, p, r) 2. It is not a comparison-based sorting. Its running
3. RANDOMIZED_QUICKSORT (A, p, q) time complexity is O(n) with space proportional to
the range of data.
4. RANDOMIZED_QUICKSORT (A, q+1, r) 3. This can be used as subroutine to some other algo-
rithm. It is often used as a subroutine to another
4.5.9.2 Time Complexity sorting algorithm, such as radix sort.
4. It uses a partial hashing to count the occurrence of
The best- and average-case running time of randomized the data object in O(1).
quick sort = O(n log n). 5. It can be extended to work for negative inputs also.
The worst-case running time of randomized quick
sort = O(n2).
Note: The worst case of randomized quick sort occurs Example 4.14
when all the given elements are equal. Consider the data in the range 0 to 9.
4.5.10 Counting Sort Input data: 1, 4, 1, 2, 7, 5, 2
The steps to sort elements through counting sort are as
Counting sort is a sorting technique based on keys follows:
within a specific range. It works by counting the number 1. Take a count array to store the count of each
of objects having distinct key values (kind of hashing), unique object:
and then doing some arithmetic to calculate the position
Index: 0 1 2 3 4 5 6 7 8 9
of each object in the output sequence.
Count: 0 2 2 0 1 1 0 1 0 0
4.5.10.1 Pseudocode for Counting Sort 2. Modify the count array such that each element at
each index stores the sum of the previous counts:
COUNTING_SORT(A[], B[], k)
Index: 0 1 2 3 4 5 6 7 8 9
1. for i = 1 to k do
2. C[i] = 0 Count: 0 2 4 4 5 6 6 7 7 7
3. for j = 1 to length(A) do The modified count array indicates the position of each
4. C[A[j]] = C[A[j]] + 1 object in the output sequence.
5. for 2 = 1 to k do
6. C[i] = C[i] + C[i-1]
7. for j = 1 to length(A) do
B[C[A[j]]] = A[j]
C[A[j]] = C[A[j]] - 1
Chapter 4.indd 193 4/9/2015 9:54:32 AM
Table 4.1 | Sorting techniques

Sorting Average Best Time Worst Time Space Stability In-Place
Algorithm Time Sorting
Bubble sort O(n2) O(n2) O(n2) Constant Stable Yes
Selection sort O(n2) O(n2) O(n2) Constant Stable Yes
2 2
Insertion sort O(n ) O(n) O(n ) Constant Stable Yes
Heap sort O(n × log(n)) O(n × log(n)) O(n × log(n)) Constant Instable Yes
Merge sort O(n × log(n)) O(n × log(n)) O(n × log(n)) Depends Stable No
Quick sort O(n ×log(n)) O(n × log(n)) O(n2) Constant Stable Yes
Count sort O(n) O(n) O(n) Constant Stable No
3. Output each object from the input sequence fol- • It does not have any parallel edge.
lowed by decreasing its count by 1.
(a) Process the input data: 1, 4, 1, 2, 7, 5, 2. Posi-
tion of 1 is 2.
(b) Put data 1 at index 2 in the output. Decrease
count by 1 to place the next data 1 at an index 1
smaller than this index.
(c) Put data 2 at index 4 in the output. Decrease Figure 4.5 | Simple graph.
count by 1 to place the next data 2 at an index 3
smaller than this index. 2. Multigraph: A graph with a self-loop and parallel
(d) Now, move to the next element 3 which has count 4, edges is called a multigraph (Fig. 4.6).
and index 4 is already filled by 2 which means ele-
ment 3 does not exist in the given input list.
(e) Follow this process for all the elements.
4.5.11 Comparison of Sorting Techniques
The comparison of the various sorting techniques is

Figure 4.6 | Multigraph.
given in Table 4.1.
4.6 GRAPH 4.6.2 Types of Simple Graph
4.6.2.1 Null Graph
A graph G = (V, E) is a set of vertices and edges where
each edge consist of pair of vertices.
If in the given graph there is no edge, then it is called a
1. Finite set of vertices are known as nodes of graph. null graph. In other words, a null graph does not have
2. (u, v) is a pair of vertices called edge. any edge between any two vertices (Fig. 4.7).
3. For directed graph (u, v) is an ordered pair that
represents an edge from node u to node v. A
D
4.6.1 Types of Graph
B
There are two types of graphs:
1. Simple graph: A graph is called a simple graph E
C
(Fig. 4.5) if
• It contains no self-loop. Figure 4.7 | Null graph.
Chapter 4.indd 194 4/9/2015 9:54:33 AM

4.7 GREEDY APPROACH 195
4.6.2.2 Regular Graph graph. If adj[i][j] = 1, this indicates there exists an edge

between vertices i and j; otherwise, the value will be 0.
A graph is called a regular graph if all of its vertices have The adjacency matrix for undirected graph is symmet-
the same degree (Fig. 4.8). ric. Weights can also be represented by this matrix. The
value adj[i][j] = w means weight of edge from i to j is w.
Vertex 1 2 3 4 5
1 2
1 0 1 1 1 0
5 2 1 0 1 1 1
3 1 1 0 1 0
1-Regular 2-Regular 3-Regular 4 1 1 1 0 1
Figure 4.8 | Regular graph.
3 4
5 0 1 0 1 0
If a graph has n number of vertices and it is k-regular,

This type of representation is very easy for implementa-
where k is a positive constant, then the number of edges
tion purpose. O(1) time is required to remove an edge.
= (n × k)/2.
Testing that an edge exists between two nodes or not
can be computed in O(1) time complexity.
4.6.2.3 Complete Graph
Space requirement is more for both dense and sparse
In the given graph, if every vertex is adjacent to all other matrices is O(V 2). Addition of a vertex takes O(V2) time.
remaining vertices, then that graph is called a complete
graph (Fig. 4.9). If a graph is a complete graph and is 4.6.3.2 Adjacency List
having n vertices, then it is called a Kn graph.
To implement adjacency list, an array of linked list is
used. Array size is kept equal to the total number of ver-
tices. For each vertex i, linked list of its adjacent nodes
is stored in array[i]. Weights of edges can also be stored
in linked list nodes. The representation of adjacency list
K4 graph K5 graph is as follows:
Figure 4.9 | Complete graphs.
A complete graph having n vertices has [n ×(n −1)]/ 1 2 3 4 Null
2 edges. 2 1 3 4 5 Null
Note: The number of simple graphs possible with n ver- 3 1 2 4 Null
tices = 2[n(n −1)]/2. 4 1 2 3 5 Null
5 2 4 Null
4.6.3 Graph Representation
The space required for adjacency list is: O(V+E),
Graph data structure is useful in many real-time applica- which is less than that of adjacency matrix. In worst-
tions like they are used to represent complex networks. case, space consumed is O(V2). Addition of vertex is
The network represented by graph can be of city defin- easy, but finding an edge between vertices, say u and v,
ing paths, of telephone or circuit network. is not efficient. It requires O(V) complexity.
The following two are the most commonly used repre-
sentations of a graph:
1. Adjacency matrix
4.7 GREEDY APPROACH
2. Adjacency list
Incidence matrix and incidence list are two other rep- A greedy algorithm is an algorithm for solving prob-
resentations for graph. Any representation can be chosen lems in an optimised way. In this approach,the algo-
depending on the situation. rithms execute repeatedly to maximize the outcome,
considering the local conditions for some global prob-
4.6.3.1 Adjacency Matrix lem. For some problems, it guarantees the optimal
solution while for some it does not. The main strat-
An adjacency matrix is a two-dimensional array of size egy is to implement the problem so that it requires
V ×V, where V represents the number of vertices in a minimal resources.
Chapter 4.indd 195 4/9/2015 9:54:35 AM

The important terms are given as follows: knapsack exactly; otherwise, we could add a fraction of
1. Solution space: The set of all possible solutions one of the remaining objects and increase the value of
for given n inputs is called solution space. the load. Thus, in an optimal solution, nSi = 1xiwi = W.
2. Feasible solution: The feasible solution is one of
the solutions from the solution space which satis- 4.7.2.1 Pseudocode for Fractional
fies the given condition.
Knapsack
3. Optimal solution: The optimal solution is one
of the feasible solutions which optimizes our goal. Greedy-fractional-knapsack (w, v, W)
1. for i =1 to n
4.7.1 Applications of Greedy Approach
2. Do x[i] =0
The following are the applications of the greedy approach: 3. weight = 0
4. while weight < W
1. Knapsack problem (fractional)
5. Do i = best remaining item
2. Job sequence with deadline
6. if ( weight + w[i] ≤ W)
3. Huffman coding
4. Minimum cost spanning tree: 7. Then x[i] = 1
a. Kruskal’s algorithm 8. weight = weight + w[i]
b. Prim’s algorithm 9. else
5. Singlesource shortest path:
10. x[i] = (w - weight) / w[i]
a. Dijkstra’s algorithm
b. Bellman—Ford algorithm 11. weight = W
12. Return x
4.7.2 Fractional Knapsack
In this problem, items are given along with its weight
and profit. The target is to maximize the profit consider- If the ratio of vi/wi is already sorted in decreasing order,
ing the weight constraint. Unlike 0-1 knapsack problem, then the time taken by the while loop will be O(n). So,
fraction of item can be considered. Fractional knapsack the total time required will be O(n log n). If heap data
problem can be solved by Greedy approach. structure is used, which keeps highest value/weight ratio
Greedy Algorithm: The solution can be obtained by at root. The construction of heap will take O(n) time
making different choices. At each stage, the best possible and while loop will take O(log n) time (heap property is
choice is made. In fractional knapsack, first of all profit restored for each root removal). This does not improve
value/weight ratios are calculated and then sorted in the worst case, but it is faster if small number of items
descending order. Item with highest value/weight ratio are to be filled in knapsack.
is chosen and added in the collection. The maximum
weight is checked after adding each item. If the entire Example 4.16
item cannot be included, then fraction of it is added to
Consider the following details:
the collection.
1. Knapsack size = 20
2. Number of objects = 3
Example 4.15
There are n items in a store. For i = 1, 2, …, n, item i Object Obj1 Obj2 Obj3
has weight wi > 0 and worth vi > 0. A thief can carry a Profit 25 24 15
maximum weight of W pounds in a knapsack. In this ver- Weight 18 15 10
sion of a problem, the items can be broken into smaller
pieces, so the thief may decide to carry only a fraction xi Solution:
of object i, where 0 ≤ xi ≤ 1. Item i contributes xiwi to
the total weight in the knapsack, and xivi to the value Step 1: Calculate ratio (profit/weight):
of the load. Object Obj1 Obj2 Obj3
In symbol, the fraction knapsack problem can be Profit 25 24 15
stated as follows.
Weight 18 15 10
Maximize nSi = 1xivi subject to constraint nSi = 1xiwi
≤ W. It is clear that an optimal solution must fill the
Ratio 1.38 1.6 1.5
Chapter 4.indd 196 4/9/2015 9:54:35 AM

4.7 GREEDY APPROACH 197
Step 2: Put according to decreasing ratio (profit/ 4.7.3.2 Pseudocode for Huffman Coding
weight):
HUFFMAN(f[1...n])
Object Obj2 Obj3 Obj1 1. T = empty binary tree
2. Q = priority queue of pairs (i, f[i]),
Profit 24 15 25
i = 1...n, with f as comparison key
Weight 15 10 18 3. for each k = 1...n − 1
Ratio 1.6 1.5 1.38 4. i = extractMin(Q)
5. j = extractMin(Q)
Step 3: Put items into the knapsack till the knapsack 6. f[n + k] = f[i] + f[j]
is full: 7. insertNode(T, n + k) with children i, j
8. insertRear(Q,(n + k, f[n + k]))
(a) Put object 2 into the knapsack; it will add (value = 9. return T
24 and weight = 15) to the knapsack. Knapsack has
5 units capacity remaining.
(b) Now object 3 cannot be put completely into the 4.7.3.3 Time Complexity
knapsack, so put a fraction of it. Add (value = 7.5
and weight = 5) to the knapsack. Knapsack is com- From the above pseudocode, we can see that with each
pletely full, which means no other object can be put iteration,the problem size is reduced by 1. Hence, there
into it. are exactly n iterations. The ith iteration consists of
locating the two minimum values in a list of length n −
Step 4: Total knapsack value = 31.5 (which is optimum).
i + 1. This is a linear operation, and so Huffman’s
algorithm clearly has a time complexityof O(n2).
4.7.3 Huffman Coding However, it would be faster to sort the weights ini-
tially, and then maintain two lists. The first list consists of
Huffman coding is a common technique of encoding, weights that have not been combined yet, and the second
includinglossless data compression. The algorithm’s list consists of trees that have been formed by combining
output can be viewed as a variable-length code table weights. This initial ordering is obtained at the cost of
for encoding a source symbol (such as a character in a O(n log n). Obtaining the minimum two trees at each step
file). Huffman coding is based on the frequency of occur- then consists of two comparisons (comparing the heads of
rence of a data item (pixel in images). The main aim the two lists, and then comparing the larger item with the
of Huffman coding is to encode the data that occurs item after the smaller). The ordering of the second list can
frequently using less number of bits. Code books store be maintained cheaply by using a binary search to insert
codes that are constructed for images. Both the code new elements. Since at step i there are i − 1 elements in
book and encoded data are transmitted to decode the the second list, O(log i) comparisons are needed for inser-
information. tion. Over the entire duration of the algorithm, the cost of
Huffman codes can be constructed using following two keeping this list sorted is O(n log n). Therefore, the overall
ideas: time complexity of Huffman’s algorithm is O(n log n).
1. For an optimum code, symbols that have high
probability should be assigned shorter code words. Example 4.17
2. To make optimum prefix code, two symbols that
In the following example five different symbols are given
occur least frequently should be assigned same
along with their frequency.
length.
Symbol Frequency
4.7.3.1 Principle of Huffman Codes
A 24
1. Starting with the two least probable symbols, g B 12
and d, of an alphabet A, if the code word for g is C 10
[m]0, the code word for d would be [m]1, where [m] D 8
is a string of 1s and 0s.
2. Now, the two symbols can be combined into a E 8
group, which represents a new symbol y in the
There are a total of 186 bits (with 3 bits per code word).
alphabet set.
3. The symbol y has the probability P(g ) + P(d ). Here, the least frequency symbols are E and D. First
Recursively, determine the bit pattern [m] using of all, we will connect them and frequency of node will
the new alphabet set. become 16. Then B and C are combined to make the
Chapter 4.indd 197 4/9/2015 9:54:36 AM

frequency 22. These will be two new nodes. In the next 3. At each step, an edge is added to the graph that
step, these two nodes are brought together and act as has minimum weight and is the neighbour to the
subordinate to the root node. A also acts as subordinate previous vertices.
to root node.
Steps of Prim’s Algorithm
1. Find the minimum weight edge from graph and
4.7.3.4 Code Tree According to Huffman
mark it.
2. Consider the neighboring edges of the selected
edges and mark the minimum weight edge among
those edges.
0 62 1 3. Continue this until we cover n-1 edges.
4. The resultant tree is minimum spanning tree and it
A
should always be connected.
24 0 38 1 Input: Undirected connected weighted graph G = (V, E )
in adjacency list representation, source vertex s in V and
w is an array representing weights of edges.
Output: p[1⋅⋅⋅|V |], representing the set of edges com-
0 22 1 16 1
0
posing an MST of G.
B C E
Note: The field p(v) names the parent of v in the tree.
D
12 10 8 8
Key(v) is the minimum weight of any edge connect-
ing v to a vertex in tree. Key(u)←∞ means there is no
Symbol Frequency Code Code Total such edge.
Length Length
PRIM (G, w, s)
A 24 0 1 24 1. for each u in V(G)
B 12 100 3 36 2. do key(u) ∞
3. π(v) NIL
C 10 101 3 30
4. key [r] 0
D 8 110 3 24 5. Q V [G]
E 8 111 3 24 6. while Q != empty
7. do u Extract-Min(Q)
8. for each v in Adjacent[u]
9. do if v ∈ Q and w(u, v) < key [u]
The total length is 138 bits.
4.7.4 Minimum Spanning Tree 10. then π[v] u

11. key [v] w(u, v)
Spanning tree is a sub-graph that connects all the verti-
ces together. A single graph can have multiple spanning Time Complexity
trees. Minimum spanning tree for a weighted undirected
graph is the spanning tree with minimum total weights. Prims algorithm use min heap data structure and its
The weight of spanning tree is addition of weights of all time complexity is O(VlogV+E logV ).
the edges in the spanning tree. An MST with N vertices, Maximum edges in tree, E = V − 1
has N − 1 edges for a given graph. Therefore, O(VlogV + (V − 1)logV) = O(2VlogV − logV)
4.7.4.1 Prim’s Algorithm Thus, complexity = O(VlogV )
Prim’s algorithm is a method to find minimum spanning Problem 4.5: Use Prim’s algorithm to find an MST in
tree. In this algorithm minimum spanning tree formed the following weighted graph. Use alphabetical order
should always be connected. Prim’s algorithm is an to break ties.
example of Greedy approach. b 5 d
Properties of Prim’s Algorithm 2 2
Prim’s algorithm has the following properties: a 6 3 1 z

1. Prime’s algorithm always results into a single tree. 3 4
2. Tree grows until it covers all the vertices of the
c 2 e
graph.
Chapter 4.indd 198 4/9/2015 9:54:37 AM

KRUSKAL(V, E, w)
1. A ¬ {φ} Set A will ultimately contain the
edges of the MST 4.7 GREEDY APPROACH 199
2. For each vertex v in V[G]

Solution: Prim’s algorithm will proceed as follows. 3. Do MAKE-SET(v)
4. Sort E into non-decreasing order by weight w
1. First we add edge {d, e} of weight 1.
5. For each edge (u, v)∈ E, taken in

2. Next, we add edge {c, e} of weight 2.
non-decreasing order by weight w
3. Next, we add edge {d, z} of weight 2.
4. Next, we add edge {b, e} of weight 3. 6. do if FIND-SET(u) ≠ FIND-SET(v)
5. And finally, we add edge {a, b} of weight 2. 7. Then A ¬ A ∪ {(u, v)}
8. UNION(u, v)
This produces a MST of weight 10.
9. Return A
b d
2 2 Time Complexity
a 1 z The edge weight can be compared in constant time.
3
Initialization of priority queue takes O(E log E ) time
by repeated insertion. At each iteration of while loop,
minimum edge can be removed in O(log E ) time, which
c 2 e
is O(log V ), since the graph is simple. The total running
time is O((V + E ) log V ), which is O(E log V ) since the
4.7.4.2 Kruskal’s Algorithm graph is simple and connected.
One another algorithm to find the minimum spanning Example 4.18

tree is Kruskal’s. This method considers all minimum
Let us use the Kruskal’s algorithm to solve the following
edges of graph one by one in increasing order and gives
example.
a resultant minimum spanning tree. In Kruskal’s algo-
rithm, it is not necessary for a tree to be connected. b 8 c 7 d
Implementation steps: 4 2 9
1. Start from the source node. i e
a 4 14
2. Scan the weights and include the minimum weights
7 6
to the tree in increasing order.
8 10
3. Stop if N − 1 edges are included.
h g f
Uses a disjoint-set data structure to determine whether 1 2
an edge connects vertices in different components.
Step 1: Take only the vertices of graph.
Data Structure
b c d
Before formalizing the above idea, let us quickly under-
stand the disjoint-set data structure:
1. Make-SET(v): It creates a new set whose only i
a e
member is pointed to by v. Note that for this oper-
ation, v must already be in a set.
2. FIND-SET(v): It returns a pointer to the set
containing v. h g f
3. UNION(u, v): It unites the dynamic sets that
contain u and v into a new set that is a union of Step 2: Choose the minimum weight edge from the
these two sets. original graph and draw that edge in disconnected graph
containing n connected components as shown below:
Algorithm
Starting with an empty set A, select the shortest edge b c d
that has not been chosen or rejected at every step,
regardless of the position of this edge in the graph. The
pseudocode for Kruskal algorithm is given: a i e
KRUSKAL(V, E, w)
1. A ¬ {φ} Set A will ultimately contain the
edges of the MST h g f
2. For each vertex v in V[G] 1
3. Do MAKE-SET(v)
4. Sort E into non-decreasing order by weight w
5. For each edge (u, v)∈ E, taken in
non-decreasing order by weight w
Chapter 4.indd6.
199 do if FIND-SET(u) ≠ FIND-SET(v) 4/9/2015 9:54:39 AM
Step 3: Now select the next minimum weight edge and 2. Calculate the temporary distance of neighbours of
draw the corresponding edge in disconnected graph. active nodes. Distance is calculated by adding up
the weights associated with those edges.
b c d 3. Update the distance, and set the current node an
2 antecessor if such a calculated distance of a node is
smaller as the current one. This step is also called
i update and is Dijkstra’s central idea.
a e
4. Set the node as active node which has minimal tem-
porary distance. Mark the distance as permanent.
5. Repeat steps 2 to 4 until there are no nodes left
h g f with a permanent distance, whose neighbours still
1 2 have temporary distances.
Here, there are two edges containing weight 2, add all
possible edges of same weight until they form loop in the Dijkstra’s Pseudocode
tree (since tree doesnot contain any loop but a graph
DIJKSTRA (G , S):
may contain). That is why edge(g,f) and edge(i,c) both 1. for each vertex v in Graph:
are added. 2. distance[v] := infinity
Step 4: Continue this process and check for the next 3. previous[v] := undefined
smallest edge and add it into tree. Do this till we get 4. distance[ S ] := 0
(n − 1) edges in the tree. 5. Q := the set of all nodes in Graph
6. While Q is not empty:
The final minimum spanning tree would be: 7. u := node in Q with smallest distance[ ]
8. remove u from Q
b c d 9. for each neighbour v of u:
10. alt := distance[u] + distance_between(u, v)
11. If alt < distance[v]
i e 12. distance[v] := alt
a
13. previous[v] := u
14. Return previous[ ]
h g f
Time Complexity
Like Prim’s algorithm, Dijkstra’s algorithm runs in O(E
4.7.5 Single-Source Shortest Path log V) time.
1. Dijkstra’s algorithm with list: O(V2)
Given a weighted graph G, find the shortest path from 2. Dijkstra’s algorithm with modified binary heap:
a given vertex to every other vertex in G. Note that we O((E + V) log V )
can solve this problem quite easily with BFS traversal in 3. Dijkstra’s algorithm with Fibonacci heap: O(E +
the special case when all weights are equal. The Greedy V log V )
approach to this problem is repeatedly selecting the best
choice from those available at that time. Example
The following are the steps:
1. Start from the source node; find the distance of all
4.7.5.1 Dijkstra’s Algorithm
nodes reachable from the source.
Dijkstra’s algorithm is a single source shortest path algo- 2. Then pick the node which has the minimum dis-
rithm. It calculates the shortest distance starting from tance, from the pool of nodes. Source
the start node till another node in the graph. While cal- 3. Calculate again the distance from the selected A B C
culating the shortest distance, it excludes the longer dis- node, and follow the above two steps till the entire ∞ ∞ ∞ ∞
tance paths. Dijkstra’s algorithm does not operate on nodes are selected. Nil Nil N
negative weights. For calculating shortest distance with O 10 3 ∞
2 A Nil A A N
negative edges, Bellman-Ford algorithm is used. Steps to B D
10 O 4
perform Dijkstra’s algorithm are as follows:
1 4 8 7 9 AC Nil C
1. Assign all the nodes with distance infinity and 0 to A
3 O
initial node.Distance of start node s permanent and Nil
3 ACB
of all the other nodes is temporary.Set start node
C E O
as active node.
ACBD Nil
Chapter 4.indd 200 4/9/2015 9:54:40 AM

4.8 GRAPH TRAVERSAL 201
Source to know that the node has been explored before or not.
A B C D E The re-traversal becomes more for dense graph, which
∞ ∞ ∞ ∞ ∞ Cost increase computation time. Tree traversal is a special
Nil Nil Nil Nil Parent case of graph traversal. The following are two graph
O 10 3 ∞ ∞ traversal methods:
2 A Nil A A Nil Nil
B D 1. Breadth-first traversal (BFT)
Source O 4 11 6 2. Depth-first traversal (DFT)
A 4 B8 C 7 D 9 E AC Nil C C C
2
∞ ∞ ∞ ∞ ∞ Cost O 6 6 B D
4.8.1 Breadth-First Traversal
Nil Nil Nil Nil ParentNil
ACB B C A 1
CO 10 3 ∞E ∞
3
O 6 3 C The breath-first
E traversal starts from the root node and
Nil A A Nil Nil ACBD Nil 3 traverses all its neighbours. Then for each neigh-
C then
O 4 11 6 bour, its corresponding unvisited neighbour is processed.
Nil C C C
2 The pseudocode for BFT is given as follows:
O 6 6 B D
Nil B C A 1 BST (V)
1. visited(V) =1
O 6 3 C
3
E 2. add(V, Q)
Nil C 3. while(Q is not empty)
4. u = delete(Q)
4.7.5.2 Bellman—Ford Algorithm 5. for all x adjacent to u
6. if( x is not visited)
The Bellman—Ford algorithm is used to compute 7. visited(x) = 1
single source shortest path to all other vertices of 8. add(Q, x)
weighted directed graph. This algorithm also consid-
ers negative weights of graph edges. This algorithm 4.8.1.1 Time Complexity
is slower but is capable of even finding a negative
weight cycle in graph. The time complexity of BFT is O(V + E), where V is
the number of vertices in the graph and E is the number
Bellman-Ford Pseudocode of edges in the graph.
1. d[s] ← 0 4.8.1.2 Application of BFT

2. for each v V – {s}
3. do d[v] ← ∞ The following are the applications of BFT. It can be
4. for i ← 1 to |V| – 1 do used to find out
5. for each edge (u, v) E do
6. if d[v] > d[u] + w(u, v) then 1. whether the given graph is connected.
7. d[v] ← d[u] + w(u, v) 2. the number of connected components in a graph.
8. [v] ← u 3. whether the given graph contains a cycle.
9. for each edge (u, v) E 4. the shortest path if all the edges have the same weight.
10. do if d[v] > d[u] + w(u, v) 5. whether the given graph is a bipartite graph.
11. then report that a negative-weight
cycle exists
Problem 4.6: Consider the graph G = (V, E ) and
find the BFT starting node A.
Time Complexity
Graph (G) A
The time complexity of Bellman-Ford algorithm is
O(VE), where E = V2, so the worst-case running time of B C
Bellman—Ford algorithm is O(V3).
D E F G
4.8 GRAPH TRAVERSAL H I
J
Graph traversal means visiting all the nodes in a graph in
some order. In graph traversal, some nodes may be vis- Solution: A is the starting node and Q is a queue data
ited more than once. This is because it is not necessary structure to store the values.
Chapter 4.indd 201 4/9/2015 9:54:41 AM

BFT (A): 10. u = delete (Q) → I

1. Visited (A) = 1 For all adjacent nodes to I which are not already
Add (Q, A) added in the queue, Add (Q, X), where X is the
adjacent node to I.
A
J
2. u = delete (Q) → A
11. u = delete (Q) → J
For all adjacent nodes to A which are not already
For all adjacent nodes to J which are not already
added in the queue, Add (Q, X), where X = (B, C)
added in the queue, Add (Q, X), where X is the
is the adjacent node to A.
adjacent node to J.
B C
3. u = delete (Q) → B

12. The queue is empty, which means graph is tra-
For all adjacent nodes to B which are not already
versed completely.
added in the queue, Add (Q, X), where X = (D, E)
is the adjacent node to B.
C D E
4.8.2 Depth-First Traversal
4. u = delete (Q) → C
As already discussed, there are many ways of tree tra-
For all adjacent nodes to C which are not already
versal. Consider the tree given in Fig. 4.10:
added in the queue, Add (Q, X), where X = (F, G)
is the adjacent node to C.
Tree
D E F G
5. u = delete (Q) → D I Root
For all adjacent nodes to D which are not already
added in the queue, Add (Q, X), where X = (H) is
the adjacent node to D. F K
E F G H
6. u = delete (Q) → E A H Z
For all adjacent nodes to E which are not already
added in the queue, Add (Q, X), where X = (F) is
the adjacent node to E. D
F G H Figure 4.10 | DFT of a tree.
7. u = delete (Q) → F
For all adjacent nodes to F which are not already Preorder traversal for given tree: I F A D H K Z
added in the queue, Add (Q, X), where X = (I) is Preorder traversal also results in depth first traversal.
the adjacent node to F. This is because the traversal goes deeper before travers-
G H I ing its siblings. In this tree, the descendants of F, that
is, A, H, D are traversed first and after that sibling of F,
8. u = delete (Q) → G that is, K are traversed.
For all adjacent nodes to G which are not already
added in the queue, Add (Q, X), where X = (I) is
the adjacent node to G. 4.8.2.1 Pseudocode for DST
H I DFT (V)
1. visited (V) =1
9. u = delete (Q) → H 2. for all x adjacent to V
For all adjacent nodes to H which are not already 3. if (x is not visited)
added in the queue, Add (Q, X), where X = (J) is 4. visited (x) =1
the adjacent node to H. 5. DFT (x)
I J
The above is the recursive code for DFT.
Chapter 4.indd 202 4/9/2015 9:54:42 AM

4.9 DYNAMIC PROGRAMMING 203
4. POP (S) → D
The time complexity of DFT is O(V + E ), where V is For all adjacent nodes to D which are not already
the number of vertices in the graph and E is the number added in the stack, PUSH (S, X). Every neighbour
of edges in the graph. element will be stored in the stack from right to left.
Here, X = (H) is the adjacent node to D.
4.8.2.3 Application of DFT H E C
The following are the applications of BFT. It can be 5. POP (S) → H
used to find out For all adjacent nodes to H which are not already
added in the stack, PUSH (S, X). Every neighbour
1. whether the given graph is connected. element will be stored in the stack from right to left.
2. the number of connected components in a graph. Here, X = (F, G) is the adjacent node to H.
3. whether the given graph contains a cycle.
4. whether the given directed graph is strongly F G E C
6. POP (S) → F
connected.
For all adjacent nodes to F which are not already
Problem 4.7: Consider the graph G = (V, E) and find added in the stack, PUSH (S, X). Every neighbour
the DFT starting node A. element will be stored in the stack from right to left.
Here, X = (F) is the adjacent node to F.
Graph(G) A
G E C
7. POP (S) → G
B C For all adjacent nodes to G which are not already
D E F G element will be stored in the stack from right to left.
Here, X = (F) is the adjacent node to G.
H E C
8. POP (S) → E
Solution: A is the starting node and S is a stack data For all adjacent nodes to E which are not already
structure to store the values. Stack is internally used added in the stack, PUSH (S, X). Every neighbour
by the recursive call. element will be stored in the stack from right to left.
Here, X = (F) is the adjacent node to E.
DFT (A)
1. Visited (A) =1 C
PUSH(S, A) 9. POP (S) C
A For all adjacent nodes to C which are not already
2. POP(S) → A element will be stored in the stack from right to left.
For all adjacent nodes to A which are not already Here, X = (F) is the adjacent node to C.
element will be stored in the stack from right to left.
Here, X = (B, C) is the adjacent node to A. 10. The stack is empty, which means the graph is
traversed completely.
B C
3. POP(S) → B
For all adjacent nodes to B which are not already 4.9 DYNAMIC PROGRAMMING
element will be stored in the stack from right to left.
Here, X = (D, E) is the adjacent node to B. Dynamic programming is a method for solving complex
problems by breaking them down into simpler sub-prob-
D E C lems. It is applicable to problems exhibiting the proper-
ties of overlapping sub-problemsand optimal substructure.
Chapter 4.indd 203 4/9/2015 9:54:43 AM

Dynamic programming algorithms are used for optimiza- The total running time = O(2n).
tion (e.g. finding the shortest path between two points, or Note: Most of the recursive programs contain very less
the fastest way to multiply many matrices). A dynamic number of distinct calls and more number of repeated sub-
programming algorithm will examine all possible ways to problems or function calls. These problems are known as
solve the problem and will pick the best solution. overlapping sub-problems.
4.9.1 Applications of Dynamic Programming

4.9.2.2 Dynamic Solution of Fibonacci
The following are the applications of dynamic pro‑
gramming: In dynamic programing technique, we solve every problem
1. Fibonacci series exactly once and store the results in memory. That result
2. 0/1 knapsack is used later to get an optimum output (Fig. 4.11).
3. Largest common subsequence
Fib (6)
4. All-pair shortest path
4.9.2 Fibonacci Series Fib (5) 4
Fibonacci series is the sequence produced by adding the

previous two numbers in the sequence. First two num- Fib (4) 3
bers in sequences are 0 1 or 1 1. The recurrence relation
Array 0 1 1 2 3 5 8
to find the next term in the sequence is given below:
Fib (3) 0 1 2 3 4 5 6
Fn = Fn -1 + Fn -2 2
with initial values F1 = 0 and F2 = 1 or F1 = 1 and

F2 = 1. Fib (2) 1
Example: Fibonacci series 0, 1, 1, 2, 3, 5, 8, 13, 21, ….

Fib (1) Fib (0)
Figure 4.11 | Dynamic programming approach

4.9.2.1 Recurrence Relation of Fibonacci
0, if n < 0 for Fibonacci.

fib(n) = 1, if n = 1

fib(n - 1) + fib(n - 2), if n > 1
Pseudocode for Dynamic Solution of
Fibonacci Series
1. Algorithm Fast-Fibonacci(n)
Example 4.19 2. Let fib[0] and fib[1] be 1
3. for each i from 2 to n, do
The solution of Fibonacci through recursion tree is
4. Let fib[i] = fib[i - 2] + fib[i - 1]
shown as follows: 5. end of loop
Fib(6) 6. return fib[n]
Time Complexity
Fib(5) Fib(4)
1. Fibonacci series without dynamic programing
[space = O(n) and time = O(2n)]
Fib(4) Fib(3) Fib(3) Fib(2) 2. Fibonacci series with dynamic programing [space =
Level 6 O(n) and time = O(n)]
Fib(3) Fib(2) Fib(2) Fib(1) Fib(2) Fib(1)Fib(1) Fib(0) 4.9.3 0/1 Knapsack
One of the problems in combinatorial optimization is

k napsack problem. In this problem, the target is to
m aximize the profit by choosing items. Each item is
a ssociated with a weight and profit. The collection
Total number of nodes = (2n −1). of items is formed such that the chosen total weight
So, the total number of function calls = 2n, and one should be less than or equal to the given limit. The
function call takes O(1) time. profit can be as large as possible.
Chapter 4.indd 204 4/9/2015 9:54:44 AM

4.11 CONCEPTS OF COMPLEXITY CLASSES 205
4.9.3.1 Recurrence Relation 0/1 Knapsack 4.9.4.3 Time Complexity



The recurrence 0, relation of 0/1 Knapsack problem is = 0 or n =Complexity
if m 0 LCS with LCS with


KS(n, m) = KS(n - 1, m) if m < w[n ]
Recursive Dynamic
 Solution Programing
Max[{KS(n - 1, m - w[n ] + p[n ])} otherwise
  Time If (m = n) T(n) = O(mn)
  {KS(n - 1, m) + 0}], LCS (m, n) = O(2n)
0, if m = 0 or n = 0

KS(n, m) = KS(n - 1, m) if m < w[n ] Else
 LCS(m, n) = O(2m+n)
Max[{KS(n - 1, m - w[n ] + p[n ])} otherwise
 Space O(m + n) O(mn)
 {KS(n - 1, m) + 0}],
4.10 ALL-PAIR SHORTEST PATH

The time complexity of the knapsack problem is as follows: Given a directed graph G = (V, E ), where each edge (v, w)
1. Without dynamic programing = O(2n) has a non-negative cost C[v, w], for all pairs of vertices (v,
2. With dynamic programing = O(mn) w), find the cost of the lowest-cost path from v to w. It is
a generalization of the single-sourceshortest-path problem.
It is one of the NPC problems.
4.10.1 Floyd’s Algorithm
4.9.4 Longest Common Subsequence
Let Ak(i, j) be the minimum cost required to go from
A subsequence of a given sequence is just the given sub- vertex i to vertex j for which all intermediate vertices are
sequence in which zero or more symbols are left out. in the set (0, 1, 2, 3, …, k) (Fig. 4.12).
Let sequence (S) = {A, B, B, A, B, B}, subsequence
1
(S1) = {A, A}, subsequence (S2) = {A, B, A, B} and
subsequence (S3) = {B, A, B, A}. 2
i j
4.9.4.1 Common Subsequence
In the given two sequences X and Y, we say that a k
Figure 4.12 | Different paths from i to j.

sequence Z is a common subsequence of X and Y, if Z is
a subsequence of both X and Y.
4.10.1.1 Recurrence Relation of Floyd’s

Example 4.20 Algorithm
X = {A, B, B, A, B, B}
The recurrence relation of Floyd’s algorithm is
Y = {B, A, A, B, A, A}
Ak (i , j ) = Min{Ak -1(i , j ), Ak -1(i , k) + Ak -1(k, j)}
Z = {B, B, A} is a common subsequence. But Z = {A,
B, A, B} is not a common subsequence. 4.10.1.2 Time Complexity
The running time complexity: T(n) = O(V 3).

4.9.4.2 Recurrence Relation of LCS
4.11 CONCEPTS OF COMPLEXITY
The recurrence relation of longest common sub‑
sequence is
CLASSES
LCS(i, j)
4.11.1 P-Complexity Class
0, if i = 0 or j = 0

= 1 + LCS(i - 1, j - 1), if X[i] = Y [ j ] P in computational complexity theory is known as poly-
 nomial time or deterministic time. This is the most basic
Max[{LCS(i - 1, j)}{LCS(i, j - 1)}], if X[i] ≠ Y [ j ]
 complexity class. This consists of all the problems that
Chapter 4.indd 205 4/9/2015 9:54:46 AM

are solvable in polynomial time. These problems can be 5. Subgraph isomorphism problem
solved in O(nk) time, where k is some constant. In class 6. Subset sum problem
P, problems are efficiently solvable or traceable. 7. Clique problem
A language L is in class P iff there exists some Turing 8. Vertex cover problem
machine M such that: 9. Independent set problem
10. Dominating set problem
1. Turing machine runs in polynomial time for all the 11. Graph colouring problem
inputs.
2. For all y in L, M must return 1, otherwise 0. 4.11.4 NP-Hard
P includes many natural problems like linear programming,
finding maximum matching and calculating the greatest NP-hard is a class that includes problems which are “at
common divisor. Prime number is also a part of class P now. least as hard as hardest problem in NP”. A problem is
said to be NP hard if
4.11.2 NP Complexity Class 1. It is in NP

2. It is NP-complete
NP class consists of problems that are verifiable in poly- This means a problem H is NP-hard if any problem L
nomial time. This class contains those languages that has polynomial time reduction from L → H. The prob-
can be defined by second order logic. lem can be a decision problem, optimization problem or
Note: The most important open question in complexity a search problem.
theory is the P = NP problem (Fig. 4.13). Examples of NP-hard problem is the traveling sales-
man problem. It is an optimization problem of finding
NP-hard NP-hard the least-cost cyclic route through all nodes of a weighted
graph. There are decision problems that are NP-hard
NPC but not NP-complete, for example, the halting problem.
This is the problem which asks, `given a program and its
NP
P = NP = NPC
input, will it run forever?’ That is a yes/no question, so
this is a decision problem.
P
Note:
1. Let À’ be an NP-complete problem and `B’ be
P! = NP P = NP
an unknown problem; if À’ is a polynomial that
Figure 4.13 | Relation between P, NP, NP-complete reduces to `B’ (A ≤p B), then `B’ is NP-hard.
and NP-hard problems. 2. Let À’ be an NP-hard problem and `B’ be an
unknown problem; if À’ is a polynomial that
4.11.3 NP-Complete reduces to `B’ (A ≤p B), then `B’ is NP-hard.
3. Let À’ be a P-class problem and `B’ be an unknown
These are the hardest problems in NP. In the computa- problem; if (B ≤p A), then `B’ is a P-class problem.
tional complexity theory, the complexity class NP-complete 4. Let À’ be a P-class problem and `B’ be an unknown
(abbreviated NP-C or NPC) is a class of decision problems. problem; if (B ≤p A), then `B’ is an NP-class
A decision problem L is NP-complete if it is in the set of problem.
NP problems and also in the set of NP-hard problems.
NP-complete is a subset of NP, the set of all deci- Problem 4.8: Both P and NP are closed under the
sion problems whose solutions can be verified in polyno- operation of:
mial time; NP may be equivalently defined as the set of (A) Union (B) Intersection
decision problems that can be solved in polynomial time (C) Kleene’s closure (D) None of these
on a non-deterministic TM. A problem p in NP is also
Solution: (D) Both P and NP are closed under con-
NP-complete if every other problem in NP can be trans-
catenation and Kleene’s closure. The Kleene closure
formed into p in polynomial time.
of a language L, denoted by L* is defined as
The following list contains some well-known problems
{w1w2…wk−k∈N, wi ∈ L∀i = 1, 2, …, k}.
that are NP-complete when expressed as decision problems:
P is closed under Kleene’s closure as for all L ∈ P,
1. Boolean satisfiability problem (SAT)
L* ∈ P.
2. Knapsack problem
3. Hamiltonian path problem NP is closed under Kleene’s closure as for all L ∈ NP,
4. Travelling salesman problem L* ∈ NP.
Chapter 4.indd 206 4/9/2015 9:54:46 AM

SOLVED EXAMPLES 207
IMPORTANT FORMULAS
1. Theta notation (Q): 0 < c1(g(n)) ≤ f (n) ≤ c2 (g(n)) ∀n ≥ 11.

n0 . Insertion sort: O(n2).
(n)) ≤ f (n) ≤ c2 (g(n)) ∀n ≥ n0 .
12. Heap sort: O(n log n).
2. Big-O notation (O): 0 < f (n) ≤ c(g(n)) ∀n ≥ n0. 13. Merge sort: O(n log n).
3. Omega notation (W): 0 < c(g(n)) ≤ f (n) ∀n ≥ n0 .
14. Quick sort: O(n2) (worst case).
4. Small-o notation (o): 0 < f (n) < c(g(n)) ∀n ≥ n0 .
15. Quick sort: O(n log n) (average case).
5. Small omega notation (w): 0 < c(g(n)) < f (n) ∀n ≥ n0 . 16. Counting sort: O(n + k).
6. Master theorem: T (n) = aT (n/b) + f (n) . 17. If a graph has n number of vertices and it is
log b a -e k-regular, where k is a positive constant, then the
Case 1: f (n) = O(n ), then T (n) = Θ(n logb a ). number of edges = (n × k)/2.
Case 2: f (n) = Θ(n ), then T (n) = Θ(n logb a log n).

log b a
18. A complete graph having n vertices has
[n (n − 1)]/2 edges.
Case 3: f (n) = Ω(n logb a + e ).
19. Number of simple graphs possible with n vertices
= 2[n(n − 1)]/2.
af(n/b) ≤ cf(n) for some constant c < 1 and large n,
then T(n) = Q(f(n)). 20. Prim’s algorithm: O(E log V).
7. Linear search: O(n). 21. Kruskal’s algorithm: O((V + E) log V).
8. Binary search: Q(log n). 22. Dijkstra’s algorithm: O(E log V).
9. Bubble sort: O(n2). 23. Bellman—Ford algorithm: O(V3).
10. Selection sort: O(n2). 24. BFT and DFT: O (V + E).
SOLVED EXAMPLES
1. Consider T (n) = 6 T (n/2) + n3; then T(n) is equal to Solution:

3 T(n) = 4T(n/2) + n; By using Master Theorem.
(a) (n/2) (b) O(n log n)
(c) (n2 log n) (d) O(n3) a = 4, b = 2, f(n) = n
g(n) = (n logb a ) = n log2 4 = Θ(n2)
Solution: f (n) = O(n log2 4-e ), f(n) is smaller than g(n), but
T(n) = 6T(n/2) + n3; By using Master Theorem. is not a smaller polynomial.
a = 6, b = 2, f(n) = n3
log b a log 2 6
By case 1, T(n) = Θ(n2)
2.58
g(n) = (n ) = n = Θ(n )
Ans. (b)
log 2 6+ e
f(n) = W (n ) , f(n) is larger than g(n), but is
not a larger polynomial. 3. Consider the following statements:
3
By case 3, T(n) = Θ(f (n)) = n 1. An algorithm is the number of steps to be per-
Ans. (d)
formed to solve a problem.
2. An algorithm is the number of steps and their
2. Solve the following recurrence relation: T (n) = 4T (n/2) + n
implementation in any language to a given prob-
T (n) = 4T (n/2) + n.
lem to solve a problem.
(a) O(n2 log n) (b) O(n2) 3. To solve a given problem there may be more
(c) O(n/2) (d) O(log n) than one algorithm.
Chapter 4.indd 207 4/9/2015 9:54:51 AM

Which of the following is true? 9. The time complexity of searching an element in a

linked list of length n will be
(a) 1 is correct. (b) 2 is correct.
(c) 1 and 3 are correct. (d) 2 and 3 are correct. (a) O(n) (b) O(n2 − 1)
(c) O(log n) (d) O(log n2)
Solution: Both statements 1 and 3 are correct.
Ans. (c) Solution: The worst-case time complexity for
searching an element in a linked list is O(n), so
4. Compilers require efficient searching strategies for
option (a) is correct. If in the question it is not
which they rely on
mentioned about (best or worst or average) time
(a) Hash tables (b) String matching complexity then we always calculate as worst time
(c) Binary search tables (d) Binary search trees complexity.
Ans. (a)
Solution: Hash tables have less complexity, so
they are used by the compiler for efficient searching. 10. A hashing algorithm uses linear probing; the aver-
Ans. (a) age search time will be less if the load factor
5. Consider a binary search tree having n elements. (a) is less than one (b) equals one
The time required to search a given element will be: (c) is greater than one (d) equals two
(a) O(n) (b) O(log n) Solution: As load time is the ratio of number of
(c) O(n2) (d) O(n log n) records present to the total number of records, so is
load time is less than 1, the probability of collisions
Solution: The time required for search operation
decreases and hence, search time reduces.
in a binary search depends on the height of the
Ans. (a)
tree. In a complete binary tree with n nodes, search
operation takes Θ (log 2 n) in worst case scenario. 11. If a node has K children in B tree, then the node
But if all the nodes of tree are in linear chain, then contains exactly keys.
(a) K2 (b) K −1 (c) K + 1 (d)
the same operations will take O(n) in worst-case
K
scenario. So, option (a) is correct.
Ans. (a) Solution: The number of keys = K − 1.
Ans. (b)
6. The average time required to perform a success-
ful sequential search for an element in an array 12. Big-O estimate for the factorial function, that is, n!,
A(1: n) is given by is given by
(a) log n (b) n2 (a) O(n!) (b) O(nn)
(c) (n + 1)/2 (d) (n ×n − 1/2) (c) O(n ×n!) (d) O( n )
Solution: The average time required for sequen- Solution: Factorial function f(n) = n! = n(n − 1) ×
tial search or linear search in this case is (n + 1)/2. (n − 2) … 1
Ans. (c) n! = n (n − 1)(n − 2) … 1 ≤ n × n … × n = (nn)
7. Which sort will operate in quadratic time relative So, Big O Estimation of f (n!) = O(n n)
to the number of elements in the array (on the Ans. (b)
average)? 13. Let A and B be two n × n matrices. The efficient
(a) Heap sort (b) Bubble sort algorithm to multiply the two matrices has the
(c) Radix sort (d) Merge sort time complexity O(nx), where x is ––––.
Solution: The average time for bubble sort is

Solution: Strassen’s surprising algorithm can mul-
O(n2), so it will operate in quadratic time.
tiply two n × n matrices in Θ(nlog 7) time, which is
Ans. (b)
equal to O(n2.81) time.
8. Binary expression tree is traversed in –––– Ans. (2.81)
traversal.
14. The searching technique in which less number of
(a) Postorder (b) Preorder comparisons are done is
(c) Both A and B (d) Reverse polish order
(a) Binary searching
Solution: Binary tree is traversed in both preor- (b) Linear search searching
der as well as postorder traversal. (c) Hashing
Ans. (c) (d) Woline’s search
Chapter 4.indd 208 4/9/2015 9:54:53 AM

SOLVED EXAMPLES 209
Solution: 18. Solve the following given recurrence relation:

T (n) = 4T (n/2) + n2.
Search Complexity Average Case Worst Case
(a) Q(n2) (b) Q(n2 log2 n)
Binary O(log n) O(n)
(c) Q(n log2 n) (d) None of the above
Hash table O(1) O(n)
Solution: Using the master theorem, we have
Linear search O(n) O(n) a = 4, b = 2 and f(n) = n2. So
Thus, hash table takes less search time. g(n) = n log b⋅a = n log 2⋅4 = n2
Ans. (c) Here, order of f(n) = order of g(n), so we get
15. Which of the following statements is false about T (n) = O(n log b⋅a log2 n) = O(n2 log2 n)
Prim’s algorithm? Ans. (b)
(a) The time complexity is O(E log V) using a 19. Solve the following given recurrence relation:
binary heap. T (n) = 16 T (n/4) + n3.
(b) It may use binomial max-heap to represent the
priority queue. (a) Q(n log2 n) (b) Q(n2 log2 n)
(c) The time complexity is O(E + V log V) using (c) Q(n3) (d) None of the above
a Fibonacci heap. Solution: From the master theorem, we have
(d) Initially the roots key and nodes are set to zero. a = 16, b = 4 and f(n) = n 2. So
Solution: g(n) = n log b⋅a = n log 4⋅16 = n2
(a) True, The time complexity is O(E log V) using
a binary heap. Order of f(n) > order of g(n), so we get
(c) True, Prim’s algorithm’s time complexity is T (n) = O(f (n))
O(E + V log V) using a Fibonacci heap. Ans. (c)
(d) True, Initial distance to the roots key and
nodes is kept Zero. 20. Consider the following two functions:
n3 , for 0 ≤ n ≤ 10, 000
g1(n) =  3
Only option (b) is false.
Ans. (b) n , for 0 ≤ n ≥ 10, 000

n, for 0 ≤ n ≤ 10, 000
g2 (n) =  3
16. Express the following recurrence relation in asymp-
totic notations: n , for n > 10, 000
n
T (n) = 2T   + n
 2 
Which of the following is true?
(a) g1(n) is O(g2 (n)). (b) g1(n) is O(n3 ).
(a) Q(n log n) (b) Q(n log2n) (c) g2 (n) is O(g1(n)). (d) g2 (n) is O(n).
(c) Q(n2) (d) O(n3)
Solution: Looking at the following figure:
Solution: By using Master Theorem,
g2(n) = n3
T(n) = 2T(n/2) + n
a = 2, b = 2, f(n) = n
g(n)= (n logb a ) = n log2 2 = Θ(n) g1(n) = n3
Order of g(n) = order of f (n), so by case 2, T(n) = g1(n) = n2
Θ(f (n)⋅log n) = Θ(n log n)
Ans. (a)
g2(n) = n3
17. What is the height of n-node, k-ary heap?
(a) Q(logk n) (b) Q(nk) 100 10,000
(c) Q(k log2 n) (d) Q(n/k)
g2(n) = n
Solution: If h is the height of k-ary heap (h ≥ 2)
H, then the number of nodes `n’ in H such that Therefore, we get
k(h) - 1 k(h + 1) - 1
<n≤ n2 ≤ n3 for N ≥ 10, 000
k -1 k -1
So h = Q(logk n). g1(n) = O(g2 (n))
Ans. (a) Ans. (a)
Chapter 4.indd 209 4/9/2015 9:54:58 AM

1. Consider the following three claims: Which are depth-first traversals of the above graph?
(I) (n + k) = Q(n ),where k and m are constants
m m
(a) I, II and IV only (b) I and IV only
(II) 2n+1 = O(2n) (c) II, III and IV only (d) I, III and IV only
(III) 22n+1=O(2n) (GATE 2003: 1 Mark)
Which of these claims is correct?
Solution: Option II is incorrect. There is no edge
(a) I and II (b) I and III between f and e.
(c) II and III (d) I, II and III Ans. (d)
(GATE 2003: 1 Mark)
4. The usual Q(n2) implementation of insertion sort
Solution: Option I: Consider k = 1; the expansion to sort an array uses linear search to identify the
of (n + 1)m = f(n). position where an element is to be inserted into the
Option II: 2n+1 = 2.2n = O(2n). already sorted part of the array. If, instead, we use
Option III: 22n+1 = 2.22n = O(22n). binary search to identify the position, the worst-
Options I and II are correct. case running time will
Ans. (a)
(a) remain Q(n2) (b) become Q(n(log n)2)
2. Let G be an arbitrary graph with n nodes and k (c) become Q(n log n) (d) become Q(n)
components. If a vertex is removed from G, the (GATE 2003: 1 Mark)
number of components in the resultant graph must
necessarily lie between Solution: Worst-case complexity will remain O(n2).
This is because binary search is used to locate
(a) k and n (b) k − 1 and k + 1 the position of element will take log2(n) time, but
(c) k − 1 and n − 1 (d) k + 1 and n − k inserting an element at its desired position will
take O(n) time in worst case.
Solution: Minimum components comes out to be Ans. (a)
when the independent vertex, which itself is a com-
ponent, is removed. In that case, the number of 5. In a heap with n elements with the smallest element at
components left is k − 1. the root, the 7th smallest element can be found in time
(a) Q(n log n) (b) Q(n)
The maximum range comes when the removed
(c) Q(log n) (d) Q(1)
vertex isolates all the vertices, In that case, the
number of components comes out to be n − 1.
(GATE 2003: 1 Mark)
Ans. (c)
Solution: Minimum element as root means it is
3. Consider the following graph:
min heap. Searching 7th minimum element takes
O(log n) time.
a Ans. (c)
6. It has not been proved yet that P = NP.

e b f Consider the language L defined as follows:
(0 + 1)* if P = NP
L=
h  f otherwise
Which of the following statements is true?

g (a) L is recursive.
(b) L is recursively enumerable but not recursive.
(c) L is not recursively enumerable.
Among the following sequences:
(d) Whether L is recursive or not will be known
(I) a b e g h f (II) a b f e h g after we find out if P = NP.
(III) a b f h g e (IV) a f g h b e (GATE 2003: 1 Mark)
Chapter 4.indd 210 4/9/2015 9:54:59 AM

Solution: If a language L belongs to class P and 10. What would be the worst-case time complexity of the
class NP then it is called recursive. insertion sort algorithm, if the inputs are restricted
Ans. (a) to permutations of 1…n with at most n inversions?
7. A graph G = (V, E ) satisfies |E | ≤3 |V | −6. The (a) Q(n2) (b) Q(n log n)
min-degree of G is defined as (c) Q(n )
1.5
(d) Q(n)
min{degree(v)}
v∈ V
Solution: With at most n inversions, insertion
Therefore, min-degree of G cannot be sort takes log n time. So, if there are n elements,
complexity will be O(n log n).
(a) 3 (b) 4 (c) 5 (d) 6
Ans. (a)
11. The cube root of a natural number n is defined as
Solution: Substituting value of E = V(V − 1)/2, the largest natural number m such that m3 ≤ n.
we get V = 3 or 4. Thus, the minimum degree of The complexity of computing the cube root of n (n
each vertex could be 1 or 2. Hence, 3 cannot be the is represented in binary notation) is
min degree.
Ans. (a) (a) O(n) but not O(n0.5)
(b) O(n0.5) but not O((log n)k) for any constant k > 0
8. Consider two languages L1 and L2, each on the (c) O((logn)k) for some constant k > 0 but not
alphabet ∑. Let f: ∑→∑ be a polynomial time O((log log n)m) for any constant m > 0
computable bijection such that (∀ x)[x ∈ L1 iff f (x) (d) O((log log n)k) for some constant k > 0.5 but
∈ L2]. Further, let f −1 also be polynomial time not O((log log n)0.5)
computable. (GATE 2003: 2 Marks)
Which of the following CANNOT be true? Solution: The complexity of computing the cube
(A) L1 ∈ P and L2 is finite root of n is O(K3 n2 log(B)). Using binary search it
(B) L1 ∈ NP and L2 ∈ P takes O((log n)k) for some constant k > 0 but not
(C) L1 is undecidable and L2 is decidable O((log n)m for any constant m > 0.
(D) L1 is recursively enumerable and L2 is recursive Ans. (c)
(GATE 2003: 2 Marks) 12. Let G = (V, E) be an undirected graph with a sub-
graph G1 = (V1, E1). Weights are assigned to edges
Solution: L1 is undecidable and L2 is decidable
are contradictory, as L1 ≤p L2. Thus, if L2 is decid-
of G as follows:
able ⇒L1 is decidable. 0 if e ∈ E1
w(e) = 
Ans. (c) 1 otherwise
Common Data Questions 8 and 9: In a permuta- A single-source shortest path algorithm is executed
tion a1…an of n distinct integers, an inversion is a on the weighted graph (V, E, w)with an arbitrary
pair (ai, aj) such that i < j and ai > aj. vertex v1 of V1 as the source. Which of the fol-
lowing can always be inferred from the path costs
9. If all permutations are equally likely, what is the
computed?
expected number of inversions in a randomly
chosen permutation of 1…n? (a) N
umber of edges in the shortest paths from v1
n(n - 1) n(n - 1)
to all vertices of G.
(a) (b) (b) G1 is connected.
2 4 (c) V1 forms a clique in G.
n(n + 1)
(c) (d) 2n[log2 n ] (d) G1 is a tree.
4 (GATE 2003: 2 Marks)
Solution: Let us take four vertices v1, v2, v3 and v4.
Solution: Inversion means i < j and ai > aj in pair
The vertex v1 is in G1 such that v1 is adjacent to v3
(ai, aj).
and v2 is adjacent to v4. Now after apply shortest path
1 n(n - 1)
Average number of inversions = (n C2 ) = algorithm we get v1v2 = 1, v1v4 = 1 and v1v3 = 0. From
2 2 these results, we can say that v1 is not a clique in G1.
Ans. (b) Ans. (b)
Chapter 4.indd 211 4/9/2015 9:55:02 AM

13. What is the weight of a minimum spanning tree of Let A be an n × n array initialized as follows:
the following graph?
1 if ( j, k) ∈ E
A[ j, k ] = 
2 g 19
j 0 otherwise
b
6 8 14 Consider the following algorithm:
3
c 5 for i = 1 to n
1 2
for j = 1 to n
a
for k = 1 to n
2 15 h 4
d
i A [j, k] = max (A[j, k] (A[j, i]+ A [i, k]);
4 9
8 8 Which of the following statements is necessarily true
11 A
f
2
for all j and k after terminal of the above algorithm?
(a) A[j, k] ≤ n.
e
(b) If A[j, j] ≥ n − 1, then G has a Hamiltonian
cycle.
(a) 29 (b) 31 (c) 38 (d) 41
(c) If there exists a path from j to k, A[j, k] contains
the longest path lens from j to k.
Solution: A minimum spanning tree will contain (d) If there exists a path from j to k, every simple
n − 1 edges. Here, 10 − 1 = 9 edges will be included. path from j to k contain most A[j, k] edges.
Total weight = 1 + 2 + 2 + 2 + 3 + 4 + 4 + 5 (GATE 2003: 2 Marks)
+ 8 = 31. Solution: If there exists a path from j to k, every
Ans. (b) simple path from j to k, it must contain A[j, k] edges.
14. The following are the starting and ending times of Ans. (d)
activities A, B, C, D, E, F, G and H, respectively, 16. Consider the label sequences obtained by the fol-
in chronological order: lowing pairs of traversals on a labelled binary tree:
às bs as ae ds ae es fs be de gs ee fe hs ge he’
(i) Preorder and postorder
Here, xs denotes the starting time and xe denotes (ii) Inorder and postorder
the ending time of activity X. We need to schedule (iii) Preorder and inorder
the activities in a set of rooms available to us. An (iv) Level order and postorder
activity can be scheduled in a room only if the room
is reserved for the activity for its entire duration. Which of these pairs identifies a tree uniquely?
What is the minimum number of rooms required? (a) (i) only (b) (ii) and (iii)
(a) 3 (b) 4 (c) 5 (d) 6 (c) (iii) only (d) (iv) only
(GATE 2003: 2 Marks) (GATE 2004: 2 Marks)
Solution: We have às bs cs ae ds ce es fs be de gs ee Solution: Only if we have inorder traversal with

fe hs ge he’. For this, the minimum number of rooms any one of postorder or preorder traversal of tree,
required is 4. can we uniquely identify a tree.
as bs cs Ans. (b)
ds bs cs 17. Two matrices M1 and M2 are to be stored in arrays
ds bs es fs A and B, respectively. Each array can be stored in
es fs either row-major or column-major order in contigu-
ous memory locations. The time complexity of an
gs es fs
algorithm to compute M1M2 will be
gs
gs hs (a) b
est if A is in row-major and B is in column-
major order.
(b) best if both are in row-major order.
Ans. (b) (c) best if both are in column-major order.
(d) independent of the storage scheme.
15. Let G = (V, E) be a directed graph with n vertices.
A path from vi to vj in G is a sequence of vertices
(vi, vi+1, …, vj) such that (vk, vk+1) ∈ E for all k in Solution: Complexity of matrices multiplication
i through j − 1. A simple path is a path in which is irrespective of storage scheme of the arrays.
no vertex appears more than once. Ans. (d)
Chapter 4.indd 212 4/9/2015 9:55:03 AM

18. Suppose each set is represented as a linked list with Solution: The maximum time taken by a for
elements in an arbitrary order. Which of the opera- loop is n. So, the complexity is Q(n).
tions among union, intersection, membership and Ans. (c)
cardinality will be the slowest?
21. The time complexity of the following C function is
(a) Union only (assume n > 0):
(b) Intersection and membership int recursive (int n) {
(c) Membership and cardinality if (n == 1)
(d) Union and intersection return (1);
(GATE 2004: 2 Marks) else
return (recursive (n – 1) + recursive (n – 1));
Solution: To calculate cardinality just counts the }
number of nodes in a list and for membership, just (a) O(n) (b) O(n log n)
traverse the list and look for a match. So, cardinal- (c) O(n2) (d) O(2n)
ity and membership is not the slowest. (GATE 2004: 2 Marks)
But For getting intersection and union of L1 and
L2, we have to perform search in both list L1 and Solution: We have
L2. So, they are the slowest one in the given list. T (n) = 2T (n - 1), n>1
Ans. (d)
T (2) = 2T (1) = 21
19. Suppose we run Dijkstra’s single-source shortest-
path algorithm on the following edge-weighted T (3) = 2T (2) = 22
directed graph with vertex P as the source: …
n -1
1 T (n) = 2 = O(2n )
Q R Ans. (d)
1 4 2 22. The recurrence equation
6 1 T (1) = 1
P S
2 T (n) = 2T (n - 1) + n, n≥2
3 evaluates to:
7
T U (a) 2n+1 − n − 2 (b)2n − n
2 (c) 2n+1 − 2n − 2 (d)2n + n
In what order do the nodes get included into the (GATE 2004: 2 Marks)
set of vertices for which the shortest path distances Solution: We have
are finalized?
T (n) = 2T (n - 1) + n, n≥2

(a) P, Q, R, S, T, U (b) P, Q, R, U, S, T
T (2) = 2T (1) + 2 = 4
(c) P, Q, R, U, T, S (d) P, Q, T, R, U, S
(GATE 2004: 2 Marks) T (3) = 2T (2) + 3 = 11
Solution: Nodes with minimum weight are included T (4) = 2T (3) + 4 = 26
first. The sequence will be P, Q, R, U, S, T. …
Ans. (b)
T (n) = 2n +1 - n - 2

20. Let A [1,…, n ] be an array storing a bit (1 or 0) at Ans. (a)
each location, and f(m) is a function whose time
complexity is Θ(m).
23. Let G1 = (V, E1) and G2 = V, E2 be connected
graphs on the same vertex set V with more than
Consider the following program fragment written two vertices. If G1 ∩ G2 = (V, E1 ∩ E2) is not a con-
in a C-like language: nected graph, then the graph G1 ∪G2 = (V, E1 ∪ E2)
counter = 0; (a) cannot have a cut vertex.
for (i = l ; i < = n; i++)
(b) must have a cycle.
{if (A [i] == 1) counter++;
else {f (counter); counter = 0;} (c) must have a cut-edge (bridge).
} (d) has chromatic number strictly greater than
those of G1 and G2.
The complexity of this program fragment is
(a) W(n2) (b) W(n log n) and O(n2)
(c) Q(n) (d) o(n) Solution: Option (b) is the most appropriate option.
Chapter 4.indd 213 4/9/2015 9:55:07 AM

24. A program takes as input a balanced binary search (c) G

raph G has multiple distinct MSTs, each of
tree with n leaf nodes and computes the value of a cost n − 1.
function g(x) for each node x. If the cost of comput- (d) Graph G has multiple spanning trees of differ-
ing g(x) is min{no. of leaf-nodes in left-subtree of ent costs.
x, number of leaf-nodes in right-subtree of x}, then (GATE 2005: 2 Marks)
the worst-case time complexity of the program is
Solution: Since diagonal elements are 0, there is
(a) O(n) (b) O(n log n) no self-loop. And every node is connected to other
(c) O(n2) (d) O(n2 log n) n − 1 nodes.
Solution: The complexity of g(x) will be O(n), 28. Let G(V,E) be an undirected graph with positive
using balanced binary tree. edge weights. Dijkstra’s single source shortest path
Ans. (a) algorithm can be implemented using the binary
25. The time complexity of computing the transitive heap data structure with time complexity:
(a) O(|V |2) (b) O(|E | + |V | log |V |)
closure of a binary relation on a set of n elements
(c) O(|V | log |V |) (d) O((|E | + |V |)log |V |)
is known to be:
(a) O(n) (b) O(n log n) (GATE 2005: 2 Marks)
(c) O(n3/2) (d) O(n3)
Solution: Complexity using Dijkstra’s algorithm
(GATE 2005: 1 Mark) for single source shortest path algorithm is
Solution: The transitive closure of a binary rela- O(E + V log V)
tion can be computed using Warshall’s algorithm. Ans. (b)
The complexity of finding it is O(n3).
Ans. (d) 29. Consider the following two problems on undirected
graphs:
26. Suppose T (n) = 2T (n/2) + n , T (0) = T (1) = 1 .
a: Given G(V,E), does G have an independent set
Which one of the following is false? of size |V| −4?
(a) T(n) = O(n2) (b) T(n) = Q(n log n) b: Given G(V,E), does G have an independent set
(c) T(n) = W(n2) (d) T(n) = O(n log n) of size 5?
(GATE 2005: 2 Marks) Which one of the following is TRUE?
Solution: This can be solved using the master (a) a is in P and b is NP-complete
theorem. (b) a is NP-complete and b is in P
n (c) a Both and b are NP-complete
T (n) = aT   + f (n) (d) a Both and b are in P
 b 
Here we have a = 2, b = 2 and f(n) = n. So, it is
the second case of master theorem. Therefore, Solution: Independent set decision problems of
T (n) = Q(n log b a
log n) graph are NP-complete problems.
Ans. (c)
T (n) = Q(n log n)
Linked Answer Questions 30 and 31: Consider the
Options (a), (b) and (d) are true because options (a) following C-function:
and (d) are indicating the upper bound which is
greater than Q(n log n). But, option (c) is false double foo (int n) {
int i;
which indicates that the lower bound is n2.
double sum;
Ans. (c) if (n==0) return 1.0;
else {
27. An undirected graph G has n nodes. Its adjacency
sum = 0.0;
matrix is given by an n × n square matrix whose for (i =0; i<n; i++)
(i) diagonal elements are 0s and (ii) non-diagonal sum +=foo(i);
elements are 1s. Which one of the following is true? return sum;
} }`
(a) Graph G has no minimum spanning tree (MST).
(b) Graph G has a unique MST of cost n − 1. (GATE 2005: 2 Marks)
Chapter 4.indd 214 4/9/2015 9:55:09 AM

30. The space complexity of the above function is: 34. Let S be an NP-complete problem and Q and R
be two other problems not known to be in NP. Q
(a) O(1) (b) O(n)
is polynomial time reducible to S, and S is polyno-
(c) O(n!) (d) O(nn)
mial time reducible to R. Which one of the follow-
Solution: The space complexity is the number ing statements is true?
of elements to be stored. This will depend on the (a) R is NP-complete. (b) R is NP-hard.
height of tree produced. Maximum storage required (c) Q is NP-complete. (d) Q is NP-hard.
is O(n). (GATE 2006: 1 mark)
Ans. (b)
Solution: R is NP-hard because S is polynomial
31. Suppose we modify the above function foo() and time reducible to R.
store the values of foo(i), 0 < = I < n, as and Ans. (b)
when they are computed. With this modification,
the time complexity for function foo() is signifi- 35. A set X can be represented by an array x[n] as
cantly reduced. The space complexity of the modi- follows:
fied function would be: 1 if i ∈ X
x[i] = 
(a) O(1) (b) O(n) 0 otherwise
(c) O(n2) (d) O(n!) Consider the following algorithm in which x, y and
z are Boolean arrays of size n:
Solution: Complexity will remain the same. With
modification also, the required space will be same. algorithm zzz(x[] , y[], z [])
Ans. (b) {
int i;
Linked Answer Questions 32 and 33: We are given for (i=0; i<n; ++i)
9 tasks T1, T2, …. T9. The execution of each task z[i] = (x[i] ^ ~y[i]) V (~x[i] ^ y[i])
requires one unit of time. We can execute one task }
at a time. Each task Ti has a profit Pi and a dead- The set Z computed by the algorithm is:
line di. Profit Pi is earned if the task is completed
before the end of the (di)th unit of time. (a) (X intersection Y )
(b) (X union Y )
Task T1 T2 T3 T4 T5 T6 T7 T8 T9 (c) (X−Y ) intersection (Y−X )
(d) (X−Y ) union (Y−X )
Profit 15 20 30 18 18 10 23 16 25 (GATE 2006: 2 Marks)
Deadline 7 2 5 3 4 5 2 7 3 Solution:
x[i]^~y[i] = only 1's in x where corresponding
32. Are all tasks completed in the schedule that gives entry in y is 0 → X − Y
maximum profit? ~x[i]^y[i] = only 1's in y where corresponding
(a) All tasks are completed entry in x is 0 → Y − X
(b) T1 and T6 are left out The operator `V ’ results in a union of the above
(c) T1 and T8 are left out two sets.
(d) T4 and T6 are left out Ans. (d)
36. Let SHAM3 be the problem of finding a Hamiltonian
Solution: cycle in a graph G = (V,E) with V divisible by 3
T2 T7 T9 T5 T3 T8 T1 and DHAM3 be the problem of determining if a
Hamiltonian cycle exists in such graphs. Which one
T4 and T6 are left out. of the following is true?
Ans. (d)
33. What is the maximum profit earned? A B AoB
(a) 147 (b) 165 (c) 167 (d) 175 True True True
(GATE 2005: 2 Marks) True False True
Solution: Maximum profit = 20 +23 + 25 + 18 + False True False
30 + 15 + 16 = 147 False False True
Ans. (a)
Chapter 4.indd 215 4/9/2015 9:55:10 AM

(a) Both DHAM3 and SHAM3 are NP-hard Solution: If we use Queue data structure to im‑
(b) SHAM3 is NP-hard, but DHAM3 is not plement Dijkstra’s shortest path algorithm on
(c) DHAM3 is NP-hard, but SHAM3 is not unweighted graphs Then shortest path will be
(d) Neither DHAM3 nor SHAM3 is NP-hard calculated in linear time O(|V | + |E |). Basically we
(GATE 2006: 2 Marks) do BFS traversal of graph to get the shortest path.
Ans. (a)
Solution: Both are NP-hard problems, and finding 40. Consider the following graph:
the Hamiltonian cycle is also an NP-hard problem.
Ans. (a) 2
37. Consider the following recurrence:
(
T (n) = 2T  n  + 1,
 ) T (1) = 1
1
b 4 d
1
Which one of the following is true? 2
a 3 3 f
(a) T(n) = Q(log log n) (b) T(n) = Q(log n) 5
(c) T(n) = Q(sqrt(n)) (d) T(n) = Q(n) 6 4
c e
7
Solution: Let n = 2m. Then, we have
T (2m ) = T (2m /2 ) + 1 Which one of the following cannot be the sequence

m of edges added, in that order, to a minimum span-
Let T(2 ) = S(m). Then, we have
ning tree using Kruskal’s algorithm?
m
S(m) = 2S   + 1
 2  (a) (a—b), (d—f), (b—f), (d—c), (d—e)
(b) (a—b), (d—f), (d—c), (b—f), (d—e)
The above expression is a binary tree traversal (c) (d—f), (a—b), (d—c), (b—f), (d—e)
recursion whose time complexity is q(m). You can (d) (d—f), (a—b), (b—f), (d—e), (d—c)
also prove this using the master theorem: (GATE 2006: 2 Marks)
S(m) = q (m) = q (log n) Solution: Edges are added according to increas-
m ing number of weights. In option (d), d—e is added
since n = 2 .
Now, let us go back to the original recursive func- before d—c.
tion T(n): Ans. (d)
41. The median of n elements can be found in O(n)
T (n) = T (2m ) = S(m) = Q(log n) time. Which one of the following is correct about
Ans. (b) the complexity of quick sort, in which median is
2 3 selected as pivot?
38. Consider the polynomial p(x) = a0 + a1x + a2 x + a3 x
(x) = a0 + a1x + a2 x2 + a3 x3, where ai! = 0, for all i. The minimum (a) O(n) (b) O(n log n)
number of multiplications needed to evaluate p on (c) O(n2) (d) O(n3)
an input x is (GATE 2006: 2 Marks)

(a) 3 (b) 4 (c) 6 (d) 9 Solution: Randomized quick sort:
T (n) = 2T (n/2) + 1 = O(n log n).
Solution: We have Ans. (b)
2 2
p(x) = a0 + a1x + a2 x + a3 x 42. Consider a weighted complete graph G on the
vertex set {v1, v2, …, vn} such that the weight of
x2 = 1 multiplication the edge (vi, vj) is 2|i − j |. The weight of a mini-
a1x + a2x2 + a3x3 = 3 multiplications mum spanning tree of G is
(a) n − 1 (b) 2n − 2 (c) nC2 (d) 2
Total four multiplications are required.
Ans. (b)
39. To implement Dijkstra’s shortest path algorithm
Solution: Weight of the minimum spanning tree is
on unweighted graphs so that it runs in linear time,
the data structure to be used is: 2 |2 - 1| +2 |3 - 2| +2 |4 - 3|
(a) Queue (b) Stack (c) Heap (d) B-tree + 2 |5 - 4| + + 2 |n - (n - 1)| = 2n - 2
Chapter 4.indd 216 4/9/2015 9:55:13 AM

43. Let w be the minimum weight among all edge weights 46. In the following C function, let n ≥ m:
in an undirected connected graph. Let e be a specific int gcd(n,m)
edge of weight w. Which of the following is false? {
if (n%m ==0) return m;
(a) There is a minimum spanning tree containing e.
n = n%m;
(b) If e is not in a minimum spanning tree T, then return gcd(m,n);
in the cycle formed by adding e to T, all edges }
have the same weight.
(c) Every minimum spanning tree has an edge of How many recursive calls are made by this function?
weight w. (a) Q(log2n) (b) W(n)
(d) e is present in every minimum spanning tree. (c) Q(log2log2n) (d) Q(sqrt(n))
Solution: Options (a), (b) and (c) are correct. But
Solution:
option (d) is incorrect as there may be many edges
Here, Tn = 1 + (T0 T1 + + Tn -1 ) = Θ(log 2 n)
1
of weight w in the graph and edge e may not be in n
some of the minimum spanning trees. Ans. (a)
Ans. (d)
44. Consider the following C code segment: 47. What is the time complexity of the following recur-
sive function?
int IsPrime(n)
{ int DoSomething (int n)
int i,n; {
for(i=2;i<=sqrt(n);i++) if (n <= 2)
if(n%i == 0) return 1;
{printf(‘Not Prime\n’); return 0;} else
return 1; return (DoSomething (floor(sqrt(n)))+ n);
} }
Let T(n) denote the number of times the for loop (a) Q(n) (b) Q(n log2n)
is executed by the program on input n. Which of (c) Q(log n) (d) Q(log2log2n)
the following is true? (GATE 2007: 2 Marks)
(a) T(n) = O(sqrt(n)) and T(n) = W(sqrt(n))
Solution: Recursive relation for the DoSomething() is
(b) T(n) = O(sqrt(n)) and T(n) = W(1)
(c) T(n) = O(n) and T(n) = W(sqrt(n)) ( T (n) = T ( n ) + C1 ) if n > 2
(d) None of the above Let n = 2m, so T(n) = T( 2m )
(GATE 2007: 2 Marks) Let T(2m) = S(m)
Solution: The for loop runs maximum for sqrt(n) From the above two expressions, we have
and minimum for 1. Therefore, T(n) = S(m)
T(n) = O(sqrt(n)) and T(n) = W(1) S(m) = S(m/2) + C1
Ans. (b) S(m) = O(log2m) (by Master Theorem)
S(m) = O(log2log2n) (because n = 2m)
45. In an unweighted, undirected connected graph, the
From the above, we have S(m) = T(n) = O(log2log2n)
shortest path from a node S to every other node
Ans. (d)
is computed most efficiently in terms of time com-
plexity by Linked Answer Questions 48 and 49: Suppose
(a) Dijkstra’s algorithm starting from S the letters a, b, c, d, e, f have probabilities
(b) Warshall’s algorithm 1 1 1 1 1 1
, , , , , , respectively.
(c) Performing a DFS starting from S 2 4 8 16 32 32
(d) Performing a BFS starting from S
(GATE 2007: 2 Marks) 48. Which of the following is the Huffman code for the
letter a, b, c, d, e, f?
Solution: Time complexity of BFS is O(|E | +
|V |), whereas time complexity of Dijkstra’s is (a) 0, 10, 110, 1110, 11110, 11111
(b) 11, 10, 011, 010, 001, 000
O(½V½2 + E ) and Warshall’s algorithm is O(|V |3 )
(c) 11, 10, 01, 001, 0001, 0000
and DFS cannot be used for finding shortest path. (d) 110, 100, 010, 000, 001, 111
Ans. (d) (GATE 2007: 2 Marks)
Chapter 4.indd 217 4/9/2015 9:55:16 AM

Solution: 52. G is a graph on n vertices and 2n − 2 edges. The

edges of G can be partitioned into two edge-disjoint
a b c d e f spanning trees. Which of the following is NOT true
1/2 1/4 1/8 1/16 1/32 1/32 for G?
(a) F or every subset of k vertices, the induced sub-
abcdef: 1 graph has at most 2k − 2 edges.
0 1 (b) The minimum cut in G has at least two edges.
a:1/2 bcdef: 1/2 a: 0
(c) There are two edge-disjoint paths between
0 1 b: 10
every pair of vertices.
b:1/4 cdef: 1/4 c: 110
0 1 (d) There are two vertex-disjoint paths between
d: 1110 every pair of vertices.
c:1/8 def: 1/8
0 1 e: 11110 (GATE 2008: 2 Marks)
d:1/16 ef: 1/16 f: 11111
0 1 Solution: Option (a) is true since the induced
e: f: sub-graph on k vertices can be constructed as the
1/32 1/32
union of the induced sub-graphs on k vertices of
Ans. (a) two edge disjoint spanning tree and each of the
induced spanning trees has atmost k − 1 edges.
49. What is the average length of the correct answer to Option (b) is true otherwise G cannot be parti-
the above question? tioned into two edge disjoint spanning trees.
Option (c) is true otherwise G has minimum cut
(a) 3 (b) 2.1875 (c) 2.25 (d) 1.9375 of one edge.
(GATE 2007: 2 Marks) Option (d) is false.
Solution: The average length = Ans. (d)
1 1 1 1 1 1
1× + 2× + 3× + 4× + 5× + 5× = 1.9375 53. Consider the Quicksort algorithm. Suppose there is a
2 4 8 16 32 32 procedure for finding a pivot element which splits the
Ans. (d) list into two sub-lists each of which contains at least
one-fifth of the elements. Let T(n) be the number of
50. The most efficient algorithm for finding the number
comparisons required to sort n elements. Then
n
of connected components in an undirected graph
(a) T (n) ≤ 2T   + n
 5 
on n vertices and m edges has time complexity
(a) Q(n) (b) Q(m) n  4n 
(c) Q(m + n) (d) Q(mn) (b) T (n) ≤ T   + T   + n
 5   5 
(GATE 2008: 1 Mark)
 4n 
(c) T (n) ≤ 2T   + n
Solution: DFS can be used to find out the number  5 
of connected components in an undirected graph. n
(d) T (n) ≤ 2T   + n
The time complexity of DFS is O(m + n).  2 
Ans. (c)
51. The minimum number of comparisons required
to determine if an integer appears more than n/2 Solution: The recurrence relation is T(n) = T(a)
times in a sorted array of n integers is + T(b) + n,a ≥ n/5, b ≥ n/5 and a + b = n
 4n 
(a) Q(n) (b) Q(log n) Thus, a ≤ 4n/5 and b ≤ 4n/5, so T (n) ≤ 2T   + n
 5 
(c) Q(log*n) (d) Q(1)
(GATE 2008: 1 Mark) Ans. (c)
Solution: The minimum number of comparison 54. The subset-sum problem is defined as follows:
would be 1 since if an element exists more than n/2 Given a set S of n positive integers and a positive
times in a sorted list than it has to be in middle. integer W, determine whether there is a subset of S
The middle index of an array can be reached in a whose elements sum to W. An algorithm Q solves
constant time. this problem in O(nW ) time. Which of the follow-
Ans. (a) ing statements is false?
Chapter 4.indd 218 4/9/2015 9:55:17 AM

(A) Q solves the subset-sum problem in polynomial b -3 e

time when the input is encoded in unary
2 2
(B) Q solves the subset-sum problem in polynomial 1
-5
1 1
time when the input is encoded in binary a c h f
(C) The subset sum problem belongs to the class NP 2 3 2 3
(D) The subset sum problem is NP-hard 2
d g
Solution: Since we know that subset sum is a (a) Only vertex a

NP-hard problem, thus problem may take linear (b) Only vertices a, e, f, g and h
time when the input encoded is unary. (c) Only vertices a, b, c and d
Ans. (b) (d) All of the vertices
Linked Answer Questions 55 and 56: The sub-
setsum problem is defined as follows. Given a Solution: Following the Dijkstra’s algorithm’s single-
set of n positive integers, S = {a1, a2, a3, …, an} source shortest paths, we can reach all the vertices here.
and positive integer W, is there a subset of S Taking à' as the source: minimum distance is
whose elements sum to W? A dynamic program computed by checking all the adjacent nodes, as
for solving this problem uses a two-dimensional follows:
Boolean array X, with n rows and W + 1 col- D(a) = 0
umns. X[i, j], 1 ≤ i ≤n, 0 ≤j ≤W, is true if and
D(b) = cost(a → b) = 1
only if there is a subset of {a1, a2, …, ai} whose
elements sum to j. Here, the minimum distance from source is com-
55. Which of the following is valid for 2 ≤ i ≤ n and
puted, but since there is only one adjacent node,
ai ≤ j ≤ W?
there is no need to check any other node for
minimum.
(a) X[i, j] = X[i − 1, j] ∨X[i, j − ai] Now, here b node has two adjacent nodes e and
(b) X[i, j] = X[i − 1, j] ∨ X[i − 1, j − ai] c. But since cost(b → e) has a negative weight,
(c) X[i, j] = X[i − 1, j] ∨X[i, j −ai] we ignore that path as Dijkstra does not enter-
(d) X[i, j] = X[i −1, j] ∨ X[i −1, j −ai] tain negative weights. Thus, we will proceed
(GATE 2008: 2 Marks) with node c:
Solution: X[i, j] (2 ≤i ≤n and ai ≤j ≤W) is true D(c) = D(b) + cost(b → c) = 2 + 1 = 3
if any of the following is true:
(1) Sum of weights excluding ai is equal to j, that Also c node could be reached only through b, so
is, if X[i −1, j] is true. that is the only possible minimum weight of c.
(2) Sum of weights including ai is equal to j, that is, As similar to node b, node c is also left with one
if X[i −1, j −ai] is true so that we get (j −ai) + child d due to negative edge removal c → h:
ai as j.
Ans. (b) D(d) = D(c) + cost(c → d) = 3 + 3 = 6
56. Which entry of the array X, if true, implies that Similarly, proceeding further the same way, we can
there is a subset whose elements sum to W ? reach all nodes and compute the shortest path.
Ans. (d)
(a) X[1, W] (b) X[n, 0]
(c) X[n, W] (d) X[n −1, n] 58. Consider the following functions:
f (n) = 2n g(n) = n ! h(n) = n log n
Solution:
Which of the following statements about the
If we get the entry X[n, W] as true, then there is a
asymptotic behaviour of f(n), g(n) and h(n) is true?
subset of {a1, a2, …, an} that has sum as W.
Ans. (c) (a) f(n) = O(g(n)); g(n) = O(h(n))
(b) f(n) = W(g(n)); g(n) = O(h(n))
57. Dijkstra’s single-source shortest path algorithm (c) g(n) = O(f(n)); h(n) = O(f(n))
when run from vertex a in the following graph (d) h(n) = O(f(n)); g(n) = W(f(n))
computes the correct shortest path distance to (GATE 2008: 2 Marks)
Chapter 4.indd 219 4/9/2015 9:55:19 AM

Solution: The functions can be arranged in an 5 mod 10 → 5 [5, 6 occupied] so will go to 7

increasing order as h(n) < f(n) < g(n). Hence, 15 mod 10 → 5 [5, 6, 7, 8 are occupied] will go to 9
option (d) is the correct one. Ans. (c)
Ans. (d)
62. Consider the following graph:
59. Let X be a problem that belongs to the class NP.
Then which one of the following is true? b c
2
5 6 5
(a) There is no polynomial time algorithm for X. 6
(b) If X can be solved deterministically in polyno-
mial time, then P = NP. a 4 d 3 g
(c) If X is NP-hard, then it is NP-complete. 6
(d) X may be undecidable. 3 5 4
(GATE 2009: 1 mark) c f
6
Solution: Option (a) is incorrect because NP (GATE 2009: 2 Marks)
includes both P and NPC.
Option (b) is also incorrect because X may belong Which one of the following is NOT the sequence of
to P but it does not mean that P = NP. edges added to the minimum spanning tree using
Option (c) is correct because NP-complete set is Kruskal’s algorithm?
intersection of NP and NP-hard sets. (A) (b, e) (e, f) (a, c) (b, c) (f, g) (c, d)
Option (d) is incorrect because all NP problems are (C) (b, e) (a, c)(e, f) (b, c) (f, g) (c, d)
decidable in finite set of operations. (B) (b, e) (e, f) (a, c) (f, g) (b, c)(c, d)
Ans. (c)
(D) (b, e) (e, f) (b, c)(a, c) (f, g) (c, d)
60. What is the minimum number of swaps required to (GATE 2009: 2 Marks)
sort n elements using selection sort, in the worst case?
Solution: In Kruskal’s algorithm edges are chosen
(a) Q(n) (b) Q(n log n) in increasing order of weights. So, (b, e) = 2, (e, f)
(c) Q(n2) (b) Q(n2 log n) = 3, (b, c) = 4, (a, c) = 3, (f, g) = 4 and (c, d) = 5.
(GATE 2009: 1 mark) Ans. (d)
Solution: Selection sort select the minimum ele- 63. The running time of an algorithm is represented by
ment and the swap with first position. In the worst the following recurrence relation:
case, all n elements will be swapped.
if n <= 3 then T(n) = n
Ans. (a) else T(n) = T(n/3) + cn
61. The keys 12, 18, 13, 2, 3, 23, 5 and 15 are inserted Which one of the following represents the time
into an initially empty hash table of length 10 using complexity of the algorithm?
open addressing with hash function h(k) = k and
linear probing. What is the resultant hash table? (a) Q(n) (b) Q(n log n)
(c) Q(n2) (d) Q(n2 log n)
(A) (B) (C) (D) (GATE 2009: 2 Marks)
0 0 0 0
1 1 1 1
2 2 2 12 2 12 2 12,2 Solution: As per the master theorem, case 3, we
3 23 3 13 3 13 3 13, 3, 23 have a = 1, b= 3 and f(n) = cn. Thus
4 4 4 2 4
5 15 5 5 5 3 5 5, 15
logb a = log3 1 = 0
6 23 6
n 0 < f (n)
6 6
7 7 7 5 7
8 18 8 18 8 18
So, the answer is Q(n).
8 18
9 9 9 15 9
Ans. (a)
64. In quick sort, for sorting n elements, the (n/4)th
Solution: smallest element is selected as pivot using an O(n)
12 mod 10 → 2 time algorithm. What is the worst-case time com-
18 mod 10 → 8 plexity of the quick sort?
13 mod 10 → 3
2 mod 10 → 2 [2, 3 are occupied] so will go to 4 (a) Q(n) (b) Q(n log n)
3 mod 10 → 3 [3, 4 are occupied] so will go to 5 (c) Q(n2) (d) Q(n2 log n)
23 mod 10 → 3 [3, 4, 5 are occupied] so will go to 6 (GATE 2009: 2 Marks)
Chapter 4.indd 220 4/9/2015 9:55:21 AM

Solution: The recursive expression is Solution: The value can be computed in either
n  3n  row-major or column-major order.
T (n) = T   + T   + cn
 4   4 
Ans. (b)
T (n) = Q(n log n) 67. Two alternative packages A and B are available for
processing a database having 10k records. Package
This expression is of randomized quick sort. A requires 0.0001 n2 time units and package B
Ans. (b) requires 10nlog10n time units to process n records.
Common Data Questions 65 and 66: A subsequence What is the smallest value of k for which the pack-
of a given sequence is just the given sequence with age B will be preferred over A?
some elements (possibly none or all) left out. We (a) 12 (b) 10
are given two sequences X[m] and Y[n] of lengths (c) 6 (d) 5
m and n, respectively, with indexes of X and Y (GATE 2010: 1 mark)
starting from 0.
Solution:
0.0001 n2< 10 n logn ⇒
65. We wish to find the length of the longest common
subsequence (LCS) of X[m] and Y[n] as l(m, n),
where an incomplete recursive definition for the 10-5 n < log n ⇒ 10-5 × 10k < k (as log10 n = k)
function l(i, j) to compute the length of the LCS of
Hence, k −5 > 0 or k> 6. Thus, for k = 6, B will
X[m] and Y[n] is given as follows:
be preferred.
l (i, j) 0, if either i=0 or j=0 Ans. (c)
= expr1, if i,j>0 and X[i -1] = Y[j - 1]
= expr2, if i,j>0 and X[i - 1] = Y[j - 1] 68. The weight of a sequence a0, a1, … an −1 of real
Which one of the following options is correct? numbers is defined as
(a) expr1 ≡ (i −1, j) + 1
(b) expr1 ≡ l(i, j −1) a0 + a1 /2 + + an -1 / 2n -1
(c) expr2 ≡ max(l(i −1, j), l(i, j −1)) A subsequence of a sequence is obtained by delet-
(d) expr2 ≡ max(l(i −1, j −1), l(i, j)) ing some elements from the sequence, keeping the
(GATE 2009: 2 Marks) order of the remaining elements the same. Let X
Solution: LCS problem is solved using dynamic denote the maximum possible weight of a subse-
programing in which every row is dependent upon quence of a0, a1, …, an − 1 and Y the maximum
previous row and column. So, if two characters at ith possible weight of a subsequence of a1, a2, …, an − 1.
row and jthcolumn do not match then the length is Then X is equal to
calculated as: (a) max(Y, a0 + Y ) (b) max(Y, a0 + Y )
expr2 = max(l(i −1, j), l(i, j −1)) (c) max(Y, a0 + 2Y ) (d) a0 + Y/2
The values of l(i, j) could be obtained by Solution: This involves dynamic programming in
dynamic programming based on the correct recur- algorithms. Given that
sive definition of l(i, j) of the form given above, X= Maximum weight from the sequence (a0,a1,a2,
using an array L[M, N], where M = m + 1 and N
…an −1) = a + a1 + a2 + + n -1
= n + 1, such that L[i, j] = l(i, j). a
2n -1
Ans. (c) 0
2 4
66. Which one of the following statements would be Y = Maximum weight from the sequence (a1,a2, …
true regarding the dynamic programming solution
a2 a3 a -1
for the recursive definition of l(i, j)? an−1) = a1 + + + + nn-
2 4 2 2
(a) All elements L should be initialized to 0 for the
values of l(i, j) to be properly computed. Here, we have to find X in terms of Y.
(b) The values of l(i, j) may be computed in a row- a2 a3 a -1
major order or column-major order of L(M, N). So, X = a1 + + + + nn- =Y
2 4 2 2
(c) The values of l(i, j) cannot be computed in
+ + nn -
either row-major order or column-major order a1 a2 a 1
if a0 + +
of L(M, N). 2 4 2 -1
(d) L[p, q] needs to be computed before L[r, s] if a a a -1
either p < r or q < s. < a1 + 2 + 3 + + nn-
(GATE 2009: 2 Marks) 2 4 2 2
Chapter 4.indd 221 4/9/2015 9:55:24 AM

OR function h(k) = k mod 10, and linear probing. After

inserting 6 values into an empty hash table, the v
+ + nn -
a1 a2 a 1
X = a0 + + = a0 +Y/2 if
2 -1
is as shown below:
2 4
0
+ + nn -
a1 a2 a 1 a a a -1
if a0 + + -1
> a1 + 2 + 3 + + nn- 1
2 4 2 2 4 2 2 2 42
a a a -1 3 23
> a1 + 2 + 3 + + nn-
2 4 2 2 4 34
5 52
Hence, we sum up as: X = max(Y,a0+Y/2) 6 46
Ans. (b) 7 33
8
Common Data Questions 69 and 70: Consider a
complete undirected graph with vertex set {0, 1, 2,
9
3, 4}. Entry Wij in the following matrix W is the
weight of the edge {i, j}: 71. Which one of the following choices gives a pos-
sible order in which the key values could have been
 0 1 8 1 4 inserted in the table?
 
 1 0 12 4 9
(a) 46, 42, 34, 52, 23, 33
W =  8 12 0 7 3
(b) 34, 42, 23, 52, 33, 46
(c) 46, 34, 42, 23, 52, 33
 1 4 7 0 2 (d) 42, 46, 33, 23, 34, 52
 
 4 9 3 2 0 (GATE 2010: 2 Marks)
Solution: Check the options according to linear

69. What is the minimum possible weight of a span- probing.
ning tree T in this graph such that vertex 0 is a leaf The correct sequence is 46, 34, 42, 23, 52, 33.
node in the tree T ? Ans. (c)
(a) 7 (b) 8
72. How many different insertion sequences of the key
(c) 9 (d) 1
values using the same hash function and linear
probing will result in the hash table shown above?
Solution: To get the minimum spanning tree (a) 10 (b) 20
with vertex 0 as leaf, first remove 0th row and (c) 30 (d) 40
0th column and then get the minimum spanning (GATE 2010: 2 Marks)
tree (MST) of the remaining graph. Once we
have MST of the remaining graph, connect the Solution: We have n(n - 1) = 6 × 5 = 30
MST to vertex 0 with the edge with minimum Ans. (c)
weight (we have two options as there are two
1's in the 0th row). 73. A max-heap is a heap where the value of each
Ans. (d) parent is greater than or equal to the value of its
children. Which of the following is a max-heap?
70. What is the minimum possible weight of a path P
from vertex 1 to vertex 2 in this graph such that P (a) (b)
10 10
contains at most three edges?
(a) 7 (b) 8
(c) 9 (d) 10
8 6 8 6
Solution: Path: 1 → 0 → 4 → 2
Weight: 1 + 4 + 3 = 8 4 5 2 4 5 1 2
Ans. (b)
Linked Answer Questions 71 and 72: A hash 1

table of length 10 uses open addressing with hash
(c) (d)
10 5
Chapter 4.indd 222 5 6 2 8 4/9/2015 9:55:26 AM

8 6 8 6
4 5 2 4 5 1 2
1 GATE PREVIOUS YEARS’ QUESTIONS 223
(c) (d) 76. Which of the given options provides the increasing
10 order of asymptotic complexity of functions f1, f2,
5
f3 and f4?
f1(n) = 2n; f2(n) = n3/2; f3(n) = n log2n;
5 6 2 8 f4 (n) = n log2 n
(a) f3, f2, f4, f1 (b) f3, f2, f1, f4
(c) f2, f3, f1, f4 (d) f2, f3, f4, f1
4 8 2 1 1 4 6 10 (GATE 2011: 2 Marks)
(GATE 2011: 1 mark) Solution: Take n as a large value and evaluate

functions. Let n = 32, then
Solution: Heap is also a complete binary tree.
Ans. (b) f1=4294967296 f2 =181.019 f3 =160
74. An algorithm to find the length of the longest f4 =33554432
monotonically increasing sequence of numbers in f3 < f2 < f4 < f1
an arrayA[O: n −1] is given below. Ans. (a)
Let Li denote the length of the longest monotonically 77. Consider a relational table r with sufficient number
increasing sequence starting at index i in the array. of records, having attributes A1, A2, …, An and
let 1 ≤ p ≤n. Two queries Q1 and Q2 are given as
Initialize Ln-L
follows:
For all i such that 0 ≤ i ≤ n − 2
Q1: ΠA ,…, A (s A = c (r)), where c is a constant
1 + Li +1 if A[i] < A[i + 1]
Li = 
1 p p
1 otherwise Q2: ΠA ,…, A (s c ≤ Ap ≤ c2 (r)), where c1 and c2 are

1 p 1
Finally, the length of the longest monotonically constants

increasing sequence is Max(L0, L1, …, Ln −1). The database can be configured to do ordered
Which of the following statements is true? indexing on Ap or hashing on Ap. Which of the fol-
lowing statements is true?
(a) The algorithm uses dynamic programming
paradigm. (a) O rdered indexing will always outperform hash-
(b) The algorithm has a linear complexity and uses ing for both queries.
branch and bound paradigm. (b) Hashing will always outperform ordered index-
(c) The algorithm has a non-linear polynomial com- ing for both queries.
plexity and uses branch and bound paradigm. (c) Hashing will outperform ordered indexing on
(d) The algorithm uses divide and conquer paradigm. Q1, but not on Q2.
(GATE 2011: 1 Mark) (d) Hashing will outperform ordered indexing on
Q2, but not on Q1.
Solution: In the given algorithm, length of the (GATE 2011: 2 Marks)
longest monotonically increasing sequence is cal-
culated as :Li in terms of Li+1. So, it uses dynamic Solution:
programing technique. Hashing will outperform ordered indexing on Q1,
Ans. (a) but not Q2.
Ans. (c)
75. We are given a set of n distinct elements and an
Linked Answer Questions 78 and 79: An undirected
unlabelled binary tree with n nodes. In how many
graph G(V, E) contains n (n> 2) nodes named v1,
ways can we populate the tree with the given set
v2, …, vn. Two nodes vi and vj are connected if
so that it becomes a binary search tree?
and only if 0 <− i −j − ≤2. Each edge (vi, vj) is
1 2n
(a) 0 (b) 1 (c) n! (d) Cn assigned a weight i + j. A sample graph with n =
n +1 4 is shown as follows:
7
Solution: We can populate only in one way because V3 V4
minimum value will be in the leftmost node, and 4 5
the maximum value will be in the rightmost node. 6
Recursively, we can define for other nodes.
V1 V2
Ans. (b) 3
Chapter 4.indd 223 4/9/2015 9:55:29 AM

78. What will be the cost of the minimum spanning MST with 6 nodes have weight = 3 + 4 + 6 + 8
tree (MST) of such a graph with n nodes? + 10 = 31
Ans. (c)
(a) (1/12) (11n2 − 5n) (b) n2 −n + 1
(c) 6n −11 (d) 2n + 1 80. Assuming P! = NP, which of the following is true?
(a) NP-complete = NP
Solution: (b) NP-complete ∩ P = F
Let us have 2 nodes in the tree, then minimum (c) NP-hard = NP
spanning tree is (d) P = NP-complete
v1 - v2 (GATE 2012: 1 mark)
Total weight = 3
Minimum spanning tree for 3 nodes would be Solution:
v1 - v2
|
v3 NP
Total weight= 3 + 4 = 7 NPC
P NP-hard
Minimum spanning tree for 4 nodes would be
v1 - v2 - v4
|
v3
Total weight= 3 + 4 + 6 = 13 Diagram clearly shows that NP-complete ∩ P = e.
From the above examples, we observe that when we Ans. (b)
add ith node, the weight of spanning tree increases
81. The worst-case running time to search for an ele-
by 2i − 2. Let T(n) be the weight of minimum
ment in a balanced binary search tree with n2n
spanning tree.
elements is
ìT ( n - 1) + ( 2n - 2 ) , n>2
ï (b) O(n2n)
T (n) = í
(a) O(n log n)
3, n =1 (c) O(n) (d) O(log n)
ï
î 0, n=2 (GATE 2012: 1 mark)
The recurrence can be written as sum of series Solution: The worst-case complexity of a bal-
(2n − 2) + (2n − 4) + (2n − 6) + (2n − 8) + … 3, anced binary search tree is Q(log n), where n is the
and solution of this recurrence is n2 − n + 1. number of elements; here, n is n2n, so
Ans. (b)
log(n2n ) = log n + n log 2 = log n + n
79. The length of the path from v5 to v6 in the MST of
previous question with n = 10 is So,the complexity is Q(n).
Ans. (c)
(a) 11 (b) 25 (c) 31 (d) 41
(GATE 2011: 2 Marks) 82. The recurrence relation capturing the optimal exe-
Solution: Basic structure of all MSTs with more cution time of the Tower of Hanoi problem with n
than 5 nodes(for the above given graph specifica- discs is
tion) would be following: (a) T (n) = 2T (n - 2) + 2
(b) T (n) = 2T (n - 1) + n
1 2 4 6
(c) T (n) = 2T (n/2) + 1
even number of nodes (d) T (n) = 2T (n - 1) + 1
3 (GATE 2012: 1 mark)
Solution: Recurrence relation for Tower of Hanoi is

5
T (n) = 2T (n - 1) + 1
odd number of nodes
Ans. (d)
Chapter 4.indd 224 4/9/2015 9:55:32 AM

83. Let G be a simple undirected planar graph on 10 S and T. Which one will be reported by Dijkstra’s
vertices with 15 edges. If G is a connected graph, shortest path algorithm? Assume that, in any itera-
then the number of bounded faces in any embed- tion, the shortest path to a vertex v is updated only
ding of G on the plane is equal to when a strictly shorter path to v is discovered.
(a) 3 (b) 4 2
(c) 5 (d) 6
(GATE 2012: 1 mark) E 2 G
C 1
Solution: We have 1
A 1 4 3
Given that vertices (V ) = 10 and edges (E) = 15 3
Number of bounded and unbounded faces (F) = ? 4
We know that V - E + F = 2 7 3
D
10 - 15 + F = 2 ⇒ F = 7 S T
4 5
3 5
So, the bounded faces = F −1 = 7 −1 = 6
Ans. (d)
3
B F
84. Let w(n) and A(n) denote, respectively, the worst-
case and average-case running time of an algorithm (a) SDT (b) SBDT
executed on an input of size n. Which of the follow- (c) SACDT (d) SACET
ing is always true? (GATE 2012: 2 Marks)
(a) A(n) = W(W(n)) (b) A(n) = Q(W(n)) Solution: Apply Dijkstra’s shortest path algorithm.
(c) A(n) = O(W(n)) (d) A(n) = o(W(n))
(GATE 2012: 1 mark) S A B C D E G T
0 — — — — — — —
Solution: 0 4 3 — 7 — — —
A(n) = O(W(n))
Ans. (c) 0 4 3 5(A) 7 — — —
0 4 3 5(A) 7 6(C) — —
85. Let G be a complete undirected graph on 6 verti- 0 4 3 5(A) 7 6(C) 8(E) 10(E)
ces. If vertices of G are labelled, then the number
of distinct cycles of length 4 in G is equal to So, the minimum path will cover S → A → C → E
(a) 15 (b) 30 → T, with total weight 10.
(c) 90 (d) 360 Ans. (d)
(GATE 2012: 2 Marks) 88. Which one of the following is the tightest upper
bound that represents the number of swaps
Solution: The number of distinct cycles of length
required to sort n numbers using selection sort?
4 is
C4 × (4 - 1)! = 6C4 × 3 ! = 90
6 (a) O(log n) (b) O(n)
(c) O(n log n) (d) O(n2)
Ans. (c) (GATE 2013: 1 mark)
86. A list of n strings, each of length n, is sorted into Solution: The maximum number of swaps
lexicographic order using the mergesort algo- required for n elements is n.
rithm. The worst-case running time of this com‑ Ans. (b)
putation is
89. Which one of the following is the tightest upper
(a) O(n log n) (b) O(n2 log n) bound that represents the time complexity of insert-
(c) O(n2 + log n) (d) O(n2) ing an object into a binary search tree of n nodes?
(a) O(1) (b) O(log n)
Solution: The worst-case runtime: n log n. (c) O(n) (d) O(n log n)
For n strings: n(n log n), which gives n2 log n. (GATE 2013: 1 mark)
Ans. (b) Solution: In case of skewed binary tree, the tightest
87. Consider the directed graph shown in the figure below. upper bound of the insert node is O(n).
There are multiple shortest paths between vertices Ans. (c)
Chapter 4.indd 225 4/9/2015 9:55:33 AM

90. What is the time complexity of Bellman-Ford sin- Solution:

gle-source shortest path algorithm on a complete It can be observed by taking different examples
graph of n vertices? for graph with cycle, clique and planner graph and
(A) Q(n2) (B) Q(n2 log n)
with tree. Then we find that only option (a) is
(C) Q(n3) (D) Q(n3 log n)
correct.
Ans. (a)
(a)A (b)B (c)C (d)D
(GATE 2013: 1 mark) 93. The number of elements that can be sorted in
Q(log n) time using heap sort is
Solution: The complexity for Bellman-Ford = (A) Q(1) (B) Θ( log n )
O(E ×V). Thus, we have
 log n 
n(n - 1) (C) Q   (D) Q(log n)
E = = n2  log log n 
2
V =n (a) A (b) B (c) C (d) D
E × V = n3
Solution:
Heap sort sorts k elements in time Q(k log k)
Ans. (c)
 log n 
91. Which of the following statements are TRUE?
K = Θ  
(1) The problem of determining whether there  log log n 
exists a cycle in an undirected graph is in P.
 log n log n 
(2) The problem of determining whether there Θ(k log k) = Θ  
 log log n log log n 
log
exists a cycle in an undirected graph is in NP.
(3) If a problem A is NP-Complete, there exists  log n 
⇒ Θ  (log log n - log log log n)
a non-deterministic polynomial time algorithm  log log n 
to solve A.
 log n 
(a) 1, 2 and 3 (b) 1 and 2 only ⇒ Θ  (log log n) ⇒ Θ(log n)
(c) 2 and 3 only (d) 1 and 3 only  log log n 
(GATE 2013: 1 mark) Ans. (d)
Solution: 94. Consider the following function:

(1) True. Complexity for detecting cycle using int unknown(int n) {
DFS is O(E+V) = O(V2), which is solvable in int i, j, k = 0;
polynomial time. for (i = n/2; i <= n; i++)
for (j = 2; j <= n; j = j * 2)
(2) True. Every problem in P is also NP.
k = k + n/2;
(3) True. NP-Complete problems are solvable in return k;
non-deterministic time. }
Ans. (a)
92. The line graph L(G) of a simple graph G is defined (a) Q(n2) (b) Q(n2log n)
as follows. (c) Q(n3) (d) Q(n3log n)
There is exactly one vertex v(e) in L(G) for each
edge e in G. Solution:
For any two edges e and e′ in G, L(G) has an edge
between v(e) and v(e′), if and only if e and e′ are for(i=n/2; i<=n; i++) ---- > n/2,n/2+1…
n => n/2 times
incident with the same vertex in G.
{
for(j=2; j<=n; j=j*2) -------→ 2,4,8…
Which of the following statements is/are true?
(P) The line graph of a cycle is a cycle. n => log n times
(Q) The line graph of a clique is a clique. {
(R) The line graph of a planar graph is planar. k = k + n/2;
(S) The line graph of a tree is a tree. }
(a) P only (b) P and R only return k;
}
(c) R only (d) P, Q and S only
Chapter 4.indd 226 4/9/2015 9:55:35 AM

95. Consider the following operation along with What is the worst-case time complexity of a
enqueue and dequeue operations on queues, where sequence of n MultiDequeue() operations on an ini-
k is a global parameter: tially empty queue?
MultiDequeue(Q){ (a) Q(n) (b) Q(n + k)
m = k (c) Q(nk) (d) Q(n2)
while (Q is not empty and m > 0) { (GATE 2013: 2 Marks)
Dequeue(Q)
m = m - 1 Solution: Any combination of n enqueue and
}} dequeue operations will take Q(n).
Ans. (a)
PRACTICE EXERCISES
Set 1
1. What will be the height of a d-ary heap having n 8. The number of nodes in a complete binary tree of
elements? level 5 is —————————.
(a) O(log n ×d) (b) O(d log n) 9. A graph is non-planar if and only if
(c) O(log n/log d) (d) O(log d/log n)
(a) It does not contain subgraphs homeomorphic
2. Quick sort gives O(n log n) worst-case performance to k5 and k3,3.
if the pivot is selected as: (b) It does not contain subgraphs isomorphic to k5
and k3,3.
(a) First element of the array
(c) It does contain subgraph isomorphic to k5 and
(b) Median of first, last and middle elements
k3,3.
(c) Arithmetic mean of the first and last elements
(d) It does contain subgraph homeomorphic to k5
(d) None of the above
and k3,3.
3. A complete full binary tree with 10 leaves
10. To find all pairs of shortest distance in a graph, we
(a) cannot have more than 4 nodes can use
(b) has exactly 19 nodes
(a) dynamic programming approach
(c) has exactly 17 nodes
(b) bidirectional search approach
(d) cannot have more than 19 nodes
(c) merging approach
4. What is the minimum number of edges to be (d) recursion approach
removed from a complete bipartite graph of 6
nodes K(6), so that graph becomes planar? 11. Sorting is useful for
(a) 1 (b) 3 (c) 4 (d) 6 (a) report generation

(b) memory storage
5. A digraph is represented using adjacency matrix, (c) compiler
which is a (d) making searching easier and efficient
(a) Square matrix (b) Symmetric matrix 12. A sorting technique that guarantees that records with
(c) Asymmetric matrix (d) Unit matrix the same primary key occur in the same order in the
6. Merge sort uses sorted list as in the original unsorted list is said to be
(a) divide and conquer approach (a) internal (b) external

(b) heuristic approach (c) stable (d) inconsistent
(c) greedy approach
13. Stack cannot be used to
(d) back tracking approach
(a) evaluate an arithmetic expression in postfix form
7. If B is a circuit matrix of a graph having K compo-
(b) call different subroutines of a program
nents, e edges and n vertices, the rank of B is:
(c) allocate resources(like CPU) by the operating
(a) e + n + k (b) e − n + k system
(c) e − n − k (d) e + n −k (d) implement recursion
Chapter 4.indd 227 4/9/2015 9:55:36 AM

14. A sorting method that uses a lesser number of (c) E

lements in the first half of the array are made
swappings is equal to elements in second half of the array.
(d) The arrays are sorted and form a heap.
(a) quick sort (b) bubble sort
(c) selection sort (d) radix sort 23. An algorithm is made up of two modules M1 and
M2. If time complexities of modules M1 and M2 are
15. Merge sort uses recursion for
h(n) and g(n), respectively, the time complexity
(a) postorder traversal of a tree of the algorithm is
(b) preorder traversal of a tree
(a) Mean(h(n), g(n)) (b) max(h(n), g(n))
(c) inorder traversal of a tree
(c) h(n) + g(n) (d) min(h(n), g(n))
(d) both b and c
16. A binary search tree 24. A vertex cover of an undirected graph G(V, E) is a
subset V1 ⊆ V vertices such that
(a) g ives the descending order for pre order traversal
(b) gives the ascending order for in order traversal (a) Each pair of vertices in V1 is connected by an edge
(c) gives the ascending order for post order traversal (b) If (u, v) ∈E, then u ∈V1 and v ∈V1
(d) gives the ascending order for pre order traversal (c) If (u, v) ∈E, then u ∈V1 or v ∈V1
(d) All pairs of vertices in V1 are not connected by
17. Walk of a graph can be defined as a sequence of an edge
nodes and edges in which
25. Which one of the following is not an informed
(a) edges are not repeated search technique?
(b) edges may be repeated
(c) all nodes may be covered (a) Hill climbing search (b) AO* search
(d) nodes may be repeated (c) A* search (d) Breadth-first search
18. A sorting technique that uses binary tree in which 26. The map colouring problem can be solved using
the larger number acts as a root of a subtree to all which of the following techniques?
the numbers is called
(a) Depth-first search
(a) heap sort (b) quick sort (b) Constraint satisfaction
(c) merge sort (d) bucket sort (c) AO* search
(d) Bidirectional search
19. The reverse polish notation equivalent to the infix
expression ((A + B) ×C + D)/(E + F + G) is 27. The value of postfix expression 8 3 4 + −3 8 2 / +
× 2 3 + is –––—.
(a) AB + C ×D + EF + G + /
(b) AB + CD× + EF G + 28. Which of the following connected simple graphs
(c) AB + C ×D + EF + G + / has exactly one spanning tree?
(d) AB + CD× + E + FG + /
(a) Graph with four nodes and five edges
20. Big-O estimate for the logarithm of the factorial (b) Graph with five nodes and five edges
function, that is, log n!, is given by (c) Graph with five nodes and four edges
(a) O(n log n2) (b) O(n log n) (d) None of the above
(c) O(log n!) (d) O(log n ) 29. Assume that there are n keys and each key is in the
range [0, m −1]. The runtime of bucket sort is
21. If h is a hash function used to hash n keys into a
table of size m, where n < m, then the expected (a) O(m −1) (b) O(m log n)
number of collisions involving a particular key K is (c) O(n log m) (d) O(n + m)
(a) Less than 1 30. Consider f (x) = (x + 1) log(x2 + 1) + 3x2 ; its time
(b) Less than log n complexity in big-O notation is
(c) Greater than log n
(d) Greater than log n + 1 (a) O(x) (b) O(x2)
(c) O(x3) (d) O( n)
22. In merge sort, what process is performed before
merging? 31. How many edges are there in a forest of t trees
containing a total of n vertices?
(a) The array elements form a heap.
(b) Elements in each half of the array are sorted (a) n ×t (b) n −t
among themselves. (c) n/t (d) t/n
Chapter 4.indd 228 4/9/2015 9:55:37 AM

32. Which of the following sorting algorithms is stable? 41. Which of the following is not the in-place sorting?
(a) Quick sort and insertion sort (a) Insertion sort (b) Heap sort
(b) Insertion sort and bubble sort (c) Merge sort (d) Selection sort
(c) Quick sort and heap sort
42. Which of the following does not use the concept of
(d) Quick sort and bubble sort
dynamic programming?
33. The minimum number of edges in a connected
(a) Bellman-Ford algorithm
cyclic graph with n vertices is
(b) Kruskal’s algorithm
(a) n (b) n −1 (c) Dijkstra’s algorithm
(c) n (d) n + 1 (d) Prim’s minimum spanning tree
34. If there exists at least one path between every pair 43. We use dynamic programming approach when
of vertices in a graph, the graph is called (a) It can be implemented on 0—1 knapsack problem
(a) Complete graph (b) Hamiltonian graph (b) The solution has optimal substructure
(c) Connected graph (d) Mesh (c) The given problem can be reduced to the 3-SAT
problem
35. The depth of a complete binary tree with n nodes is (d) The brute-force algorithm is not applicable
(a) log(n + 1) −1 (b) log(n) −1 44. Which of the following algorithms is not a divide
(c) log(n −1) (d) log(n) and conquer algorithm by nature?
36. The maximum number of nodes contained by a (a) Dijkstra’s algorithm (b) Heap sort
binary tree of level l is (c) Selection sort (d) Quick sort
(a) 2l (b) 2l 45. What is recurrence for worst case of quick sort and
(c) l2 (d) log(l −1) what is the time complexity in worst case?
37. f(n) = O(g(n)) implies (a) R ecurrence is T (n) = T (n/2) + Ο(n/2) and
(a) f(n)= O(g(n)) only time complexity is O(n2).
(b) f(n)= W(g(n)) only (b) Recurrence is T (n) = T (n - 1) + O(n) and time
(c) f(n) = O(g(n)) and f(n) = W(g(n)) complexity is O(n2).
(d) f(n) = Q(g(n)) only (c) Recurrence is T (n) = 2T (n / 2) + O(log n) and
time complexity is O(n log n).
38. Assuming P! = NP, which of the following is true? (d) Recurrence is T (n) = T (n / 10) + Ο(n × 9/10)
(a) P = NP and time complexity is O(n3).
(b) NP-complete ∩P = F Set 2
(c) NP-hard = NP
(d) NP-complete = P 1. Map the following statements regarding quick sort
39. Which of the following statements is true? with true (T) or false (F), respectively.
1. The problem of determining whether there exists a S1: Quick sort is an in-place sorting algorithm.
cycle in an undirected graph is in P. S2: Quick sort is stable.
2. The problem of determining whether there exists a S3: Quick sort performs best when numbers are
cycle in an undirected graph is in NP. sorted or almost sorted.
3. If a problem A is NP-complete, there exists a non- S4: The worst running time of quick sort using ran-
deterministic polynomial time algorithm to solve A. domizedpartition is O(n log n).
S5: It is an example of greedy technique.
(a) 1, 2 and 3 (b) 1 and 3
(c) 2 and 3 (d) 1 and 3 (a) TFFFF (b) TTFTT
(c) TFFTF (d) FTFTF
40. Let X is a problem that belongs to the class NP.
Which of the following is true? 2. What will be the output after the second pass of
bubble sort to sort the following number sequence
(a) There is no polynomial time algorithm for X. in an ascending order?
(b) If X can be solved, then P = NP.
(c) If X is NP-hard, then it is NP-complete. 357642
(d) X may be represented by Bachmann-Landau (a) 3 5 6 4 2 7 (b) 3 5 4 6 2 7
notations. (c) 3 5 4 2 6 7 (d) 2 3 4 6 5 7
Chapter 4.indd 229 4/9/2015 9:55:38 AM

3. The number of spanning trees for a complete graph 10. Solve the following given recurrence relation:
with seven vertices is T (n) = 2T (n/2) + n2.
(a) 25 (b) 75 (a) Q(n2) (b) Q(n2 lognn)
(c) 35 (d) 22×5 (c) Q(n log2n) (d) None of the above
4. If BFS is run on a complete bipartite graph Kq,r, q ≥ 1 11. Given

and r ≥ 1, starting with a vertex `X’, then maximum
n = w(n)
1 2
possible height of BFS tree, rooted at `X’, is: (i)
2
(ii) n = w(n )
(a) 1 (b) 2 1 2 2
(c) q + r −1 (d) Max(q, r) 2
Which of the following is true?
Common data questions 5 and 6: Consider the fol- (a) I is correct.
lowing algorithm (b) II is correct.
Add(k){//Assume n is a positive integer values (c) Both I and II are correct.
If(k<=1) (d) None of them is correct.
Return k;
Total=0; 12. f(n)=O(g(n)) if and only if
For i=1 to k-1
Total =Total+Add(i); (a) g(n)= O(f(n)) (b) g(n)= w(f(n))
Return Total;} (c) g(n)= W(f(n)) (d) None
13. What is the generating function G(z) for the

5. If T(k) denotes the time complexity of the algo- sequence of Fibonacci number?
rithm, then T(k) is z z
(a) 2 (b)
(a) Q(n2) (b) Q(n!) z - z -1 1 - z - z2
(c) Q(2n) (d) Q(n3) z z
(c) (d)
1 + z - z2 1 + z + z2
6. If S(k) denotes the space complexity of the algo-
14. ∑ i =0 O(n) , where O(n) stands for order n is
n
rithm, then S(k) (in terms of T(k)) is
(a) Q(log log(T(n))) (a) O(n) (b) O(n2) (c) O(n3) (d) O(3n2)
(b) Q(log(T(n)))
(c ) Q(log(T(n))/log(n)) 15. In the worst case, the number of comparisons
(d) Q (T (n)) needed to search singly linked list of length n for a
given element is
7. If A is the best algorithm to check whether a given
(a) log2n (b) n/2 (c) log2n −1 (d) n
simple undirected graph G(V, E) is a tree or not,
then which of the following best describes the time 16. Consider the following algorithm for searching for
complexity T(V, E) of A? (Assume that G is rep- a given number x in an unsorted array A[1…n]
resented as adjacency list.) having n distinct values:
(a) T(V, E)=O(|V|) 1. Choose an i uniformly at random from 1…. n;
(b) T(V, E)= O(|V|+|E |) 2. If A[i] = x then Stop else Goto 1;
(c) T(V, E)= Q(|V |2) Assuming that x is present on A, what is the
(d) T(V, E) = Qlog(|V|+|E |) expected number of comparisons made by the algo-
rithm before it terminates?
(a) n (b) n −1 (c) 2n (d) n/2
8. Express the following recurrence relation in asymp-
totic notation: T (n) = 4T (n/4) + n .
(a) Q( n ) (b) Q(n) 17. The running time of the following algorithm:
(c) Q(n2) (d) Q(n3/2) Procedure A(n)
If n < = 2 return (1) else return (A( n ))
9. The depth of the binary tree with n nodes is
is best described by
(a) log (n + 1) −1 (b) log n (a) O(n) (b) O(log n)
(c) log(n −1) + 1 (d) log(n) + log(n −1) (c) O(log log n) (d) O(1)
Chapter 4.indd 230 4/9/2015 9:55:41 AM

18. Let P be the quick sort program to sort numbers in 25. A sorting technique is called stable if
ascending order. Let t1 and t2 be the time taken by
(a) it takes O(n log n) time
the program for the input [1 2 3 4] and [5 4 3 2 1],
(b) it maintains the relative order of occurrence of
respectively. Which of the following holds?
non-distinct elements
(a) t1 = t2 (b) t1 > t2 (c) it uses divide and conquer paradigm
(c) t1 < t2 (d) t1 = t2 + 5 log 5 (d) it takes O(n) space
19. Which of the following algorithm(s) can be used to 26. If one uses straight two-way merge sort algorithm
sort n integers in the range [1…n3] in O(n) time? to sort the following elements in ascending order:
20, 47, 25, 8, 9, 4, 40, 30, 12, 17, then the order of
(a) Heap sort (b) Quick sort these elements after second pass of the algorithm is
(c) Merge sort (d) Radix sort
(a) 8, 9, 15, 20, 47, 4, 12, 7, 30, 30
20. The recurrence relation that arises in relation with (b) 8, 15, 20, 47, 4, 9, 30, 40, 12, 17
the complexity of binary search is (c) 15, 20, 47, 4, 8, 9, 12, 30, 40, 17
(a) T (n) = T (n/2) + k, where k is a constant (d) 4 8, 9, 15, 20, 47, 12, 17, 30, 40
(b) T (n) = 2T (n/2) + k, where k is a constant
27. Randomized quick sort is an extension of quick sort
(c) T (n) = T (n/2) + log n where the pivot is chosen randomly. What is the
(d) T (n) = T (n/2) + n worst-case complexity of sorting n numbers using
randomized quick sort?
21. Which one of the following statements is false?
(a) O(n) (b) (n log n)
(a) O ptimal binary search tree construction can be
(c) O(n2) (d) O(n!)
performed efficiently using dynamic programming.
(b) Breadth-first search cannot be used to find con-
28. Kruskal’s algorithm for finding a minimum span-
nected component of a graph.
ning tree of a weighted graph G with n vertices and
(c) Given the prefix and postfix walks over a
m edges has the time complexity of
binary tree, the binary tree cannot be uniquely
constructed. (a) O(n2) (b) O(mn)
(d) Depth-first search can be used to find con- (c) O(m2) (d) O(m log n)
nected components of a graph.
29. A binary search tree is generated by inserting in
22. The recurrence relation T(1) = 2 and T(n) = order the following integers: 50, 15, 62, 5, 20, 58, 91,
T (n) = 3T (n/4) + n has the solution T(n) equal to
3, 8, 37, 60, 24. The number of nodes in the left sub-
(a) O(n) (b) O(log n) tree and right subtree of the root, respectively, is
3/4
(c) O(n ) (d) None of the above (a) (4, 7) (b) (7, 4)
(c) (8, 3) (d) (3, 8)
23. Quick sort is run on two inputs shown below to
sort in ascending order 30. Which one of the following algorithm design tech-
(i) 1, 2, 3, …, n niques is used in finding all pairs of shortest dis-
(ii) n, n −1, n −2, …, 2, 1 tances in a graph?
Let C1 and C2 be the number of comparisons made (a) Dynamic programming (b) Backtracking
for the inputs (i) and (ii), respectively. Then (c) Greedy (d) Divide and conquer
(a) C1 < C2
(b) C1 > C2 31. Give the correct matching for the following pairs:
(c) C1 = C2
(d) We cannot compare for arbitrary n A. O(log n) P. Selection
24. Let T(n) be the function defined by T(1) =1, B. O(n) Q. Insertion sort
T (n) = 2T ([n/2]) + n for n ≥ 2. Which of the
following statements is true? C. O(n log n) R. Binary search
(a) T (n) = O( n ) (c) T (n) = O(log n)
D. O(n2) S. Merge sort
(b) T (n) = O(n) (d) None of the above
Chapter 4.indd 231 4/9/2015 9:55:45 AM

(a) A − R; B − P; C − Q; D − S 33. How many undirected graphs (not necessarily con-

(b) A − R; B − P; C − S; D − Q nected) can be constructed out of a given set V =
(c) A − P; B − R; C − S; D − Q {V1, V2, …, Vn} of n vertices?
(d) A − P; B − S; C − R; D − Q (a) n(n −1)/2 (b) 2n
32. Consider an undirected unweighted graph G. Let a (c) n! (d) 2(n(n−1)/2)
breadth-first traversal of G be done starting from
a node r. Let d(r, u) and d(r, v) be the lengths of 34. The most appropriate matching for the following
the shortest paths from r to u and v, respectively, pairs is:
in G. lf u is visited before v during the breadth-
first traversal, which of the following statements is X: depth-first search 1: heap
correct? Y: breadth-first search 2: queue
(a) d(r, u) <d(r, v) Z: sorting 3: stack
(b) d(r, u) >d(r, v)
(c) d(r, u) ≤d(r, v) (a) X— 1; Y— 2; Z— 3 (b) X— 3; Y— 1; Z— 2
(d) None of the above (c) X— 3; Y— 2; Z— 1 (d) X— 2; Y— 3; Z— 1
Set 1
1. (c) 9. (d) 17. (b) 25. (d) 33. (c) 41. (c)
2. (d) 10. (a) 18. (a) 26. (b) 34. (c) 42. (d)
3. (c) 11. (d) 19. (a) 27. (52) 35. (a) 43. (b)
4. (b) 12. (c) 20. (b) 28. (d) 36. (b) 44. (b)
5. (c) 13. (c) 21. (a) 29. (d) 37. (c) 45. (b)
6. (a) 14. (c) 22. (b) 30. (b) 28. (b)
7. (c) 15. (b) 23. (b) 31. (b) 39. (a)
8. (63) 16. (b) 24. (c) 32. (c) 40. (c)
Set 2
1. (a) Quick sort is an in-place sorting technique but not 4. (b) Let the two sets be Q and R and if X ∈Q, then
stable. It performs worst when numbers are sorted. the BFS tree contains X in level 0, all the nodes of
The worst running time of quick sort using random- R in level 1 and remaining nodes of Q other than X
ized partition is O(n2). Quick sort is an example of a in level 2, so maximum height is 2.
divide-and-conquer algorithm technique.
5. (c) The recurrence relation T(n) for time complex-
2. (c) ity of linear(n) is
k -1
Pass 1: Pass 2: After second T (k) = c + ∑ T (i), if k > 1 = c
pass: i =1
3 5 7 6 4 2 3 5 6 4 2 7 354267 = else
3 5 7 6 4 2 3 5 6 4 2 7 k -2
3 5 6 7 4 2 3 5 4 6 2 7 T (k) = c + ∑ T (i) + T (k - 1), for k > 1
3 5 6 4 2 7 3 5 4 2 6 7 i =1
T (k) = T (k - 1) + T (k - 1) = 2 × T (k - 1)
3. (b) The number of spanning trees for Kn is nn −2 =
So
77−2 = 75.
T (k) = 2k -1 c
Chapter 4.indd 232 4/9/2015 9:55:46 AM

6. (b) Space complexity of large(k) is proportional Case 2: If A[i] is found in the second attempt,
to the height of recursion tree which is k. Thus, number of comparisons = 2. Thus, probability of
S(k) = k = Q(log(T (k)), since T(k) = Q(2k). the case is given as
n -1 1
7. (a) Can be done using DFS in O(|V |) time. ×
n n
8. (b) By case 1 of the Master Theorem Case 3: If A[i] is found in the third attempt,
9. (a) The depth of a binary tree with n nodes is number of comparisons = 3. Thus, probability of
the case is given as
2d +1 - 1 = N n -1 n -1 1
× ×
So, we have n n n
2d +1 = N + 1 There are actually infinite such cases. So, we have
Taking log, we get the following infinite series:
1 n -1 1 n -1 n -1 1
(d + 1) log 2 = (N + 1) + × ×2 + f = × × ×3 +
n n n n n n
d + 1 = log2 (N + 1)
After solving this series, we will get `n’ comparisons.
d = log2 (N + 1) - 1 17. (c) Recursive relation for procedure A(n) is
10. (a) From the master theorem, here we have a = 2, T (n) = T ( n ) + c1 if n > 2
b = 2 and f(n) = n2. So
Use substitution method to solve this.
g(n) = n log b⋅a = n
18. (a) Quick sort takes the worst time if the elements
Here, order of f(n) > order of g(n), so we get are already in sorted orders (it could be in ascend-
T (n) = O(f (n)) = O(n2 ) ≈ Q(n2 )
ing or descending). As both the inputs are already
sorted, it will take the same time in both the cases.
11. (a) For w, there is f(n) > cg(n). So there is no 19. (d) Quick sort, merge sort and heap sort running
equality. Thus, I is correct and II is false. time are (n log n), whereas radix sort is a non-
12. (c) We have f (n) = O(g(n)) which means comparative integer sorting algorithm that sorts
f (n) < cg(n) data in O(n) time.
It is true only if g(n) >c(f(n)) which means 20. (a) In binary search, only half of the array where
g(n) = W(f (n)) the element exists is searched and the other half is
left. So, we get
13. (b) The generating function for the Fibonacci
n
T   + K
number G(z) is
 2 
z
G(z) =
1 - z - z2 21. (b) Depth-first search and breadth-first search are
capable of finding whether a given graph is con-
14. (b) We have
nected and are also capable of finding connected
n
∑ O(n) = O(1) + O(2) + O(3) + + O(n)
components of a graph in linear time. So, option
(b) is false.
22. (a) Using case 1 of master theorem, we get n log 4 3 < n
i=0
 n(n + 1
= O 
 2 
= O(n2 )
23. (c) Both of the given cases are quick sort worst-
case problems, so the number of comparisons will
15. (d) The worst-case search time in a singly linked be equal.
list is when the element is in the last position or
the element does not exist in the linked list. So it 24. (b) Use case 1 of master theorem.
is similar to linear search and requires `n’ compari- 25. (b) A sorting algorithm is said to be stable if two
sons in the worst case. objects with equal keys appear in the same order
16. (a) (Number of comparisons for a case × probabil- in the sorted output as they appear in the input
ity for the case) unsorted array. Some sorting algorithms are stable
by nature such as insertion sort, merge sort, bubble
Case 1: If A[i] is found in the first attempt, number of sort, etc. And some sorting algorithms are not,
comparisons = 1. Thus, probability of the case = 1/n. such as heap sort, quick sort, etc.
Chapter 4.indd 233 4/9/2015 9:55:51 AM

26. (b) Given: 20, 47, 15, 8, 9, 4, 40, 30, 12, 17. The number of nodes in the left subtree and the
right subtree of the root, respectively, is (7, 4).
Here, two-way merge sort is used, so a group of Therefore, option (b) is correct.
two is taken at once. The second pass is shown as
follows: 30. (a) Bellman-Ford algorithm is used to find all-pair
8, 15, 20, 47 4, 9, 30, 40 12, 17 shortest distances in a graph, and it is a dynamic
programming technique.
20, 47 8, 45 4, 9 30, 40 12, 17 31. (b) Merge sort took O(n log n) to sort a list of ele-
ment. Binary search time complexity is O(log n).
20, 47 15, 8 9, 4 40, 30 12, 17 Insertion sort worst-case running time complexity
is O(n2 ) .
After second pass of the algorithm element, the
32. (c) d(r, u) and d(r, v) will be equal when u and
order is 8, 15, 20, 47, 4, 9, 30, 40, 12, 17.
v are at same level; otherwise, d(r, u) will be less
27. (c) Randomized quick sort worst-case complexity
than d(r, v).
is O(n2 ); when all the elements are same, it took
worst time. 33. (d) In an undirected graph, if it is a complete
graph, then it has n(n −1)/2 edges. We can choose
28. (d)
to have (or not have) any of the n(n −1)/2 edges.
29. (b) Look at the following figure: So, the total number of undirected graphs with n
vertices is 2(n(n−1)/2).
50 34. (c) Depth-first search uses stack for storing values;

queue is used by breadth-first search. Heap is used
for sorting the set of elements.
15 62
5 20 58 91
3 8 37 60
24
Chapter 4.indd 234 4/9/2015 9:55:53 AM

UNIT V: THEORY OF COMPUTATION
6
Number of questions
4 Marks 1
Marks 2
2
0
2015 2014 2013 2012 2011 2010
Year Concepts
2015 Turing machine, Regular expression and languages, DFA,
Context free grammar, Recursive language, NDPDA
2014 Regular languages, DFA, RE and REC
2013 Regular languages, DFA, CFL, CFG
2012 Regular languages, DFA and NFA, CFL
2011 Power of DFA, NFA and TM, Regular language, PDA
2010 Regular expression, Recursive language, LR and LL, CFL
Chapter 5.indd 235 4/9/2015 9:57:44 AM

Chapter 5.indd 236 4/9/2015 9:57:44 AM
CHAPTER 5
THEORY OF COMPUTATION
Syllabus: Regular languages and finite automata, context-free languages and pushdown automata, recursively
enumerable sets and Turing machines, undecidability.
5.1 INTRODUCTION the question of the extent to which a problem is solv-

able on a computer, whereas complexity theory consid-
ers not only whether a problem can be solved at all on
Theory of computation is a branch of computer science a computer, but also how efficiently it can be solved.
and mathematics that studies whether a problem can Two major aspects are considered in this—time complex-
be solved using an algorithm on a model of computa- ity and space complexity.
tion. This branch also studies that how efficiently that Basic introduction and various technical terms for
problem can be solved. This deals with two types of understanding of concept are described in the following
theories. First is computability theory, deals with up section.
to which extent problem can be solved. Second is com-
plexity theory, which deals with efficiency of solution.
In this finite automata, push down automata, Turing 5.2 FINITE AUTOMATA
machines, regular expressions and various grammars will
be discussed.
Automata theory is the study of abstract machines (or There are several things to be done in the designing and
more appropriately, abstract “mathematical” machines implementation of an automatic machine, but the most
or systems) and the computational problems that can be important question is that, what will be the behaviour
solved using these machines. These abstract machines of the machine? “Theory of Computation” is the subject
are called automata. Computability theory deals with which solves this purpose.
Chapter 5.indd 237 4/9/2015 9:57:44 AM

238 Chapter 5: Theory of Computation
“Automata theory is the subject which describes the where e is called the identity element. Here, Σ*
behaviour of automatic machines mathematically”. In contains every string possible by a, b.
other words, we can say that “Automata Theory is the
Σ+ = Σ* − {Null character (l)}, so Σ+ will con-
study of self-operating virtual machines to help in logical
tain {a, b, aa, bb, ab, …}.
understanding of input and output process without or
with intermediate stages of computation”.
2. Strings: Sequences of zero or more symbols are
The model of automation is shown in Fig. 5.1.
called string of a language. Let there be languages
such as
Finite
automation L1 = a *, then it will consist {e, a, aa, aaa, aaaa, …}
Input lines Output lines string set.
with n states
(q1, q2, ..., qn) L2 = ab*, then the string will be {a, ab, abb, abbb, …}.
L3 = 0*1*, then it will consist {e, 0, 1, 01, 001,
0001, 011, 0111, …} string set.
Figure 5.1 | Model of automation.
3. Concatenation: Let there be two strings U = ab
5.2.1 Finite Automaton Model Characteristics and V = aab. Then concatenation of these two
languages is UV = abaab and VU = aabab.
The characteristics of finite automaton are described below: Concatenations of two strings satisfy associative
property but not commutative. So, UV ≠ VU.
1. Input: Input values from input alphabet Σ con- 4. Length of a string: Let U, V and W be three
tains I1, I2, I3, … In at discrete time. are the input strings. The length of a string tells the number of
values from the input alphabet Σ at discrete time. alphabets in that string.
2. Output: Output O1, O2, O3, …; On is the output
set for this computation model are different out- If U = 011, length of |U | = 3.
puts of this model. If V = ababa, length of |V | = 5.
3. States: At any instance of time, the finite auto If W = abababb, length of |W | = 7.
maton will be in one of the states q1, q2, q3, …, qn, 5. Reversal of a string: Let there be a string W = aab,
are the set of states on which the finite automa- then the reverse of W is
ton will be at some instance of time. W r = baa
4. States relation: State relation describes that how to If W = W r, then the string is a palindrome. WW r is
reach on next state using present input and output. called an even palindrome, and W × W r is called an
At any instance of time, the next state of automaton odd palindrome, where x is an alphabet. Let there
is determined by the present input and present state. be two strings U and V, then the reverse of UV is
5. Output relation: Either one state or number of
states can be obtained as output. The next state is |UV |r = V rU r
achieved using state relation. Either only state or 6. Power of a string: Let there be a string W, then
both the input and state are obtained as output. W0 = e
When an automaton reads an input symbol, it moves W1 = W
to the next state, which is given by the state relation. W2 = W ⋅ W
W 3 = W ⋅ W 2, so W n + 1= W ⋅ W n = W n ⋅ W
5.2.2 Technical Terms Let there be a string U = {010}, then the power of U:
Technical terms that are required in finite automaton U0 = e
are given as following: U 1 = {010}
U 2 = {010010}
1. Alphabets: A finite non-empty set of symbols are
the alphabets of a language. 7. Substring: Any set of conjugative symbols is
called a substring.
Example 5.1
Example 5.2
Σ = {a, b}
Let W = {abc} is a string. Substrings of W are (e, a, b,
Π = {0, 1} c, ab, bc, abc), where e is a substring of every string. If
the length of a string is n, then the number of substrings
Σ and Π are the sets which hold two alphabets. They
possible is
are called binary alphabets for language. n(n + 1)
+1
(a + b)* = Σ* = {e , a, b, aa, bb, ab, ba, ¼} 2
Chapter 5.indd 238 4/9/2015 9:57:45 AM

5.2 FINITE AUTOMATA 239
8. Prefix of a string: Prefix of a string W = {x | xy = 5.2.3 Grammar

W}
Grammar is a set of rules which generates every string
Consider a string W = {aabbcd}. in a language. A grammar is described by four terms
Prefix of W = {e, a, aa, aab, aabb, aabbc, aabbcd} G = (V, T, S, P), where
If the length of a string is n, then the number of
V → finite non-empty set of variables (represented by
prefix of that string will be (n + 1).
capital letters)
9. Suffix of a string: Suffix of a string W = {y | xy = W} S → starting symbol
Consider a string W = {aabbcd}. T → finite non-empty set of terminals (represented by
Suffix of W = {aabbcd, abbcd, bbcd, bcd, cd, d, e} small letters)
If the length of string is n, then the number of P → finite non-empty set of production rule
suffix of that string will be (n + 1).
10. Language: A language is a collection of strings of Example 5.4
finite length constructed from alphabets of symbol.
Or, we can say that a language is a subset of every S → AB (production 1)
string made by selected alphabets. A → a (production 2)
B → b (production 3)
Let an alphabet set be {0, 1}, then Σ* will con-
tain all the strings made by 0 and 1. So language where S is the starting point, A and B are variables and
made by 0 and 1 will be the subset of Σ*. Language a, b are terminals. We have three productions here. In
L ⊆ Σ*. grammar, V, T, P should be non-empty and finite.
11. Language union: Let L1 = {ab, ba, aab} and
L2 = {ab, abb, bb} be two languages, then L1 ∪ L2 5.2.4 Noam Chomsky Grammar Classification
= {ab, ba, aab, abb, bb}.
12. Language intersections: Let L1 = {ab, ba, aab} Grammar is classified into the following four types:
and L2 = {ab, abb, bb} be two languages, then 1. Type 0: This type of grammar is unrestricted, or
L1 ∩ L2 = {ab}. phase structured, which generates all types of lan-
13. Language complement: Let L be a language and guages that can be recognized by a Turing machine.
L¢ the complement of language L. So, L¢ = {Σ* − L} These languages are called recursive enumerable
language. Unrestricted means these grammar have
very less restrictions. Unrestricted grammar (G) =
Example 5.3 u → v, where u, v ∈ (V ∪ T)* and “u” must have
Given language L as {aa, ab, baa}, then complement of
at least one variable. V stands for variable and T
language L (L¢) = {a, b}* − {aa, ab, baa}.
for terminals.
Example 5.5
14. Language subtractions: Let L1 = {ab, ba, baa}
and L2 = {aab, ba} be two languages. S → aSb/l
A → aA/aB/Cb
Then L1 − L2 = {ab, baa} and L2 − L1 = {aab}.
Ba → a/Da
15. Language XOR: Let L1 = −ab, ba, baa} and L2
= −aab, ba} be two languages.
2. Type 1: This is context-sensitive grammar (CSG)
L1 ⊕ L2 = (L1 − L2) ∪ (L2 − L1) = (L1 ∪ L2) − which is used to generate context-sensitive lan-
(L1 ∩ L2) guages. CSG (G) = u → v, where u, v ∈ (V ∪ T)*
L1 ⊕ L2 = {ab, baa, aab} and “u” must have at least one variable. There
is one additional rule also, u ≤ v , which means
16. L1⋅L2: Let L1 = {ab, ba, aab} and L2 = {ba, ab}, length of “u” is less than or equal to “v”.
then
L1⋅L2 = {abba, abab, baba, baab, aabba, aabab} Example 5.6
S → aSb/l
+
17. L* and L :
L*= {L ∪ L ∪ L ∪ L …}
0 1 2 3 Aa → aA/aB/Cb
L+= {L* − l} B → a/Da
Chapter 5.indd 239 4/9/2015 9:57:46 AM

3. Type 2: This is context-free grammar (CFG) which

is used to generate context-free languages. CFG (G)
= u → v, where v ∈ (V ∪ T)* and | u | ≤ | v |. In
All languages
addition to the CSG grammar, one extra rule, that
is, only single variable is allowed in “u”. RE
Example 5.7 REC

S → aSb/l
A → aA/aB/Cb CSL
B → a/Da
CFL
4. Type 3: This is regular grammar, therefore lan-

guage generated by these grammars is called reg- REG
ular languages. Regular grammar (G) = u → v,
where u ≤ v , {v ∈ (T*, T*V, VT*) and u ∈ V}
and “u” has only one variable.
REG–regular language; CFL–context-free language;
Example 5.8 CSL–context-sensitive language; REC–recursive lan-
S → aS/l
guage; RE–recursive enumerable.
A → aA/aB/Cb Figure 5.2 | Chomsky hierarchy.
B → a/Da
5.2.6 Different Grammar Types

5.2.5 Chomsky Hierarchy
There have been numerous grammar types proposed for
The hierarchy of grammar was described by Noam different languages. Table 5.1 provides a summary of the
Chomsky in 1956, shown in Fig. 5.2. same.
Table 5.1 | Comparison of different types of grammar
Type Grammar Accepted by Restriction on Production (x → y) Example

Type 0 Unrestricted Turing Machine 1. x must have at least one variable (V) AB → a | Ba
Grammar 2. y can be any string which consist any
combination of variable (V) and terminal (T)
Type 1 Context Linear Bounded 1. x must have at least one variable (V) Aa → aAb | aB | Cb
Sensitive Automata 2. y can be any string which consist combination
Grammar of variable (V) and terminal (T)
3. Length of y should be less than or equal to x.
Type 2 Context Free Push Down 1. x can have exactly one Variable (V) A → aBa | aB | Ba
Grammar Automata 2. y can be any string which consist combination
of variable (V) and terminal (T)
Type 3 Regular Finite Automata 1. x can have exactly one Variable (V) A → a | Bb | bB
Grammar 2. y can be any string which consist any
combination from (VT*, T*V, T*), where V =
one variable and T* = any number of terminals
Chapter 5.indd 240 4/9/2015 9:57:47 AM

5.3 FINITE AUTOMATA AND REGULAR LANGUAGE 241
5.3 FINITE AUTOMATA AND 1. Q is a finite non-empty set of states.

REGULAR LANGUAGE 2. Σ is a finite non-empty set of input called alphabets.
3. d is a transition function which maps Q × (Σ ∪
{l}) → 2Q.
Finite automata is classified into two categories— 4. q0 ∈ Q, known as the initial state or starting state.
deterministic finite automata and non-deterministic 5. F ⊆ Q, known as a set of final states.
finite automata.
Example 5.10
5.3.1 Deterministic Finite Automata
Consider an NFA M = ({q0, q1, q2} {a, b} {d} {q0} {q2})
A deterministic finite automata (DFA) is defined by a as shown in figure below.
five-tuple set (Q, Σ, d, q0, F), where
a, b a, b
1. Q is a finite non-empty set of states. a, b a
2. Σ is a finite non-empty set of input called alphabets. q0 q1 q2
3. d is a transition function which maps Q × Σ → Q.
4. q0 ∈ Q, known as the initial state or starting state.
The NFA can be represented in a transition table
5. F ⊆ Q, known as a set of final states.
shown below, which can be determined by the transition
function (d ).
Example 5.9
δ a b
Consider a DFA M = ({q0, q1, q2, q3} {a, b} {d } {q0} q0 q1 q1
{q3}) as shown in the figure below.
a q1 {q1, q2} q1
a a q2 q2
q0 q1 q3 q2
b
q0 — Starting state
b b
q2 — Final state (accepting state)
q2
a, b
5.3.3 Comparison of DFA and NFA
The DFA can be represented in a transition table
shown below, which can be determined by the transition The comparison between DFA and NFA is given in
function (d ). Table 5.2.
δ a b 5.3.4 DFA and NFA Design

q0 q1 q2 We have taken a few examples and discussed how to
q1 q1 q3 design a DFA and an NFA for these cases. The following
are four types of states which are used to design a DFA
q2 q2 q2 and an NFA:
q3 q1 q3 1. Permanent accept state: In this case, the initial
and final states are the same. Due to self-loop this
q0 — Starting state machine will always be on the same state, which
is also final state. Hence, this is called permanent
q3 — Final state (accepting state)
accept state.
5.3.2 Non-deterministic Finite Automata
A non-deterministic finite automata (NFA) is defined by

a five-tuple set (Q, Σ, d, q0, F ), where
Chapter 5.indd 241 4/9/2015 9:57:49 AM

Table 5.2 | DFA vs. NFA
Title DFA NFA

Power NFA and DFA powers are same NFA and DFA powers are same
Supremacy Some NFA are DFA, but not all All DFA are NFA
Time complexity Time needed to execute an input string is Time needed to execute an input string is more
less than that required by NFA than that required by DFA
Transition function Maps Q × Σ → Q, number of next states Maps Q × (Σ ∪ {l}) → 2Q, the number of
at an input is exactly one next states at an input is zero, one or more.
Null moves Does not allow null moves Allows null moves
Moves Only one move for single input alphabet There can be choice (more than one move
can be there) for single input alphabet
2. Permanent reject state (trap state): The given Example 5.11
state is permanent reject state because there is self-
DFA which accepts string starting with “a”
loop to some non-final state.
a, b
q0 a q1
q3
a, b
Regular expression: a(a + b)*
Example 5.12
3. Temporary reject state: The given state is tem- DFA which accepts string ending with “a”
porary reject state because for input `b’, it has a loop
on the non-final state. However, if after input `b’ b a
a
there is another input à’, then it will halt on final q0 q1
state. b
Regular expression: (a + b)*a
b
a Example 5.13
DFA which accepts string starting with “a” and ending
with “b”
a
b
a b
4. Temporary accepting state: The given state is q0 q1 q3
a
temporary accepting state, as it reaches the final
state if it has input symbol à’, but if it gets another b
input symbol `b’ on that final state it will halt on
non-final state. q2
a, b
a
Regular expression:
1. a(a + b)* ∩ (a + b)*b (from statement)
b 2. aa*b(b + aa*b)*
Chapter 5.indd 242 4/9/2015 9:57:50 AM

Example 5.14 5.3.5 Applications of DFA/NFA

DFA which accepts string starting with “001” NFA and DFA have various applications as given below:
1. Lexical analyser
0, 1
0 0 1 2. Text editor
q0 q1 q3 q4
3. Spell checker
1 1 4. Sequential circuit design
0 5. Unix graph (pattern search)
q5
0, 1 5.3.6 Regular Expression and Grammar
Regular expression: 001(0 + 1)*
Finite automata accepts regular grammar. First of
all, regular expressions are discussed and then how to
construct regular grammar from finite automaton is
Example 5.15 described (Table 5.3).
1. Regular expression: A language is regular iff ∃

DFA which accepts string ending with “001”
a regular expression r, that is, L = L(r).
0
• There can be more than one regular expression
1
0 0 1
q0 q1 q2 q3 for a language L.
1 0 • There can be more than one machine for a lan-
guage L.
1 • A language L can be produced by more than one
grammar.
Regular expression: (0 + 1)*001 Symbols to represent regular expression are {a, b, e, Σ,
∅, l, +, . , *, ( )}.
Table 5.3 | Some standard regular languages and their regular expression, grammar and machine
Regular Language (L) Grammar (G) Machine (M)

Expression (r)
∅ {} S→A
a, b
a, b
Λ {l} S→e a, b
a, b
A {a} S→a a a, b
b
a, b
B {b} S→b b a, b
(Continued)
Chapter 5.indd 243 4/9/2015 9:57:52 AM

Table 5.3| Continued
Regular Language (L) Grammar (G) Machine (M)

Expression (r)
a+b {a, b} S→a a, b
S→b
a, b a, b
a + b + ab {a, b, ab} S → aab

S → ab a a a, b
b
S → ba
b a, b
a* {l, a, aa, aaa, aaaa, S → aS a

…} a, b
b
S→e
a+ {a, aa, aaa, aaaa, …} S → aS a

a, b
a
S→a b
a* + b* {l, a, aa, aaa, …, b, S → S1/S2 a

bb, bbb, …}
S1 → aS1/e a b a, b
S2 → bS2/e
b a
b
(a + b)* {l, a, b, ab, ba, aab, S → aS/bS a, b
aaab, baa, bab, bbb,
baba, …} S → a/b
2. Regular grammar: Type 3 grammar is also called 5.3.7 Pumping Lemma
regular grammar and it generates regular language.
Regular grammar is divided into two types: Pumping lemma is very useful to prove that a certain
language is not a regular language. We know that the
Left linear grammar: V → VT * + T *
loop in finite automata (FA) makes it able to accept
Right linear grammar: V → T *V + T * those strings which have length equal to and greater
A grammar is a regular grammar iff ∃ a left linear than its total number of states.
or right linear grammar for it.
Chapter 5.indd 244 4/9/2015 9:57:53 AM

Let there be a string w which has a bigger length (more There are two types of FSMs:
than its states), then we can break this string into three 1. Mealy machines
parts x, y (y should not be null) and z. Let FA has loop 2. Moore machines
for y and w = xyz ∈ L accepted by FA, so w = xyiz for
i = 0, 1, … is also accepted by FA.
5.3.9.1 Mealy Machines
5.3.8 Myhill—Nerode Theorem
In a Mealy machine, the output of an FSM is dependent
Myhill—Nerode theorem tells that the given language on the present state and present input. A Mealy machine
might be regular or not regular. A given language L can be described by a six-tuple set (Q, Σ, Δ, d, l, S ),
is a regular language if and only if the set of equiva- where
lence classes of L is finite, or it satisfies following three 1. Q is a finite and non-empty set of states.
conditions: 2. Σ is an input alphabet.
1. The language L divides the set of all possible strings 3. Δ is an output alphabet.
into mutually separate classes. 4. d is a transition function which maps the pres-
2. If L is a regular language, then the number of ent state and input symbol on to the next state
classes created is finite. Q × Σ → Q.
3. If the number of classes that L creates is finite, 5. l is the output function which maps Q × Σ → Δ
then L is a regular language. ((present state + present symbol) → output).
6. S ∈ Q is the starting state or initial state.
Problem 5.1: Consider a language L consists all
strings over {a, b} ending in b. Prove that L is a Problem 5.2: Consider the Mealy machine shown in
regular language. the figure. Construct the transition table and find the
Solution: Let DFA M = {Q, S, d, q0, F} recognizes output for input aabba.
L = {b, bb, ab, aab, abb, …}.
a/0
We have two classes C0 and C1 defined as follows: a/0 b/1
q0 q1
C0 = {All strings that end in a}
C1 = {All strings that end in b, i.e., accepted
strings (b, ab, abb, bbb, …)}
b/1
Transition diagram for the DFA M is shown below. a/1
a
b
b q2
q0 q1
a b/1
Classes C0 and C1 correspond to states q0 and q1,

respectively. Applying Myhill—Nerode theorem: Solution: Transition table for the above Mealy
machine is:
1. Two classes C0 and C1 are mutually exclusive.
2. L is regular and creates finite classes (here two Present Next State Output
classes). State
3. The number of classes created by DFA is finite. x=a x=b x=a x=b
As all the three statements of Myhill—Nerode theorem q0 q1 q2 0 1
are satisfied, hence L is regular.
q1 q1 q1 0 1
q2 q1 q2 1 1
5.3.9 Transducer or Finite State Machine
Input: aabba
A finite state machine (FSM) is similar to FA except
Output: 00110
that it has the additional capability of producing an
output. (Note: Output length of the Mealy machine is equal
to the input length.)
FSM = FA + Output capability
Chapter 5.indd 245 4/9/2015 9:57:53 AM

5.3.9.2 Moore Machines 5.3.10.1 NFA to DFA
In Moore machine, the output of an FSM is dependent To convert NFA to DFA, following points should be
on the present state only. A Moore machine can be considered:
described by a six-tuple set (Q, Σ, Δ, d, l, S), where 1. In NFA and DFA there is strictly one initial or
1. Q is a finite and non-empty set of states. starting state.
2. Σ is an input alphabet. 2. If there are n states in NFA, then equivalent DFA
3. Δ is an output alphabet. contains ≤ 2n states.
4. d is a transition function which maps the pres- 3. DFA can simulate the behaviour of NFA by
ent state and input symbol on to the next state increasing the number of states.
Q × Σ → Q.
5. l is the output function which maps Q → Δ (pres- Problem 5.4: Construct a DFA equivalent of given NFA.
ent state → output)
a
6. S ∈ Q is the starting state or initial state. b
a
q0 q1
Problem 5.3: Consider the Moore machine shown in
the figure. Construct the transition table and find the a
output for input aabba.
b b b b
a b
q0 q1
Output = 0 Output = 1 q2
a
Solution:
Step 1: Construct a transition table of the given NFA
q2
Output = 0
a, b δ a b
Solution: Transition table for the above Mealy q0 {q0, q1} q2

machine:
q1 q0 q1
Present Next State Output
State
q2 — {q0, q1}
x=a x=b
q0 q1 q0 0
q0 — Starting state
q1 q2 q1 1
q2 q2 q2 0 q2 — Final state (accepting state)
Step 2: C
onstruct a DFA transition table with the
Input: aabba
help of an NFA transition table. Put the first
Output: 010000 row same from the NFA transition table,
(Note: Output length of Moore machine is one then put one by one states from the right
greater than the input length because the first output side. After the first row, put {q0q1} in the
symbol is additional without reading any symbol from transition column and find the next state on
the input.) a from q0 and q1 separately and write them
under column a, repeat the same process for b.
Now pick another state q2 from the right side
5.3.10 Conversion in Finite Automata and follow the same process. This process has
to be repeated until all the unique value set
Power of NFA and DFA is the same, which means comes as a state under the transition table.
both can accept the same set of strings. Here we discuss
Now make the starting state of DFA the same state
interconversion of NFA and DFA.
as in NFA. And make all those states final which
Chapter 5.indd 246 4/9/2015 9:57:54 AM

include the final state of NFA; in our example it was

Solution:
q2, so q1q2 and q2 states will be final states in DFA.
Step 1: Make two sets of final and non-final states.
δ a b
q0 {q0q1} q2 Iteration 1 = ({A, D, E} {B, C})
{q0q1} {q0q1} {q1q2} Step 2: C

heck within the set if two of them are
equivalent states or not.
φ {q0q1}
Let us first check in {A, D, E}, first we pick A and D.
q2
{q1q2} q0 {q0q1}
φ φ φ Present State Next State on Input
Represent starting state X=0 X=1

Represent final or accepting state A B D
D D E
Step 3: Draw DFA from the transition table.
E E E
a
q0 a q0q1 We can clearly see that state A is going outside on

b state B at input 0, which is in another set, whereas
b b b the other two states are indicating within the non-
final state set. So, we will put state A in a new set.
a
q2 q1q2 Iteration 2 = ({A} {D, E} {B, C})
Repeat step 2 until last two iterations become same.
a
Now check for B and C.
φ Present State Next State on Input

a, b
X=0 X=1
B B C
5.3.10.2 DFA to Minimum DFA C C B
Minimization of DFA means reducing the unnecessary states.

During minimization, the following steps are considered. States B and C are not going outside from their own
1. DFA with n states when converted into minimal set at inputs 0 and 1, so there will be no change.
DFA will contain N states, where 1 ≤ N ≤ n.
2. We have to eliminate the same states from the Iteration 3 = ({A} {D, E} {B, C})
given DFA to convert it into minimal DFA.
3. Two states q1 ≡ q2 iff ∀ (w ∈ Σ*), {d(q1, w), d(q2, w)}
We get two same iterations so we will stop this pro-
cess. We will merge those states which lie in a set.
both are from the same set.
There are three states in minimal DFA (A), (DE) and
4. If a final state (accepting state) is equivalent to a
(BC ). Now we can draw the minimal DFA.
non-final state, then it cannot be reduced.
Problem 5.5: Convert the given DFA into minimal DFA.

0, 1
0
0 0 A BC
0 1
A B C
1
1
1
D E DE
1 0, 1
0 0, 1
Chapter 5.indd 247 4/9/2015 9:57:55 AM

5.3.10.3 Mealy to Moore Machine Now we will make a transition table of the Moore
machine.
In a Mealy machine, the output depends upon both the
present state and the input, and the Moore machine Present Next State Output
output depends only on the present state. While con- State
verting Mealy to Moore we develop a procedure so that
x=0 x=1
all the states of Mealy machines, which are associated
with different outputs, find them. After that, split those →q1 q1 q21 0
states into n parts if the states have n outputs. q20 q21 q20 0
q21 q21 q20 1
qi0 State qi associated
with output 0 Moore machine of 2’s complement:
For any state qi qi1 State qi associated with 0 0

output 1
qin State qi associated with q1 q21
output n
Output = 0 Output = 1
Through this procedure all the states are associated with 1
0
a single output, so all the states are considered as unique
state. We can understand this concept clearly with an
example. q20
Output = 0
Problem 5.6: Consider the 2’s complement Mealy 1
machine given below and convert it into Moore Note: m states, n outputs Mealy machine ≡ Moore
machine. At input “1101”, the output is “0011”. machine contains ≤ mn + 1 state.
Input string will be read from the right side.
0/0 0/1 5.3.10.4 Moore to Mealy Machine

1/0
q1 q2 Consider the Moore machine given in Fig. 5.3 and con-
1/1
vert it into Mealy machine.
Solution: First we will make a transition table for the Output = 0
given Mealy machine. a, b
a
q0 q1
State Output = 1
b b
x=0 x=1 x=0 x=1
q1 q1 q2 0 1
q2 q2 q2 1 0 q2 a
Output = 0
Figure 5.3 | Moore machine.
In the transition table, we have to find those states
which are associated with more than one output in
the next state column. In the transition table of Moore machine (Table 5.4),
For q1 at (q1, 0), we have output 0. the states (q0, q1, q2) have outputs (0, 1, 0), respectively.
For state q2 at (q2, 0) output is 1 and at (q2, 1) output Table 5.4 | Transition table of Moore machine
is 0. So there is a need to split q2.
q20 State
q2
x=a x=b
q21
→q0 q1 q2 0
Now the Moore machine has three states {q1, q20, q21} q1 q1 q1 1
with output {0, 0, 1}, respectively. q2 q2 q1 0
Chapter 5.indd 248 4/9/2015 9:57:56 AM

ollow above steps to remove first e move

So, now we can make a transition table for the Mealy
Step 1: F
machine (Table 5.5).
between q0 and q1.
Table 5.5 | Transition table of Mealy machine
0 1 0
q0 1 q1 0 q2
State
x=a x=b x=a x=b e
→q0 q1 q2 1 0
q1 q1 q1 1 1 ollow above steps to remove e move
Step 2: F
q2 q2 q1 between q1 and q2.
0 1
0 1 0
Mealy machine equivalent to given Moore machine
(Fig. 5.4): q0 1 q1 0 q2
ε
a/1
q0 a/1 b/1
q1
ollow above steps to remove e move
Step 3: F
between q0 and q2.
b/0
b/1 0 1 0
q0 1 q1 0 q2
q2
a/0 0
Figure 5.4 | Mealy machine. After removing all e moves, this is the final NFA.
Note: m states, n outputs Moore machine ≡ Mealy Note: If NFA with e-moves has n states then NFA
machine contains ≤ m states. without e-moves will have exactly n states.
5.3.10.5 NFA with e-Moves to NFA

Without e-Moves 5.3.11 Properties of Regular Sets/Languages
There are three steps to convert NFA with e moves to 1. Classes of two regular sets S1 and S2 are closed
NFA without e moves: under union operation, means S1 ∪ S2 is also a
1. If there is an e move from u to v, then every arrow regular set.
starting from v is also starting from u. 2. Classes of two regular sets S1 and S2 are closed
under concatenation operation, means S1S2 is also
ε
u v a regular set.
3. A class of regular set S1 is closed under Kleene
2. When e move removes from u to v, and if u is a closure operation, means S1* is also a regular set.
starting state, then make v also a starting state. 4. If S is a regular set on some alphabet Σ, then
3. When e move removes from u to v, and if v is a complement of S is denoted by S is also a regular
final state, then make u also a final state. set. Steps to make complement of an NFA are as
follows:
• Change all final states to non-final states of
Problem 5.7: Convert given NFA with e moves to NFA for Σ.
NFA without e moves. • Change all non-final states to final states of
Solution: NFA for Σ.
5. Classes of two regular sets S1 and S2 are closed
0 1 0 under intersection operation, means S1 ∩ S2 is also
e e a regular set.
q0 q1 q2
6. If S is a regular set on some alphabet Σ, then trans-
pose of S is denoted by SR is also a regular set.
Chapter 5.indd 249 4/9/2015 9:57:58 AM

5.3.12 Some Important DFA 6. L = {w | na(w) mod 2 = 0 and nb(w) mod 2 = 1}:
Some examples of important DFAs are given below: a

1. Containing even number of a’s. This can be written
as L = {w | na(w) mod 2 = 0}:
a
b
b b b b
a b
a
a
2. DFA having exactly two a’s:
a
b b b b
7. DFA which accepts exactly two a’s and one b:
a a a
a a
3. L = {w | na(w) mod 2 = 0 and nb(w) mod 2 = 0}:
a
a
b b a, b
b
a
a a, b
a
b b b b
b b
a 8. DFA which accepts at least two a’s and at least

one “b”:
a a
a a
4. L = {w | na(w) mod 2 = 0 or nb(w) mod 2 = 0}:
a
b b b
a
a
a
b b b b b a, b
b
5.4 PUSHDOWN AUTOMATA

a
a
Pushdown automata (PDA) are devices that recognize
5. L = {w | na(w) mod 2 = 0 or nb(w) mod 2 = 1}: context-free languages. PDA has finite memory in terms
of a stack. A pushdown automaton is a non-determinis-
a tic device. The deterministic version of automata accepts
only a subset of CFL known as deterministic context-
a free language (DCFL).
b b b 5.4.1 Model of PDA

b
PDA can be defined as a seven-tuple set {Q, Σ, Γ, d, q0,
Z0, F}, where
a
1. Q is a finite and non-empty set of states.
a 2. Σ is a finite and non-empty set of input symbol.
Chapter 5.indd 250 4/9/2015 9:58:00 AM

5.4 PUSHDOWN AUTOMATA 251
3. Γ is a finite non-empty set of stack alphabets. leads to some final state of PDA or cause PDA to empty
4. d is the transition function which maps Q × (Σ ∪ its stack. A non-deterministic automaton is equivalent
{l}) × Γ → Q × Γ* to a CFG and more powerful than a deterministic PDA.
5. q0 is the initial state (start state).
6. Z0 ∈ Γ is the starting (top most) stack symbol. Example 5.16
7. F is the set of final states and F ⊆ Q.
Figure 5.5 shows a model of PDA.
a/0,10 a/1,11
q0 q1 δ {(q0,a,0)={(q1,10),(q3,λ)}
$ Ψ I/P Tape δ {(q0,λ,0)={(q3,λ)}
δ {(q1,a,1)={(q1,11)}
λ/0,λ a/0,λ
b/1,λ δ {(q1,b,1)={(q2,λ)}
δ {(q2,b,1)={(q2,λ)}
Reading
z0
Top of
λ/0,λ
b/1,λ δ {(q2,λ,0)={(q3,λ)}
head the stack q4 q3
a
Finite state control This NPDA will recognize the string “aaabbb” by the
following moves:
Figure 5.5 | Model of PDA.

A A A b B B
5.4.2 Transition Function
Three types of operations are performed under the tran- Input head start from left and starting stack symbol is 0.
sition function (d ) of PDA. 1. (q0, a, 0) . (q1, 10)
1. Push element into stack: Let the current stack 2. (q1, a, 10) . (q1, 110)
symbol be z0, current state be q0 and current input 3. (q1, a, 110) . (q1, 1110)
symbol be “a”. So after reading “a”, z2 is to be 4. (q1, b, 1110) . (q2, 110)
pushed into stack and next state may or may not 5. (q2, b, 110) . (q2, 10)
be changed according to the transition function 6. (q2, b, 10) . (q2, 0)
written as 7. (q2, l, 0) . (q3, l)
d (current state, current input symbol, top of stack
As q3 ∈ F, the string is accepted.
symbol) → (next state, new top of stack)
d (q0, a, z0) = (q1, az0)
2. Pop element from stack: If “a” is input symbol 5.4.4 Deterministic PDA
on state q0 with stack top z1 and z1 is popped up
after processing “a” and next state is q1. Then A deterministic pushdown automata (DPDA) is one for
transition function d is written as which every input string has a unique way through the
machine. A PDA M = {Q, Σ, Γ, d, q0, Z0, F} is deter-
d (current state, current input symbol, top of stack
ministic if it satisfies the following two conditions:
symbol) → (next state, top of stack)
d (q0, a, z1) = (q1, l) 1. d(q, a, z) is either empty or only single move.
2. d(q, l, z) ≠ ∅ means d(q, a, z) = ∅ for each a ∈ Σ.
3. No change in state element: If on an input
symbol “a” at state q0 on stack top z1 nothing is The language accepted by DPDA is known as determin-
pushed or popped, then transition function can be istic context-free language (DCFL), and it is noticeable
written as that not all CFLs are DCFLs.
d (q0, a, z1) = (q1, z1)
Example 5.17
5.4.3 Non-deterministic PDA Language L = {an × bn: n ≥ 0}, it can be accepted by the
DPDA because all à’ characters are inserted into stack
Non-deterministic pushdown automata (NPDA) have till character `x’ come and then symbol à’ is removed as
more than one choice for a single input such as NFA. many times as many symbol `b’ occur. In the last stack
Transition function of NPDA is Q × (Σ ∪ {l}) × Γ → will be empty.
Q × Γ*. An NPDA accepts an input string if a sequence
Chapter 5.indd 251 4/9/2015 9:58:02 AM

5.4.5 Parsing 5.5 CONTEXT-FREE GRAMMAR AND

LANGUAGES
Parsing is a technique to construct a parse or a deriva-
tion tree. It is to check whether there is some leftmost
derivation for a given string using a certain grammar. A grammar G = (Vn, T, S, P) is said to be a CFG if the
Parse tree generated through parsing is also used by a production of G is of the form A → a, where a ∈ (Vn ∪ S)*
syntax analyzer in a compiler for checking a string is and A ∈ Vn.
according to a given grammar or not.
As we know that a CFG has no context either on left
or on right, this is the reason why it is also known as
5.4.5.1 Top-Down Parsing context-free.
In leftmost derivation, we start with initial and replace Problem 5.8: Consider a grammar G = (Vn, T, S, P)
the leftmost variable in each step and finally get the having production {S → aSa |bSb |x}. Check the pro-
string. This approach is known as top-down parsing. duction and find the language generated.
In the parse, start symbol is root, terminals are leaves
(input symbols) and other nodes are variables. We start Solution: Let P1: S → aSa
from the root and replacing the intermediate nodes one P2: S → bSb
by one from left to right reach the leaves. This approach P3: S → x
is also known as recursive descent.
(a, b, x) are terminals and S is a variable. As all the pro-
Here, we have constructed the leftmost derivation ductions are of the form A → a, where a ∈ (Vn ∪ S)*
looking ahead one symbol in the given string. A gram- and A ∈ Vn, hence G is a CFG, and it will produce
mar having this property is known as LL(1). In general, context-free language.
if a leftmost derivation is constructed by looking at the
k symbol ahead in the given string, then the grammar is Language generated: L(G) = {w x wR: w ∈ (a + b)*}
known as LL(k) grammar, where k ≥ 1.
Steps of construction of top-down parser:
5.5.1 Context-Free Grammar
Step 1: E
liminate left recursion (if any) from the given
grammar. A grammar G = (Vn, T, S, P) is said to be a CFG if
production P {u → v} follows these conditions:
Step 2: E
liminate left factoring (if any) from the given
grammar. 1. u → v, where v ∈ (V ∪ T)*
Step 3: If resulting grammar is LL(k) for some k ≥ 1, 2. u ≤ v (length of u is less than v)
then construct a parsing table. 3. Only single variable is allowed in the left side
(means u has single variable only)
5.4.5.2 Bottom-Up Parsing
Example 5.18
In bottom-up parsing, we start with the input string S → aSb/l
and replace the terminals by respective variables such
that these replacements lead to the starting symbol of A → aA/aB/Cb
the grammar. So every step takes input string close to B → a/Da
the starting symbol. This approach is the reverse of top-
down approach. It reads the input from left and uses
the rightmost derivation in reverse order. Bottom-up
parser is also known as shift-reduce (SR) parser. There 5.5.2 Standard Context-Free Language
are three actions in bottom-up parsing:
There are some standard languages under CFL such as:
Step 1: S
hift the current input (token) on the stack and 1. anbn, anb2n, a2nbn, anb2n + 3, a2n + 2b3n + 5, etc.
read the next token. 2. na(w) = nb(w) (means number of “a” equals to
Step 2: R
educe by some suitable LHS of production number of “b” in a language), ambn, where (m = n,
rule. m > n, m < n, etc.)
Step 3: A
ccept, final reduction which leads to starting 3. Palindrome such as [wwR (even palindrome), wawR
symbol. (odd palindrome), etc.]
Chapter 5.indd 252 4/9/2015 9:58:02 AM

5.5 CONTEXT-FREE GRAMMAR AND LANGUAGES 253
4. Generate language from grammar: 5.5.3.1 Leftmost Derivation Tree

• Grammar is S → aSbb|e
Language generated by the above grammar is A derivation tree is called as leftmost derivation (LMD)
anb2n. tree if the ordering of decomposed variable is from left
• Grammar is S → aSb|bSa|a|b to right. Thus, for generating a string w = aab from
grammar:
Language generated by the above grammar is
w x wR, (x ∈ a, b). S → AB (production 1)
5. Make grammar for given language: A → aaA (production 2)
•• Language L = {a2nbn}, where n ≥ 1 A → e (production 3)
Grammar for given language: S → aaSb|e
B → bB (production 4)
•• Language L = {w#wR: w ∈ (a, b), # ≠ Σ}
Grammar for a given language: S → aSb|bSa|# B → e (production 5)
LMD:
5.5.3 Derivation Tree or Parse Tree
S → AB by production 1)
The string generated by CFG G = (Vn, T, S, P) is S → aaAB (by production 2)
S → aaB (by production 3)
represented by a hierarchical structure called tree.
A derivation tree or parse tree for a CFG is a tree that
satisfies the following conditions: S → aabB (by production 4)
S → aab (by production 5)
1. If A → a1, a2, a3, …, an is a production in G, then
A becomes the father of nodes labelled as a1, a2,
a3, …, an. 5.5.3.2 Rightmost Derivation Tree
2. The root has label S (starting symbol).
3. Every vertex (or node) has a label. A derivation tree is called rightmost derivation (RMD)
4. Internal nodes should be labelled with variables tree if the ordering of decomposed variable is from right
only. to left. Thus, for generating a string w = aab from the
5. The leaf nodes are labelled with e or terminal above grammar:
symbols.
6. The collection of leaves from left to right yields the RMD:
string w.
S → AB (by production 1)
S → AbB (by production 4)
Problem 5.9: Consider the grammar S → S + S → Ab (by production 5)
S |S*S |a|b. Construct derivation (or parse) tree for the
string w = a + b*b. S → aaAb (by production 2)
Solution:
S S
A B
S + S
a a A b B
a S * S
∈ ∈
b b
Left to right
Chapter 5.indd 253 4/9/2015 9:58:03 AM

5.5.4 Ambiguous Grammar

Problem 5.10: Let grammar S → S + S |S *S |a |b |c
A grammar G is called ambiguous if for some string Solution: After removing left factoring, the produc-
w ∈ L(G), there exists two or more derivation tree (two tions are replaced by
S | → + SS ||*SS ||e
or more LMD or two or more rightmost derivation tree).
Let us consider a CFG grammar having production:
S → aS ||bS ||cS |
S → S + S |S*S |a |b, for string w = a + a*a have more
than one LMD tree.
5.5.5.2 Removal of Left Factoring
S S
In grammar G, two or more production of variable A are
said to have left factoring if the production are in the
S + S S * S following form:
A → ab1|ab 2|ab 3|…|abm

a S * S S + S a
where {b1|b 2|b 3|…|bm} ∈ (Vn ∪ S)* and does not start
a a a a with a. All these production have common left factor a.
Elimination of left factoring: Let variable A have (left
factoring) production as follows:
Note: A language (L) is called ambiguous language if
and only if every grammar which generates it is ambigu- A → ab1|ab2|ab3|…|abm|g 1|g 2|g 3|…|gm, where g1, g2,
ous. The only known ambiguous language is {anbmcn} ∪ g3, …, gm and b 1, b 2, b 3, …, bm do not contain a as a
{anbmcm}. prefix, then we replace this production by
Left recursion and left factoring are the major cause A → aA||g1|g2|g3|…|gm
for a grammar to be ambiguous. But presence of these in A → b1|b2|b3|…|bm
a grammar does not mean that grammar is ambiguous,
and similarly absence of these does not mean that gram-
mar is unambiguous. Problem 5.11: Let grammar A → abc |abd |abe remove
left factoring.
5.5.5 Removal of Ambiguity Solution: After removing left factoring, the produc-
tions are replaced by
Ambiguities are left recursion and left factoring. In this A → abA|
A → c |d |e
subsection, removal of various ambiguities is discussed
step by step.
5.5.5.1 Removal of Left Recursion 5.5.6 Context-Free Grammar Simplification
A production of grammar G = (Vn, T, S, P) is said to For simplification of context-free language, it is neces-

be left recursive grammar if it has one of the production sary to eliminate variable, terminal and production
in a given form. having the following properties:
1. Unit production (A → B)
A → Aa, where A is a variable and a ∈ (Vn ∪ S)*.
2. Null production (A → l)
Elimination of left recursion: Let the variable A have 3. Useless production, a variable which does not occur
left recursive problem as follows: even in a single derivation
A → Aa1|Aa2|Aa3|…|Aan|b1|b2|b3|…|bm, where b1, b2, 5.5.6.1 Removal of Unit Production

b3, …, bm do not begin with A, then we replace A pro-
duction by: A production (P) in CFG is said to be a unit production
| | | | if it is in the following form:
{A → b1A |b2A |b3A | …|bnA |},
A → B where A, B ∈ V (variable). There are three
where A → a1A| |a2A||a3A||…|anA||e
|
steps to remove a unit production.
Chapter 5.indd 254 4/9/2015 9:58:03 AM

5.5 CONTEXT-FREE GRAMMAR AND LANGUAGES 255
= P − (unit production)
Step 1: Find P Step 1: Identify the nullable variable.
Step 2: Find P production without null.
Step 2: F
ind unit production and draw a unit produc-
tion dependency graph.
Step 3: Add new rules to the grammar. Problem 5.13: Remove null (l) production from the
given CFG grammar.
Problem 5.12: Remove the unit production from the
S → ABaC
given CFG grammar.
A → BC
S → Aa |B
B → b | l
B → A|bb
C → D | l
A → a |bc |B
D→d
Solution:
) = P − (unit production
Step 1: Find a non-unit production (P
Solution:
)
) = P − (unit production)
(P Step 1: Identify nullable variables.
S → Aa Variables B and C directly produce null N = {B, C}.
B → bb But variable A is related to B and C, so A will also
produce a null production.
A → a |bc
Step 2: Find a unit production and draw a unit pro- N = {A, B, C}
duction dependency graph. production without null.
Step 2: Find P
S→B
S → ABaC
B→A
A → BC
A→B
B→b
S A B C→D
D→d
Generate grammar without null production (G |) by
Step 3: Add new rules to the grammar. Check that putting null value to nullable variables one by one in
from each variable through other variables production without null. For example, put A = l in
on which terminals we can reach. S → ABaC, we will get S → BaC. Do this process for
all nullable variables.
S → a|bc|bb
S → ABaC |BaC |AaC |ABa|aC |aB|Aa|a
A → bb
A → BC |B|C
B → a|bc
B→b
Final grammar after removing a unit pro-
duction is as follows: C→D
S → Aa|a|bc|bb D→d
A → a|bc|bb
B → bb|a|bc 5.5.6.3 Removal of Useless Production
Note: Removal of a unit production has made B and
Let G = (Vn, T, S, P) be a CFG. A variable A ∈ V is
the associated production useless.
said to be useless if there is not even a single derivation
which uses that production.
5.5.6.2 Removal of Null Production

Example 5.19
A → l is a null production and A is a nullable vari- S → A
able; l should be removed because it creates difficulties
A → aA|a
for the compilers. There are two steps to remove a null
production. B → bA
Chapter 5.indd 255 4/9/2015 9:58:04 AM

The variable B is useless and so is the production

B → bA. Although b can drive a terminal string, there is A → bAB or {A → KB, K → LA, L → b}
no way to reach on B from the starting variable S. This grammar can be written as in CNF G | as
Vn = {S, A, B, D, K, L, X, Y, Z}
Σ = {a, b, d}
Problem 5.14: Consider a given grammar and elimi- P | = { S → XD, X → YA, Y → a, A → ZB, Z → a,
nate useless production. A → KB, K → LA, L → b}
S → aS |A|C
A→a
5.5.8 Greibach Normal Form
B → aa
C → aCb A CFG is called in Greibach normal form (GNF) if all
productions have the form {A → ax or A → a}, where a
Solution: ∈ T and x ∈ V *. Means RHS of production are either a
Step 1: First find out those variables which lead to terminal or a terminal followed by variables.
terminal strings. Because A → a and B →
aa, variables A and B directly come under
this set. But due to the dependency on A, Problem 5.16: Convert the given grammar into GNF
variable S also generates terminal string S → AB
S → A → a. So, A → aA|bB|b
S → aS |A B→b
A→a
Solutions: The GNF of the given grammar is
B → aa
S → aAB|bBB|bB
Step 2: Now we have to find those variables which
cannot be reached from the starting variable. A → aA|bB|b
B is the only variable which cannot be B→b
reached from S. So B is a useless variable.
Grammar after eliminating useless produc-
tion and variables is as follows: 5.5.9 Identification of Language
S → aS |A
1. If language is finite then it will surely be a regular
A→a language. Example: Let L = {ambn, m + n = 10} as
m and n are finite, so it is a regular language. But
it does not mean that if a language is infinite then
5.5.7 Chomsky Normal Form it cannot be regular such as (a*b*) is an infinite
and regular language.
Chomsky normal form (CNF) has restriction on the 2. If there is a comparison between m and n then it
length of the right-hand side of a production, either two will be CFL. Example: Let L = {ambn (m ≥ n)}.
variables or single terminal is allowed in the right side of 3. Let language L = {ambn}, and if m or n is non-
a production such as S → AB, S → a. linear then it will not be accepted by PDA, so it
will fall in the category of CSL.
4. If there are more than one comparison then it will
be CSL. Example: Language L = {ambnco, m ≥ n
Problem 5.15: Convert the given grammar into CNF
and o ≥ n}.
form:
S → aAD 5. All modular machines are regular language.
A → aB |bAB 6. All palindrome languages are CFL language.
B→b
D→d 5.6 TURING MACHINE
Solution:
S → aAD or {S → XD, X → YA, Y → a} In 1936, an English mathematician Alan Turing sug-
A → aB or {A → ZB, Z → a } gested the concept of Turing machine (TM). This
machine can simulate the behaviour of a general purpose
Chapter 5.indd 256 4/9/2015 9:58:04 AM

5.6 TURING MACHINE 257
computer. A TM is a generalization of a PDA, which The transition function (d) is defined as d(p, a) = (q, b,
uses a tape instead of tape and stack. The length of the D), where p is the present state, q is the next state a, b ∈ Γ
tape is assumed to be infinite. A tape is divided into and D is the movement {(left (L) or right (R)}.
cells, and one cell can hold one input symbol. The head After reading an input a ∈ Σ, TM does one of the
of the TM is capable to read and write on the tape and following:
can move left or right or remain static.
1. Replaces a by b and moves right (R) as d(p, a) =
TMs accept the languages defined by type 1 gram- (q, b, R).
mar, and these languages are called recursively enumer- 2. Without write, anything moves right (R) as d(p, a) =
able or type 1 languages. (q, a, R).
3. Replace a by b and move left (L) as d (p, a) = (q, b, L).
5.6.1 Model of a Turing Machine 4. Without write, anything moves left (L) as d(p, a) =
(q, a, L).
A TM consists of a tape which is finite at the left end The change of state in the above transition is optional
and infinite at the right end and has a finite control and (means may or may not change).
read/write head (Fig. 5.6).
5.6.1.2 Language Reorganization by a
Turing Machine
A TM can be used as a language recognizer. It recognizes

all languages (regular, CFL, CSL, type 0).
R/W Head (L to R) Let w be a string to be written on tap. The Turing
machine is started from initial state q0, read write head
is placed on the left most symbol of w. with the sequence
Finite control of moves, if TM enters to a final state and halts then
string w is said to be accepted by Turing machine.
Figure 5.6 | Model of a turing machine. Problem 5.17: Design a TM for L = {w: w ∈ (a + b)*
ending in substring abb}
5.6.1.1 Mathematical Description of a Solution: Let TM M = {(q0, q1, q2, q3, h), (a, b), (a, b, #),
Turing Machine d, q0, #, h} accepts language L, where d is defined as
follows:
A TM can be described by a seven-tuple set (Q, Σ, Γ, d,
d(q0, b) = (q0, b, R)
q0, #, F), where
d(q0, a) = (q1, a, R)
1. Q is the finite set of states (not including the halt
state (h)). d(q1, a) = (q1, a, R)
2. Σ is the input alphabet which is a subset of the d(q1, b) = (q2, b, R)
tape alphabet (not including the blank symbol d(q2, a) = (q1, a, R)
d(q2, b) = (q3, b, R)
#).
3. Γ is the finite set of symbols called the tape
alphabet. d(q3, a) = (q1, a, R)
4. d is the transition function which maps from Q × d(q3, b) = (q0, b, R)
Γ → Q × Γ × [L, R].
d(q3, #) = (h, #, S)
5. # ∈ Γ is a special symbol called blank.
6. q0 ∈ Q is the initial state. The transition diagram is shown in the following figure.
7. F ⊆ Q is the set of final states.
A TM can be deterministic or non-deterministic depend- b,b,R b,b,R
ing on the number of moves in a transition. If a TM has
at the most one move in a transition, then it is a deter- a,a,R q b,b,R b,b,R a,a,R
ministic Turing machine (DTM). If there is more than q0 1 q2 q3 h
one move, then it is a non-deterministic Turing machine a,a,R
(NTM). A standard TM reads one cell of the tape and
changes its state (optional) and moves left or right by a,a,R
one cell.
Chapter 5.indd 257 4/9/2015 9:58:05 AM

5.6.2 Designing a Turing Machine 5.6.3 Recursive and Recursive Enumerable

Languages
Let us consider a problem for which we will design a TM.
Suppose we have to design a TM which will halt on a The properties of recursive and recursive enumerable
final state if a string has an even number of “a”s’. So, languages are as follows:
first we will design a DFA which accepts a string having
an even number of “a”s’, and after that we will design a 1. A language L is recursive enumerable (RE) if and
TM for the same problem. only if there exists a TM that accepts it, machine
must halt and accept ∀ w ∈ L.
b b 2. A language L is said to be recursive (REC) if and
a only if there exists a TM that accepts it and which
a halts on ∀ w ∈ Σ*.
This regular machine will accept only those strings which 3. REC is a decidable language.
have an even number of a’s. Now we will make a final state 4. RE is a semi-decidable language.
as non-final and add one final state on blank input. 5. Both RE and REC have enumerable procedure
(EP). A procedure is called an EP iff it lists all
b,b,R b,b,R members of L in a finite amount of time.
a,a,R 6. Only REC has membership algorithm.
a,a,R 7. A set S is countable if it has an EP.
8. If A and B are countable then (A × B) is also countable.
, ,R 9. Real numbers are uncountable infinite.
10. If a language L is RE then L has an EP but L does
not have an EP.
11. If a language L is REC then both L and L have an EP.
Every subset of countable set is countable, so every
Problem 5.18: Design a TM which accepts a lan- language L is countable because L ⊆ Σ* which is
guage L = {anbn, n ≥ 1}. countable. It means every language has an EP
but not those languages which are “not RE” or
Solution: First make a transition function move based
outside RE.
on an input in the tape which is shown in the figure.
12. Set of all TM is countable.
d(q0, a) = (q1, x, R) 13. If L is REC then L also will be REC.
d(q1, a) = (q1, a, R) 14. If L is RE then L may or may not be RE.
d(q1,
15. If L is RE then L may or may not be RE.
y) = (q1, y, R)
16. Languages (regular, CFL, CSL, REC, RE) are
d(q1, b) = (q2, y, L) countable infinite.
d(q2, y) = (q2, y, L) 17. If both L and L is Turing enumerable then they are
d(q2, a) = (q2, a, L) REC.
d(q2, x) = (q0, x, R)
d(q0, y) = (q3, y, R) 5.6.4 Variation of Turing Machine
d(q3, y) = (q3, y, R)
Different variants of Turing machines are available.
d(q3, #) = (qf, #, R) These variations are given as follows:
A TM for a given language L = {anbn, n ≥ 1}.
1. Multi-track Turing machine
a,a,R y,y,L 2. Multi-tape Turing machine
y,y,R a,a,L 3. Non-deterministic Turing machine
1 4. Universal Turing machine
a,x,R b,y,L 5. Offline Turing machine
q0 q1 q2
6. Turing machine with semi-infinite machine
x,x,R
y,y,R
5.7 CLOSURE AND DECIDABILITY
q3 qf
, ,R There are different closure and decidability properties
y,y,R for various languages. Tables 5.6 and 5.7 describe these
properties for different languages.
Chapter 5.indd 258 4/9/2015 9:58:07 AM

5.7 CLOSURE AND DECIDABILITY 259
Table 5.6 | Closure property of languages
Operation Languages
Regular DCFL CFL CSL REC RE
∪      
∩      
Lc      
-      
*      
L1 ∪ R      
L1 ∩ R      
R
L    — — —
h(L)    — — —
h−1(L)    — — —
L1/L2   — — — —
{Surely exists}
 {May or may not exist}
— {Not known}
Table 5.6 is interpreted as follows: second language. For example, if one language is
1. {∪} means that if there are two languages L1 and regular and the other is DCFL, then their union
L2 which are regular then their union will be reg- will be a DCFL.
ular, whereas if languages are DCFL then their 3. LR represents reverse of a language; if there are
union will not be DCFL. two CFLs L1 and L2 then their union will be a CFL
2. {L1 ∪ R} means that from two languages if one is and their intersection will not be a CFL.
regular (R) and the other is different, then their 4. (.) shows concatenation operation.
union will be result in a language similar to the 5. (*) denotes multiplication operation.
Table 5.7 | Decidability table for languages
Algorithm Exists Languages

Regular CFL CSL REC RE
Membership: w ∈ Σ*, L then w ∈ L?     

Finiteness: Given language L is finite or infinite?     
Emptiness: Language L is empty or not?     
Equivalence: Two languages L1 and L2 are equal     
or not?
Ambiguity: Language L is ambiguous or not?     
Regularity: Language is regular or not?     
Disjointness: Two language intersection is empty     
or not?
Everything: Language L, then L = Σ*     
 {Denotes that algorithm exists to decide about the question (decidable)}
 {Denotes that algorithm does not exist about that question (means undecidable)}
Chapter 5.indd 259 4/9/2015 9:58:08 AM

IMPORTANT FORMULAS
1. Number of states in a machine which accept m a’ s 12. If there is a comparison between m and n then it
and n b’ s: will be a CFL. Example: Let L = {ambn, (m ≥ n)}.
[(m + 1)*(n + 1) + 1] 13. Let language L = {ambn} and if m or n is non-
2. Mod n machines have n states. linear then it will not be accepted by PDA, so it
will fall in the category of CSL.
3. Mod machine have no trap states.
14. If there are more than one comparison then it will
4. If a machine accepts string length exactly n, then be a CSL. Example: Language L = {ambnco, m ≥ n
it has (n + 2) states. and o ≥ n }.
5. If a machine accepts string length ≤ n, then it has 15. All modular machines are regular language.
(n + 2) states.
16. All palindrome languages are CFL language.
6. If a machine accepts string length ≥ n, then it has
(n + 1) states. Variation of NFA/DFA
7. m States, n outputs Mealy machine ≡ Moore 17. DFA/NFA + (left ↔ right) move ≡ (2 − DFA)/
machine containing ≤ mn + 1 states. (2 − NFA) means whatever a 2 − DFA can do,
that can be done by simple DFA with left — right
8. m States, n outputs Moore machine ≡ Mealy move.
machine containing ≤ m states.
18. DFA/NFA + (left ↔ right) move + (read/write)
9. Output length of Mealy machine is equal to input head ≡ (2 − DFA)/(2 − NFA).
length.
19. DFA + 1 stack ≡ DPDA).
10. Output length of Moore machine is one greater 20. NFA + 1 stack ≡ NPDA).
than the input length because first output symbol
is additional without reading any symbol from the 21. DFA + 2 stack ≡ TM.
input. 22. DFA/NFA + 2 counter ≡ TM.
11. If language is finite then it will surely be regular. 23. DFA/NFA + 2 stack ≡ DFA/NFA + 2 counter.
Example: Let L = {ambn, m + n =10} as m and n
are finite so it is a regular language. But it does not 24. Power of {(DFA/NFA) < (DFA/NFA + 1 counter)
mean that if a language is infinite then it cannot < (DFA/NFA + 1 stack)}
be regular such as (a*b*) is an infinite and regular 25. {(DFA/NFA + 1 counter) < (DFA/NFA + 2
language. counter)}.
SOLVED EXAMPLES
1. Which of the following is a regular expression? In option (c), the given expression will generate
(a) L = {0 1 | m, n ≥ 0}
m n any string “x ” twice, which will be accepted by
(b) L = {0n1n | n ≥ 0}
PDA. So it is not a regular expression.
(c) L = {xx | x ∈ {0, 1}*} In option (d), the given expression will generate
(d) L = {1 0 | n ≥ 0}
n2 n twice the n number of 1’s followed by n number of
0’s, which cannot be accepted by FA. So, it is not
Solution: In option (a), the given expression a regular expression.
will generate any number of 0’s followed by any Ans. (a)
number of 1’s, which can be accepted by FA. So, it
2. The class of context-free language is not closed
is a regular expression.
under
In option (b), the given expression will gener-
ate number of 0’s followed by number of 1’s (a) union
(both should be same in number), which will (b) star
be accepted by PDA. So, it is not a regular (c) repeated concatenation
expression. (d) intersection
Chapter 5.indd 260 4/9/2015 9:58:08 AM

SOLVED EXAMPLES 261
Solution: Context-free language is not close under Let S denote the seven bits binary strings in which
intersection. Please refer table 5.6 (closure prop- the first, fourth and last bits are 1. The number of
erty of languages) in text. strings in S that are accepted by M is
Ans. (d)
(a) 6 (b) 5
3. Which of the following is true for regular sets A = (c) 4 (d) 7
(1101 + 1)* and ((1101)*1*)*?
(a) A ⊂ B (b) B ⊂ A Solution: Strings which will be accepted by
(c) A = B (d) A and B are incomparable machine M can be found by fixing 1 at places first,
fourth and last. We will have following four strings:
Solution: Both the regular expressions will gener- 1001001 1001011 1001111 1111001
ate the same set of strings, so both the regular sets Ans. (c)
are equal.
Ans. (c) 7. Which of the following CFG will generate the lan-
guage L = {ambncpdq | m + n = p + q}
4. The language generated by the following grammar
is S → aSa |bSb |e (a) S → aSd |A|B (b) S → aSd |A|B
A → aAc |C A → aAc |D
(a) a b n ≥ 0, m ≥ 0
m n
B → bBd |D B → bBd |C
(b) anbm n ≥ 1, m ≥ 1
C → bCc | e C → CBc | e
(c) S → aSd |A|B (d) S → aSd |A|B
(c) Odd length palindrome
A → aAB |C A → aAc |C
(d) Even length palindrome
Solution: In the given grammar, variable S will B → bBd |C B → bBd |C

generate a pair of symbol à’ or symbol `b’ with C → bCc | e C → bCc
variable S in the middle of them, and this is in
loop so it can be generated any number of times. Solution: Strings generated by L should have sum
In the last variable, S will be replaced by e. Some of number of à’ and `b’ equal to total number of `c’
of the strings which can be generated by the given and `d’. Only in option (c), the grammar can gener-
grammar are {aa, bb, abba, aabbaa, baab, etc}. ate such string which has number of (a + b) equal
So, the given grammar will generate even length to number of (c + d ).
palindrome. Ans. (c)
Ans. (d) 8. Consider the FA shown in the figure below, which
5. Number of states in DFA accepting the following language is accepted by the FA
language is L = {anb2m |n, m ≥ 1} b
a
(a) 2 (b) n Start a
(c) m (d) 5
b
Solution: DFA for given language L = anb2m |n,
m ≥ 1 is (a) b + (a + b)*b (b) (b + a + b)*a
a b (c) b + a*b (d) b + a*b*
q0 a q1 b q2 b q3 Solution: Given that FA will accept any string
b which has the last symbol à’. So, only option (b)
a
a expression (b + a + b)* a can generate all combina-
q4
tions of strings which end with symbol à’.
a,b q0 {Starting state} Ans. (b)
q3 {Final state} 9. Let L = L1 ∩ L2, where L1 and L2 are languages
Ans. (d) defined below
6. Consider the following DFS automaton M. L1 = {ambmcanbm | m, n ≥ 0} and L2 = {aibjck | i,

j, k ≥ 0}. Then L is
1 0 0, 1
Start (a) regular but not recursive
0 0 1 (b) context-free
(c) context-free but not regular
1 (d) recursively enumerable but not context-free
Chapter 5.indd 261 4/9/2015 9:58:09 AM

Solution: Language L1 is CFL and language L2 is (a) 1 (b) 3

regular. So, the result of L1 ∩ L2 will be context (c) 2 (d) 22
free.
Ans. (c) Solution: DFA will construct as:
10. Which of the following is/are correct? 1 1 0
q1 0 q2 0 q3
I. A language is context-free if and only if it is
accepted by the PDA. 1
II. PDA is a finite automata with push-down Ans. (b)
stack.
14. Which of the following CFG cannot be simulated
(a) Only I is true by an FSM?
(b) Only II is true
(c) Both I and II are true (a) S → Sa |b
(d) Both I and II are false (b) S → aSb |ab
(c) S → abX, X → cY, Y → d |aX
Solution: Both I and II are true. (d) None of these
Ans. (c)
Solution: Given Grammar S → aSb |ab generates
11. How many strings of length less than 4 contains language {anbn} means language contains equal
the language described by the regular expression number of a’s and b’s. So, it cannot be simulated
(x + y)*y(a + ab)*? by finite state machine.
(a) 3 (b) 9 Ans. (b)
(c) 10 (d) 11 15. Let r = 1(1 + 0)*, s = 11*0 and t = 1*0 be three
regular expressions. Which one of the following is
Solution: From the regular expression (x + y)* true?
y(a + ab)*, it can be seen that it contains 11 strings
as follows: (a) L(s) ⊆ L(r) and L(s) ⊆ L(t)
{xy, yy, y, ya, yab, xya, yya, yaa, xxy, yyy, xyy} (b) L(r) ⊆ L(s) and L(s) ⊆ L(t)
(c) L(s) ⊆ L(t) and L(s) ⊆ L(r)
Ans. (d)
(d) L(t) ⊆ L(s) and L(s) ⊆ L(r)
12. Consider the following laws:
Solution:
L1: (L*)* = L* r = 1(1 + 0)*, the language corresponds to r having
L2: ∅* = ∅ all strings starting with 1.
L3: e* = e s = 11*0, the language corresponds to s having all
L4: L+ = L* + e strings starting with 1 followed by any number of
L5: L* = e + L+ 1 and ending with 0.
Which of the above laws hold for regular expression? t = 1*0, the language corresponds to t having all
strings ending with 0.
(a) L1, L2 and L5 (b) L1, L3 and L5 So, L(s) ⊆ L(r) and L(s) ⊆ L(t). Therefore, option
(c) L2, L3 and L4 (d) L1, L4 and L5 (a) is correct.
Ans. (a)
Solution:
16. Which two of the following four regular expressions
L1: both the expression generates any string made are equivalent?
up of input symbol of L. it also include e.
L2: ∅* can generate {∅ and e}
(i) (00)*(e + 0) (ii) (00)*
(iii) 0* (iv) 0(00)*
L3: e * can only generate e.
(a) (i) and (ii) (b) (ii) and (iii)
+ contain all the string made up of input
L4: L (c) (i) and (iii) (d) (iii) and (iv)
symbol of L excluding e.
Solution:
L5: L
* contain all the string made up of input
symbol of L including e. (i) (00)*(e + 0) represents any number of 0’s
(ii) (00)* represents even no. of 0’s
Son only L1, L3 and L5 are applicable. (iii) 0* represents any number of 0’s
Ans. (b) (iv) 0(00)* represents odd number of 0’s
13. How many states does the DFA constructed for the So, (i) and (iii) are the same expressions.
set of all strings ending with “00” have? Ans. (c)
Chapter 5.indd 262 4/9/2015 9:58:10 AM

17. Let L be the set of all binary strings whose last two (c) All the three languages are context-free.
symbols are the same. The number of states in the (d) Turing machine can be used to recognize all the
minimum state deterministic finite state automa- three languages.
tion accepting L is
Solution: Language L1 is regular because it can
(a) 2 (b) 5 be accept by NFA. Language L2 and L3 are CFL
(c) 8 (d) 3 because they have comparison in between p and q.
Ans. (c)
Solution: Minimum number of states in DFA
which accept all binary strings whose last two sym- 19. Which of the following language over {a, b, c} is
bols are the same is 5. accepted by a deterministic pushdown automata?
(a) wcwR | w ∈ {a, b}*}
(b) wwR | w ∈ {a, b, c}*}
a
(c) {anbncn | n ≥ 0}
(d) {w | w is a palindrome over {a, b, c}
a
B C
a Solution: The language L = wcwR | w ∈ {a, b}*}
b is accepted by DPDA because it is the only one
b which is DCFL. For the above language everything
A a is pushed into the stack until “c” appears. When
a “c” appears all elements are removed from the
b stack if they match every symbol in wR.
D E Ans. (a)
b
20. Which one of the following is the strongest correct
b statement about a finite language over some finite
Ans. (b) alphabet ∑?
18. Consider the languages L1, L2, L3 as given below: (a) It could be undecidable.
(b) It is Turing machine recognizable.
L1 = {1p0q | p, q ∈ N } (c) It is a regular language.
L2 = {1p0q | p, q ∈ N and p = q} (d) It is a context-sensitive language.
L3 = {1p0q | p, q, r ∈ N and p = q = r} Solution: Every finite language is regular and
Which of the following statements is NOT TRUE? also CFL, CSL, RE (Turing machine recognizable)
because regular language ⊂ CFL ⊂ CSL ⊂ RE. So,
(a) Pushdown Automata (PDA) can be used to the strongest correct statement is option (c).
recognize L1 and L2.
(b) L1 is a regular language. Ans. (c)
1. Ram and Shyam have been asked to show that Solution: Reduction from the 3-SAT problem
a certain problem Π is NP-complete. Ram to Π is NP hard and reduction from Π to 3-SAT
shows a polynomial time reduction from the shows that Π is in NP.
3-SAT problem to Π, and Shyam shows a poly-
As Π is in NP and it is NP hard, so it can be
nomial time reduction from Π to 3-SAT. Which
inferred that Π is NP complete.
of the following can be inferred from these
Ans. (c)
reductions?
(a) Π is NP-hard but not NP-complete. 2. If strings of a language L can be effectively enumer-
(b) Π is in NP, but is not NP-complete. ated in lexicographic (i.e., alphabetic) order, which
(c) Π is NP-complete. of the following statements is true?
(d) Π is neither NP-hard nor in NP. (a) L is necessarily finite.
(GATE 2003: 1 Mark) (b) L is regular but not necessarily finite.
Chapter 5.indd 263 4/9/2015 9:58:10 AM

(c) L is context-free but not necessarily regular. Solution: Option (a) is False. G is ambiguous due
(d) L is recursive but not necessarily context-free. to the possibility of two parsing trees for the same
string.
(GATE 2003: 1 Mark)
Option (b) is True. Let x = ab and y = ba, so abba
Solution: L is recursive because recursive lan- does not belong to L(G).
guages can be effectively enumerated in lexico- Ans. (b)
graphic order. 6. A single tape Turing machine M has two states
Ans. (d) q0 and q1, of which q0 is the starting state. The
3. Consider the set ∑* of all strings over the alphabet tape alphabet of M is {0, 1, B} and its input
∑ = {0, 1}. ∑* with the concatenation operator for alphabet is {0, 1}. The symbol B is the blank
strings symbol used to indicate end of an input string.
(a) does not form a group The transition function of M is described in the
(b) forms a non-commutative group following table
(c) does not have a right identity element 0 1 B
(d) forms a group if the empty string is removed
from ∑* q0 q1, 1, R q1, 1, R Halt
(GATE 2003: 1 Mark) q1 q1, 1, R q0, 1, L q0, B, L
Solution: Fact
The table is interpreted as illustrated below.
Ans. (a)
The entry (q1, 1, R) in row q0 and column 1 signi-
4. Consider the following deterministic finite state
fies that if M is in state q0 and reads 1 on the cur-
automaton M.
rent tape square, then it writes 1 on the same tape
1 0 0, 1 square, moves its tape head one position to the
right and transitions to state q1.
0 0
1 Which of the following statements is true about M?
(a) M does not halt on any string in (0 + 1)+.
1 (b) M does not halt on any string in (00 + 1)*.
(c) M halts on all string ending in a 0.
Let S denote the set of seven bit binary strings in
(d) M halts on all string ending in a 1.
which the first, the fourth, and the last bits are 1.
The number of strings in S that are accepted by (GATE 2003: 2 Marks)
M is
(a) 1 (b) 5 Solution: M does not halt on any string in (0+1)+.
(c) 7 (d) 8 Ans. (a)
(GATE 2003, 2 Marks) 7. Define languages L0 and L1 as follows:

L0 = (<M, w, 0>|M halts on w)
L1 = {<M, w, 1>|M does not halt on w}
Solution: Strings that are accepted are: 1001001,
1001011, 1001101, 1001111, 1111001, 1101001, and
1011001. Here <M, w, i> is a triplet, whose first component,
Ans. (c) M, is an encoding of a Turing machine, second com-
ponent, w, is a string and third component, i, is a bit.
Let L = L0 ∪ L1. Which of the following is true?
5. Let G = ({S}, (a, b}R, S) be a context-free gram-
mar where the rule set R is
S → aS b | S S | e
(a) L is recursively enumerable, but L is not.
(b) L is recursively enumerable, but L is not.
Which of the following statements is true? (c) Both L and L are recursive.
(d) Neither L nor L is recursively enumerable.
(a) G is not ambiguous.
(b) There exists x, y, ∈ L(G) such that xy ∉ L(G). (GATE 2003: 2 Marks)
(c) There is a deterministic pushdown automaton
Solution: L contains strings that halts on w and
that accepts L(G).
does not halt on w. L can be recursively enumer-
(d) We can find a deterministic finite state autom-
able if it contains valid strings. So, L is not recur-
aton that accepts L(G).
sively enumerable, but L is recursively enumerable.
Chapter 5.indd 264 4/9/2015 9:58:12 AM

8. Consider the NFA M shown below. (a) divisible by 3 and 2

(b) odd and even
0, 1
0 (c) even and odd
(d) eivisible by 2 and 3
0, 1 (GATE 2004: 2 Marks)
1
0 Solution: The machine accepts strings that have
1’s divisible by 3 and 0’s divisible by 2. For example,
strings 11100, 10011, 00111 will be accepted by this
machine and strings 1001, 10111 will be rejected.
Let the language accepted by M be L. Let L1 be Ans. (a)
the language accepted by the NFA M1 obtained by 11. The language [ambncm+n | m, n ≥ l) is
changing the accepting state of M to a non-accept-
ing state and by changing the non-accepting states (a) regular
of M to accepting states. Which of the following (b) context-free but not regular
statements is true? (c) context-sensitive but not context-free
(d) type 0 but not context-sensitive
(a) L1= {0,1}* − L (b) L1 = {0,1}*
(c) L1 ⊆ L (d) L1 = L (GATE 2004: 2 Marks)

Solution: The given language can be accepted
Solution: by PDA, but cannot be accepted by finite autom-
0 ata. So, it is only context-free not regular. Finite
automata cannot accept context free language, it
0, 1 only accepts regular language.
Ans. (b)
12. Consider the following grammar G:

0, 1
S → bS | aA | b
Strings accepted by resulting NFA:
{∈,0,1,00,01,10,11….} = (0+1)* A → bA | aB
Ans. (b) B → bB | aS | a
9. The problems 3-SAT and 2-SAT are
Let Na(w) and Nb(w) denote the number of a’s
(a) both in P. and b’s in a string w, respectively. The language
(b) both NP-complete L(G) ⊆ {a, b} * generated by G is
(c) NP-complete and in P, respectively
(a) {w |Na(w) > 3Nb(w)}
(d) undecidable and NP-complete, respectively
(b) {w |Nb(w) > 3Nb(w)}
(GATE 2004: 1 Mark) (c) {w |Na(w) = 3k, k ∈ {0, 1, 2, ...}}
(d) {w |Nb(w) = 3k, k ∈ {0, 1, 2, ...}}
Solution: 3-SAT are NP-complete problems.
2-SAT are P problems.
Ans. (c) Solution: Strings formed from given grammar are:
10. The following finite state machine accepts all those b, bb, babaab, abbb, and so on. This means the
binary strings in which the number of 1’s and 0’s number of a’s are multiples of 3.
are, respectively, According to given grammar, string accepted
is baaa. Options (a), (b) and (d) are false. Only
1 option (c) is correct.
Ans. (c)
1 1
13. L1 is a recursively enumerable language over Σ. An
0 0 algorithm A effectively enumerates its words as w1,
0 1
0 0 w2, w3, …. Define another language L2 over Σ ∪
0
{#} as {wi # wj: wj ∈ L1, i < j}. Here # is a new
symbol. Consider the following assertions:
1 1
Chapter 5.indd 265 4/9/2015 9:58:13 AM

S1: L1 is recursive implies L2 is recursive The output computes 2’s complement of input
S2: L2 is recursive implies L1 is recursive number.
Ans. (b)
Which of the following statements is true?
16. Consider the machine M:
(a) Both S1 and S2 are true.
(b) S1 is true but S2 is not necessarily true. b a
(c) S2 is true but S1 is not necessarily true.
(d) Neither is necessarily true. a b bc b
a a
Solution: For L1, the membership algorithm can
be constructed as follows. For an arbitrary wi # wj, a, b
(i) If wi, wj ∈ L1 and i < j then wi # wj

(ii) If wi ∉ L1 or wj ∉ L2 or i ≥ j, then wi # wj ∉ L2. The language recognized by M is
(a) {w ∈ {a, b}* every a in w is followed by exactly
This implies that L2 is also recursive, so S1 is true. two b’s}.
A similar membership algorithm is not possible for (b) {w ∈ {a, b)* every a in w is followed by at least
L2, so S2 is not necessarily true. two b’s}.
Ans. (b) (c) {w ∈ (a, b}* w contains the substring “abb”}.
14. Consider three decision problems P1, P2 and P3. It (d) {w ∈ {a, b}* w does not contain “aa” as a
is known that P1 is decidable and P2 is undecid- substring}.
able. Which one of the following is TRUE? (GATE 2005: 2 Marks)
(a) P3 is decidable if P1 is reducible to P3.
Solution: DFA contains every “a” followed by at
(b) P3 is undecidable if P3 is reducible to P2.
least two b’s due to self-loops.
(c) P3 is undecidable if P2 is reducible to P3.
Ans. (b)
(d) P3 is decidable if P3 is reducible to P2’s
complement. 17. Let Nf and Np denote the classes of languages
(GATE 2005, 2 Marks) accepted by non-deterministic finite automata and
non-deterministic push-down automata, respec-
Solution: Any problem P is undecidable if any
tively. Let Df and Dp denote the classes of languages
known undecidable problem can be reduced to P.
accepted by deterministic finite automata and
So, option (c) is correct.
deterministic push-down automata, respectively.
Ans. (c)
Which one of the following is TRUE?
15. The following diagram represents a finite state
machine which takes as input a binary number (a) Df ⊂ Nf and Dp ⊂ Np
from the least significant bit. (b) Df ⊂ Nf and Dp = Np
(c) Df = Nf and Dp = Np
0/1
0/0 (d) Df = Nf and Dp ⊂ Np
1/0 (GATE 2005: 2 Marks)
1/1
Q0 Q1 Solution: Power of NFA and DFA is same. But

non-deterministic PDA is more powerful than
Which one of the following is TRUE? deterministic PDA.
(a) It computes 1’s complement of the input number Ans. (d)
(b) It computes 2’s complement of the input number 18. Consider the languages
(c) It increments the input number
(d) It decrements the input number L1 = {anbncm | n, m > 0} and L2 = {anbmcm | n,
m > 0}
Which one of the following statements is FALSE?
Solution: Output function is: (a) L1 ∩ L2 is a context-free language.
PS 0 1 (b) L1 ∪ L2 is a context-free language.
(c) L1 and L2 are context-free languages.
0 0 1 (d) L1 ∩ L2 is a context-sensitive language.
1 1 0
Chapter 5.indd 266 4/9/2015 9:58:13 AM

Solution: L1 and L2 both are context-free 22. If s is a string over (0 + 1)*, then let n0(s) denote
languages. Their intersection may or may not be the number of 0’s in s and n1(s) the number of 1’s
context-free. So, option (a) is false. in s. Which one of the following languages is not
Ans. (a) regular?
19. Let L1 be a recursive language, and let L2 be a (a) L = [s ∈ (0 + 1)* | n0(s) is a 3-digit prime}
recursively enumerable but not a recursive lan- (b) L = {s ∈ (0 + 1)* | for every prefix s’ of s, |n0(s’)
guage. Which one of the following is TRUE? — n1(s’)| ≤ 2}
(c) L = {s ∈ (0 + 1)*| |n0(s) - n1(s) ≤ 4}
(a) L1 is recursive and L2 is recursively enumerable. (d) L = {s ∈ (0 + 1)*|n0(s) mod 7 = n1(s) mod 5 = 0}
(b) L1 is recursive and L2 is not recursively
enumerable.
(c) L1 and L2 are recursively enumerable. Solution: Option (a) is finite, can be accepted by
(d) L1 is recursively enumerable and L2 is recursive. DFA.
Option (b), only prefix comparison is possible by
(GATE 2005: 2 Marks) DFA, so it is regular.
Option (c), whole string is compared, which is not
Solution: Recursive languages are closed under possible by DFA because it does not have memory.
complementation but recursive enumerable lan- Option (d), mod machines can be made using DFA.
guages are not. Ans. (c)
Ans. (b)
Linked Answer Questions 23 and 24:
20. Consider the languages 23. Which one of the following grammars generates the
L1 = {wwR | w ∈ {0, 1}*} language L = {ai bi | i ≠ j)?
L2 ={w # wR | w ∈ {0, 1}*}, (a) S → AC |CB (b) S → aS |Sb |a |b

C → aCb |a |b
where # is a special symbol A → aA | ∈
L3 = {ww | w ∈ {0, 1}*} B → Bb | ∈
Which one of the following is TRUE? (c) S → AC |CB (d) S → AC |CB
C →aCb | ∈ C →aCb | ∈
A → aA | ∈ A → aA | a
(a) L1 is a deterministic CFL.

B → Bb | ∈ B → Bb | b
(b) L2 is a deterministic CFL.
(c) L3 is a CFL, but not a deterministic CFL.
(d) L3 is a deterministic CFL. (GATE 2006: 2 Marks)
(GATE 2005: 2 Marks) Solution: Both options (a) and (b) can produce equal
number of a’s and b’s, which should not be produced
Solution: L2 can be accepted by DPDA, so it is according to grammar. Option (d) cannot produce all
a deterministic context free language. L1 is non- the strings. So, option (c) is the correct answer.
deterministic context free and L3 in not context Ans. (c)
free language. 24. In the correct grammar above, what is the length
Ans. (b) of the derivation (number of steps starring from S)
21. Let L1 = {0n+m1n0m |n, m ≥ 0}, to generate the string al bm with l ≠ m?
L2 = {0n+m1n+m0m |n, m ≥ 0} and (a) max(l, m) + 2 (b) l + m + 2
L3 = {0n+m1n+m0n+m |n, m ≥ 0} (c) l + m + 3 (d) max(l, m) + 3
Which of these languages are NOT context-free? (GATE 2006: 2 Marks)
(a) L1 only (b) L3 only Solution: This can be checked by producing vari-
(c) L1 and L2 (d) L2 and L3 ous strings of different length.
(GATE 2006: 1 Mark) S → AC

S → aC
Solution: Only L1 is accepted by pushdown S → aaCb
automata as only L1 can be written as a context- S → aab
free language. L2 is not context free, while L3 is Here l = 2 and m = 1, which satisfies the equation
context sensitive. max (l, m) + 2 = 4.
Ans. (d) Ans. (a)
Chapter 5.indd 267 4/9/2015 9:58:15 AM

25. For s ∈ (0 + 1)*, let d(s) denote the decimal value 28. Consider the regular language L = (111 + 11111)*.
of s (e.g. d(101) = 5). The minimum number of states in any DFA accept-
Let L = {s ∈ (0 + 1)*l d(s) mod 5 = 2 and d(s)
ing this languages is
mod 7 ≠ 4} (a) 3 (b) 5
Which one of the following statements is true? (c) 8 (d) 9
(a) L is recursively enumerable, but not recursive. (GATE 2006: 2 Marks)
(b) L is recursive, but not context-free. Solution: Nine states are required. The length of
(c) L is context-free, but not regular. string will be 8. Total states required are 8 + 1 =9.
(d) L is regular. Ans. (d)
29. Which of the following problems is undecidable?
Solution: DFA is possible for L. If for any
(a) Membership problem for CFGs
language, we can make DFA then that language
(b) Ambiguity problem for CFGs
will be regular language. So, L is regular.
(c) Finiteness problem for FSAs
Ans. (d)
(d) Equivalence problem for FSAs
26. Consider the following statements about the con-
(GATE 2007: 1 Mark)
text-free grammar:
G = {S → SS, S → ab, S → ba, S → e } Solution: No algorithm exists to check the ambi-
guity of grammar.
I. G is ambiguous. Ans. (b)
II. G produces all strings with equal number of a’s
and b’s. 30. Which of the following is TRUE?
III. G can be accepted by a deterministic PDA. (a) Every subset of a regular set is regular.
Which combination below expresses all the true (b) Every finite subset of a non-regular set is regular.
statements about G? (c) The union of two non-regular sets is not regular.
(d) Infinite union of finite sets is regular.
(a) I only (b) I and III only
(c) II and III only (d) I, II, and III (GATE 2007: 1 Mark)
(GATE 2006: 2 Marks) Solution: In option (a), anbn is subset of a* b*;
Solution: but it is not regular.
I. G is ambiguous. Null can be produced using Option (b) is correct.
multiple rules [S → e and S → SS]. Two parse Option (c) is union of two context-free languages,
trees will be formed. and thus is regular.
II. It cannot produce all the combinations of equal Option (d) is infinite union of finite sets, and thus
a’s and b’s such as string aabb is not possible. is not regular.
III. Language accepted is (ab + ba)*, which is Ans. (b)
regular. Regular languages are also accepted 31. A minimum state deterministic finite automaton
by DPDA. accepting the language
L = {w | w ∈ {0, 1}*, number of 0’s and 1’s in w are
Ans. (b)
27. Let L1 be a regular language, L2 be a deterministic divisible by 3 and 5, respectively} has
context-free language and L3 a recursively enumer-
able, but not recursive, language. Which one of the (a) 15 states (b) 11 states
following statements is false? (c) 10 states (d) 9 states
(a) L1 ∩ L2 is a deterministic CFL.
(b) L3 ∩ L1 is recursive. Solution: Divisible by 3 and 5 = 3 × 5 = 15 states
(c) L1 ∪ L2 is context-free. are required.
(d) L1 ∩ L2 ∩ L3 is recursively enumerable. Ans. (a)
(GATE 2006: 2 Marks) 32. The language L = {0i 21i i ≥ 0} over the alphabet
Solution: {0, 1, 2} is
Option (a) is true. DCFL is closed under regular (a) not recursive
intersection. (b) recursive and is a deterministic CFL
Option (b) is false. Intersection is recursively (c) a regular language
enumerable. (d) not a deterministic CFL but a CFL
Options (c) and (d) are true.
Ans. (b) (GATE 2007: 2 Marks)
Chapter 5.indd 268 4/9/2015 9:58:15 AM

Solution: L is accepted by DPDA. So, it is deter- (c) It is context-free but not regular.
ministic CFL. (d) It is neither regular nor context-free, but
Ans. (b) accepted by a Turing machine.
33. Which of the following languages is regular? (GATE 2008: 1 Mark)
(a) {wwR ⁄ w ∈ {0, 1}+} Solution: The options are checked by using
(b) {wwRx ⁄ x, w ∈ {0, 1}+} pumping lemma for regular and context-free lan-
(c) {wxwR ⁄ x, w ∈ {0, 1}+} guages. It is found that the language is not regular,
(d) {xwwR ⁄ x, w ∈ {0, 1}+} and hence not context free; but is accepted by the
(GATE 2007: 2 Marks) Turing machine.
Ans. (d)
Solution: Finite automata do not have memory
37. Which of the following are decidable?
element. So, they cannot remember the symbols.
Hence, options (a) and (d) are incorrect. I. Whether the intersection of two regular
Ans. (c) languages is infinite.
Common Data Questions 34 and 35: Consider the II. Whether a given context-free language is
following finite state automaton regular.
III. Whether two pushdown automata accept the
same language.
q3 IV. Whether a given grammar is context-free.
(a) I and II (b) I and IV
b b a b (c) II and III (d) II and IV
(GATE 2008: 1 Mark)
q0 q1 a q2 b
a Solution: Options II and III are undecidable.
a The intersection of two regular languages is deter-
34. The language accepted by this automaton is given mined by an algorithm, and it can be determined
by the regular expression whether the given grammar is context-free or not.
So, options I and II are decidable.
(a) b*ab*ab*ab* (b) (a + b)* Ans. (b)
(c) b*a (a + b)* (d) b*ab*ab*
38. If L and L are recursively enumerable, then L is
(a) Regular (b) Context-free
Solution: One a is compulsory. Minimum string is (c) Context-sensitive (d) Recursive
à’ that is accepted by the automaton. b*a (a + b)*
is regular and accepted by the automaton. (GATE 2008: 1 Mark)
Ans. (c)
Solution: A theorem states that when L and L
35. The minimum state automaton equivalent to the both are RE, then L is recursive.
above FSA has the following number of states Ans. (d)
(a) 1 (b) 2 39. Which of the following statements is false?
(c) 3 (d) 4
(a) E very NFA can be converted to an equivalent
(GATE 2007: 2 Marks) DFA.
Solution: Minimized DFA is: (b) Every non-deterministic Turing machine can
be converted to an equivalent deterministic
b Turing machine.
a, b
(c) Every regular language is also a context-free
q0 q1 language.
a
(d) Every subset of a recursively enumerable set is
Ans. (b) recursive.
36. Which of the following is true for the language
{aP | P is a prime}?
Solution: Option (d) is not necessarily true.
(a) It is not accepted by a Turing Machine. Every recursive set is recursively enumerable, but
(b) It is regular but not context-free. the reverse is not necessarily true.
Chapter 5.indd 269 4/9/2015 9:58:16 AM

40. Given below are two finite state automata (→ indi- 42. Match the following.
cates the start state and F indicates the final state) List I List II
a b a b (E) Checking that (P) L = {anbmcndm | n ≥

identifiers are declared 1, m ≥ 1}
Y: →1 1 2 Z: →1 2 2 before their use
2(F) 2 1 2(f) 1 1
(F) Number of formal (Q) X → XbX |XcX |dXf |g
parameters in the
declaration of a function
Which of the following represents the product
automaton Z × Y?
agrees with the number
of actual parameters in
(a) (b) use of that function
a b a b
(G) Arithmetic (R) L = {wcw | w ∈ (a|b)*}
→P S R →P Q S expressions with matched
pairs of parantheses
Q R S Q R S
(H) Palindromes (S) X → bXb|cXc|e
R(F) Q P R(F) Q P
S Q P S Q P Codes:
(a) E — P, F — R, G — Q, H — S
(c) (d) (b) E — R, F — P, G — S, H — Q
a b a b (c) E — R, F — P, G — Q, H — S
→P S Q →P S Q (d) E — P, F — R, G — S, H — Q
Q R S Q S R
Solution:
R(F) Q P R(F) Q P E — R: In R, first w is checking the declaration of
S P Q A Q P an identifier.
F — P: Actual parameters anbm and formal param-
eters cndm
(GATE 2008: 2 Marks) G — Q: Arithmetic expressions with matched pair
Solution: Option (a) represents the product of parenthesis
correctly. H — S: S is generating palindrome strings.
Ans. (a) Ans. (c)
43. Match the following NFAs with the regular expres-
41. Which of the following statements are true? sions they correspond to
I. Every left-recursive grammar can be converted
to a right-recursive grammar and vice versa. (P) (Q)
1 1 0
II. All e productions can be removed from any con- 0 1
text-free grammar by suitable transformations. 1 0
0 0 0 0
III. The language generated by a context-free
grammar all of whose productions are of the
form X → w or X → wY (where w is a string
of terminals and Y is a non-terminal) is always 1
(R) (S) 1
regular.
0 0
IV. The derivation trees of strings generated by 1 1
a context-free grammar in Chomsky normal 0 1
form are always binary trees. 0 1
(a) I, II, III and IV (b) II, III and IV only

(c) I, III and IV only (d) I, II and IV only
1. e + 0(01*1 + 00)*01*
Solution: Option II: If e belongs to L(G), then the 2. e + 0(10*1 + 00)*0
statement becomes false. 3. e + 0(10*l + 10)*l
Ans. (c) 4. e + 0(10*1 + 10)*10*
Chapter 5.indd 270 4/9/2015 9:58:17 AM

Codes: (a) T
he set of all strings containing the substring 00.
(a) P — 2, Q — 1, R — 3, S — 4 (b) The set of all strings containing at most two 0’s.
(b) P — 1, Q — 3, R — 2, S — 4 (c) The set of all strings containing at least two 0’s.
(c) P — 1, Q — 2, R — 3, S — 4 (d) The set of all strings that begin and end with
(d) P — 3, Q — 2, R — 1, S — 4 either 0 or 1.
(GATE 2009, 1 Mark)
Solution: P → e + 0(01*1 + 00)*01*
Solution: Given regular expression must have
Q → e + 0(10*1 + 00)*0
atleast two zeros 0(0 + 1)*0….
R → e + 0(10*1 + 10)*1
Ans. (c)
S → e + 0(10*1 + 10)*10* 47. Which one of the following is FALSE?
Ans. (c) (a) T here is a unique minimal DFA for every regu-
44. Which of the following are regular sets? lar language.
(b) Every NFA can be converted to an equivalent
I. {anb2m |n ≥ 0, m ≥ 0} PDA.
II. {anbm | n = 2m} (c) Complement of every context-free language is
III. {anbm| n ≠ m} recursive.
IV. {xcy | x, y ∈ (a, b}*} (d) Every non-deterministic PDA can be converted
to an equivalent deterministic PDA.
(a) I and IV only (b) I and III only
(c) I only (d) IV only (GATE 2009: 1 Mark)
(GATE 2008: 2 Marks) Solution: Both non-deterministic and determinis-

tic PDA have different powers. Non-deterministic
Solution: In options II and III, m and n are depen- PDA can accept ambiguous grammars also.
dent. As DFA do not have memory, cannot be Ans. (d)
accepted by DFA. Options I and IV are regular sets.
48. Match all items in Group 1 with correct options
Ans. (a)
from those given in Group 2.
45. S → aSa|bSb|a|b Group 1 Group 2
The language generated by the above grammar P. Regular expression 1. Syntax analysis
over the alphabet {a, b} is the set of
Q. Pushdown automata 2. Code generation
(a) all palindromes. R. Dataflow analysis 3. Lexical analysis
(b) all odd length palindromes.
S. Register allocation 4. Code optimization
(c) strings that begin and end with the same symbol.
(d) all even length palindromes. Codes:
(GATE 2009: 1 Mark) (a) P — 4, Q — 1, R — 2, S — 3
(b) P — 3, Q — 1, R — 4, S — 2
Solution: Strings generated by this grammar are (c) P — 3, Q — 4, R — 1; S — 2
odd length palindromes, for example, aaa, aba, (d) P — 2, Q — 1, R — 4, S — 3
ababa. (GATE 2009: 1 Mark)
Ans. (b)
Solution: Regular expressions: Lexical analysis
46. Which one of the following languages over the Pushdown automata: Syntax analysis
alphabet {0, 1} is described by the regular expres- Dataflow analysis: Code optimization
sion (0 + 1)*0(0 + 1)*0(0 + 1)*? Register allocation: Code generation
Ans. (b)
49. Given the following state table of an FSM with two states A and B, one input and one output;
Present Present Input Next Next Output

State A State B State C State A State B
0 0 0 0 0 1
0 1 0 1 0 0
1 0 0 0 1 0
(Continued)
Chapter 5.indd 271 4/9/2015 9:58:17 AM

Present Present Input Next Next Output

State A State B State C State A State B
1 1 0 1 0 0
0 0 1 0 1 0
0 1 1 0 0 1
1 0 1 0 1 1
1 1 1 0 – 1
If the initial state is A = 0, B = 0, what is the minimum length of an input string which will take the machine to
the state A = 0, B = 1 with output = 1?
(a) 3 (b) 4 (c) 5 (d) 6
Solution:
Present State Output
(0, 0) → 1
(0, 1) → 0
(1, 0) → 0
(0, 1) → 1
With A = 0, B = 1 and output = 1 will require a string of length 3.

Ans. (a)
50. Let L = L1 ∩ L2, where L1 and L2 are languages as Solution: The given DFA accepts all the strings
defined below: ending with 00.
Ans. (c)
L1 = {ambmcanbn | m, n3 ≥ 0}
52. Let L1 be a recursive language. Let L2 and L3 be
L2 = {aibick | i, j, k ≥ 0} languages that are recursively enumerable but not
Then L is recursive. Which of the following statements is not
necessarily true?
(a) not recursive.
(b) regular. (a) L2 − L1 is recursively enumerable.
(c) context-free but not regular. (b) L1 − L3 is recursively enumerable.
(d) recursively enumerable but not context-free. (c) L2 ∩ L3 is recursively enumerable.
(d) L2 ∪ L3 is recursively enumerable.
(GATE 2010: 1 Mark)
Solution: Intersection of two regular languages is
Solution:
regular, but L1 is not regular.
Subtraction of recursive language and recursive
Ans. (c)
enumerable is not necessarily recursive enumerable.
51. In the following figure, DFA accepts the set of all Ans. (b)
strings over {0, 1} that
53. Let L = {w ∈ (0 + 1)*|w has even number of 1’s},
1 1 0 i.e., L is the set of all bit strings with even number
of 1’s. Which one of the regular expressions below
0 0 represents L?
(a) (0*10*1)* (b) 0*(10*10*)*
1 (c) 0*(10*1)*0* (d) 0*1(10*1)*10*
(a) begin either with 0 or 1 (GATE 2010: 2 Marks)

(b) end with 0 Solution: Option (c) produces odd 1’s also, option
(c) end with 00 (a) produces strings ending with 1 and option
(d) contain the substring 00 (d) does not produce zero 1’s.
(GATE 2009, 2 Marks) Ans. (b)
Chapter 5.indd 272 4/9/2015 9:58:18 AM

54. Consider the languages 58. Which of the following pairs have DIFFERENT
expressive power?
L1 = {0i1j | i ≠ j}, L2 = {0i1j | i = j}, L3 = {0i1j | i = 2 j + 1}, L4 = {0i1j | i ≠ 2 j}
(a) Deterministic finite automata (DFA) and non-
}, L2 = {0i1j | i = j}, L3 = {0i1j | i = 2 j + 1}, L4 = {0i1j | i ≠ 2 j} deterministic finite automata (NFA)
Which one of the following statements is true? (b) Deterministic pushdown automata (DPDA) and
non-deterministic pushdown automata (NPDA)
(a) Only L2 is context-free (c) Deterministic single-tape Turing machine and
(b) Only L2 and L3 are context-free Non-deterministic single-tape Turing machine
(c) Only L1 and L2 are context-free (d) Single-tape Turing machine and multi-tape
(d) All are context-free Turing machine
Solution: All are context-free languages as they

Solution: NDPDA is more powerful than DPDA.
are recognized by pushdown automata.
Ans. (b)
Ans. (d)
59. A deterministic finite automaton (DFA) D with
55. Let w be any string of length n in {0, 1}*.
alphabet Σ = {a, b} is given below.
Let L be the set of all substrings of w.
What is the minimum number of states in a non- a, b
deterministic finite automaton that accepts L? b b
p q r
(a) n − 1 (b) n
(c) n + 1 (d) 2n−1
a a
a, b a, b
s t
Solution: n + 1 states are required for string of
length n.
Ans. (c) Which of the following finite state machines is a
valid minimal DFA which accepts the same lan-
56. Let P be a regular language and Q be a context-
guage as D?
free language such that Q ⊆ P. (For example, let P
be the language represented by the regular expres- (a) a, b
sion p*q* and Q be [pnqn | n ∈ N]. Then which of p b q b r
the following is ALWAYS regular? a a
(a) P ∩ Q (b) P − Q s
(c) Σ* − p (d) Σ* − Q
a, b
a, b
(GATE 2011: 1 Mark) s
(b) b
a, b a, b
Solution: Regular languages are closed under p q a r
complementation.
Ans. (c) (c) a, b
a, b a, b
57. The lexical analysis for a modern computer lan- p q b r
guage such as Java needs the power of which one
of the following machine models in a necessary and (d)
sufficient sense? b b
p q
(a) Finite state automata a a
(b) Deterministic pushdown automata s
(c) Non-deterministic pushdown automata a, b
(d) Turing machine
(GATE 2011: 1 Mark)
Solution: Given DFA does not accept “bba” which
Solution: Tokens are recognized in lexical analy- is accepted by option (c) and (d), and string “b”
sis phase of compiler. accepted by option (b). So, option (a) is answer.
Ans. (c) Ans. (a)
Chapter 5.indd 273 4/9/2015 9:58:18 AM

60. Definition of a language L with alphabet {a} is (a) ∅ (b) {e}

given as following. (c) a (d) {a, e}
L = {ank | k < 0, and n is a positive integer constant} (GATE 2012: 1 Mark)
What is the minimum number of states needed in
a DFA to recognize L? Solution: Language accepted by this machine is a+.
First convert NFA to DFA:
(a) (k + 1) (b) (n + 1) a
(c) 2k+1 (d) 2n+1
a
Solution: Let n = 4, so L = a4k
Five states are required for this machine. (n + 1)
states are required for such a machine. Take compliment of this DFA:
Ans. (b)
61. Consider the languages L1, L2 and L3 as given a
below.
a
L1 = {0p1q |p, q ∈ N)
L2 = {0p1q |p, q ∈ N and p = q}
L3 = {0p1q0r |p, q, r ∈ N, r ∈ N and p = q = r).
Language accepted is {e}.
Which of the following statements is NOT TRUE? Ans. (b)
(a) Pushdown automata (PDA) can be used to rec- 64. Consider the set of strings on {0, 1} in which
ognize L1 and L2. every substring of 3 symbols has at most two
(b) L1 is a regular language. zeros. For example, 001110 and 011001 are in
(c) All the three languages are context-free. the language, but 100010 is not. All strings of
(d) Turing machines can be used to recognize all length less than 3 are also in the language. A
the languages. partially completed DFA that accepts this lan-
(GATE 2011: 2 Marks) guage is shown below.
Solution: L1 is a regular language; L2 is a con-
text-free language, and L3 is a context-sensitive 00
language. 0
Ans. (c) 1
0 01
62. Given the language L = {ab, aa, baa}, which of the 0
following strings are in L*? e 0 1 q −0,1
1. abaabaaabaa 0
1 1 10
2. aaaabaaaa
3. baaaaabaaaab
4. baaaaaba 11
1
(a) 1, 2 and 3 (b) 2, 3 and 4
(c) 1, 2 and 4 (d) 1, 3 and 4 The missing arcs in the DFA are
(GATE 2012: 1 Mark)
Solution: baaaaabaaaab is not derived from the (a) 00 01 10 11 q

given language. 00 1 0
Ans. (c) 01 1
10 0
63. What is the complement of the language accepted 11 0
by the NFA shown below?
(b) q
Assume Σ = {a} and e is the empty string. 00 01 10 11
00 0 1
a e 01 1
10 0
11 0
ε (c) 00 01 10 11 q
00 1 0
01 1
10 0
11 0
Chapter 5.indd 274 4/9/2015 9:58:20 AM
(d) 00 01 10 11 q
01 1
10 0
11 0
(b) 00 01 10 11 q
00 0 1
01 1
10 0 GATE PREVIOUS YEARS’ QUESTIONS 275
11 0
(c) 00 01 10 11 q I1 is correct but I2 is not a correct inference.
00 1 0 Ans. (b)
01 1
10 0 66. Which of the following problems are decidable?
11 0 1. Does a given program ever produce an output?
(d) 00 01 10 11 q 2. If L is a context-free language, then, is L also
00 1 0 context-free?
01 1 3. If L is a regular language, then, is L also regular?
10 0
4. If L is a recursive language, then, is L also
11 0
recursive?
(GATE 2012: 2 Marks) (a) 1, 2, 3, 4 (b) 1, 2
(c) 2, 3, 4 (d) 3, 4
Solution: Complete DFA is given below: (GATE 2012: 2 Marks)
00 Solution: Regular and recursive languages are

closed under complementation, but CFLs are
0 0 not.
1 0 0, 1 Ans. (d)
0 01
0
q
67. Consider the languages L1 = Φ and L2={a}. Which
ε 0 1
one of the following represents L1L*2 ∪ L*1 ?
1 0
1 10 (a) {e} (b) Φ
0 1 (c) a* (d) {e, a}
1
(GATE 2013: 1 Mark)
11
1 Solution:
Ans. (d) L*1 = e . So, L1L*2 ∪ L*1 = e
Ans. (a)
65. Consider the following logical inferences.
I1: If it rains then the cricket match will not be played. 68. Which of the following statements is/are FALSE?
The cricket match was played. 1. For every non-deterministic Turing machine, there
Inference: There was no rain. exists an equivalent deterministic Turing machine.
I2: If it rains then the cricket match will not be played.
2. Turing recognizable languages are closed under
It did not rain.
union and complementation.
Inference: The cricket match was played.
3. Turing decidable languages are closed under
Which of the following is TRUE?
intersection and complementation.
(a) Both I1 and I2 are correct inferences
4. Turing recognizable languages are closed under
(b) I1 is correct but I2 is not a correct inference
union and intersection.
(c) I1 is not correct but I2 is a correct inference
(d) Both I1 and I2 are not correct inferences (a) 1 and 4 only (b) 1 and 3 only
(GATE 2012: 2 Marks) (c) 2 only (d) 3 only
(GATE 2013: 1 Mark)
Solution:
R: It rains. Solution:
C: Cricket match played. Option (a) is true. The power of non-deterministic
I1: R → ∼ C can be written as ~R U ∼C and deterministic Turing machines is equivalent.
C Option (b) is True.
∼R Option (c) is false. They are not closed under
I2: R → ∼ C can be written as ∼R U ∼C
complementation.
∼R
Option (d) is true.

∼R U C
Ans. (c)

Chapter 5.indd 275 4/9/2015 9:58:21 AM

69. Consider the DFA A given below. 3. M is a Turing machine. Is L(M) regular?
1 4. A is a DFA and N is an NFA. Is L(A) = L(N)?
1 (a) 3 only (b) 3 and 4 only
(c) 2 and 3 only (d) 2 and 3 only
0 0 (GATE 2013: 2 Marks)
Solution: CFG is empty or not is decidable, and

to test whether language accepted by DFA and
0, 1
NFA is same is also decidable.
Ans. (d)
Which of the following are FALSE?
1. Complement of L(A) is context-free. 71. Consider the following languages.
{{ }}
2. L(A) = L((11*0 + 0)(0 + 1)*0*1*)
L = 0 p1q 0r p, q, r ≥ 0
3. For the language accepted by A, A is the L11 = 0 p1q 0r p, q, r ≥ 0
L22 = {{0 p1q 0r p, q, r ≥ 0, p ≠ r}}

minimal DFA. L = 0 p1q 0r p, q, r ≥ 0, p ≠ r
4. A accepts all strings over {0, 1} of length at least 2.
(a) 1 and 3 only (b) 2 and 4 only
Which one of the following statements is FALSE?
(c) 2 and 3 only (d) 3 and 4 only
(GATE 2013, 2 Marks) (GATE 2013: 2 Marks)
Solution: Option (3) is false because minimal (a) L2 is context-free.

DFA can be constructed using two states only. (b) L1 ∩ L2 is context-free.
Option (4) is false, DFA accepts string “0”, of (c) Complement of L2 is recursive.
length 1. (d) Complement of L1 is context-free but not
Ans. (d) regular.
70. Which of the following is/are undecidable? Solution: L1 is regular and complement of Regular
1. G is a CFG. Is L(G) = ∅? language is also regular.
2. G is a CFG. Is L(G) = Σ*? Ans. (d)
PRACTICE EXERCISES
Set 1
1. What is the meaning of the regular expression Σ* 4. The string 1101 does not refer to the set repre-
0101 Σ*? sented by
(a) Any string containing “1” as substring (a) 110*(0 + 1) (b) 1(0 + 1) * 101
(b) Any string containing “01” as substring (c) (10)*(01)*(00 + 11) (d) (11 + 00) + 01
5. The grammar S → 0S1, S → 0S, S → S1, S → 0
(c) Any string containing “0101” as substring
(d) All strings containing “0101” as substring
will generate a
2. Which of the following is called a regular operation? (a) Context-free language
(a) Union, star (b) Regular language
(b) Concatenation (c) Context-sensitive language
(c) Star, concatenation (d) Push-Down Automata
(d) Concatenation, star, union 6. Which of the string is accepted by a given NFA?
3. In Mealy machine, if we enter the string of length a, b
5, then what will be the output string length?
c
(a) 4 (b) 5
(c) 6 (d) 1
Chapter 5.indd 276 4/9/2015 9:58:21 AM

(a) c . (a ∪ b)* (b) c . a*b* 15. Consider the following context-free languages
(c) c . (ab)* (d) c . (ab) L = {0m1n|m ≠ n; m, n ≥ 0}
7. Let L denotes the language generated by the gram- L1 = {0m1n|m > n ≥ 0} and L2 = {0m1n|n >
mar S → 0S0 | 00. Which of the following is true? m ≥ 0} then
(a) L = 0+ (a) L = L1*L2 (b) L = ∅

(b) L is regular but not 0+ (c) L = L1 ∪ L2 (d) L = L1 ∩ L2
(c) L is context-free but not regular
(d) L is context-free 16. Consider the following NFA with e moves
8. Given an arbitrary non-deterministic finite autom- 0 1 2

aton (NDFA) with N states, the maximum number Start q0 ε q1 ε q2
of states in an equivalent minimized DFA is
(a) N2 (b) 2N
(c) 2N (d) 2*N! If the given NFA is converted to NFA without e
moves, which of the following denotes the set of
9. A language accepted by pushdown automaton is final states?
(a) Regular (b) Context-free (a) {q2} (b) {q0, q2}
(c) Context-sensitive (d) Recursive (c) {q0, q1, q2} (d) {q1, q2}
10. Which one of the following is a regular language? 17. Which of the following strings will not be accepted
(a) {ww | n ≤ m, w ∈ {0, 1}*} by the given NFA in the above question?
(b) {wwR | w ∈ {0*, 1*}} (a) 001122 (b) 1122
(c) {wnwm | n ≤ m, w ∈ {0, 1}} (c) 0121 (d) 22
(d) {w(wR | w ∈ {0, 1}*}
18. If L and L are recursively enumerable then L is
11. DFA and NFA when compared on computational
(a) Regular
power are
(b) Recursive enumerable
(a) DFA ⊆ NFA (c) Context-sensitive
(b) NFA ⊆ DFA (d) Recursive
(c) DFA = NFA
(d) They cannot be compared 19. Let L1 be a recursive language. Let L2 and L3 be
languages that are recursively enumerable but not
12. Number of states in DFA accepting the strings con- recursive. Which of the following statements is not
taining at most 2a’s and at most 2b’s over {a, b} is necessarily true?
(a) 4 (b) 10 (a) L2 — L1 is recursively enumerable
(c) 6 (d) 11 (b) L1 — L3 is recursively enumerable
(c) L2 ∩ L1 is recursively enumerable
(d) L2 ∪ L1 is recursively enumerable
13. Find the minimum number of states in the PDA
accepting language
L = {anbm | n > m; m, n > 0} 20. Let w be any string of length n in {0, 1}*. Let L
Σ = {a, b}, and Γ = {z, a} be the set of all substrings of w. What is the mini-
mum number of states in a non-deterministic finite
(a) n*m (b) 0
automaton that accepts L?
(c) 4 (d) 1
(a) n − 1 (b) n
14. The context-free grammar which generates the lan- (c) n + 1 (d) 2n−1
guage L = {anbm | m = n mod 3} is
(a) S → AB (b) S → AB
21. A minimum state deterministic finite automa-
ton accepting the language L ={w | w ∈ {0, 1}*,

A → aaabbA|e A → AA|e

B → aabb |e B → BBBA|e
number of 0’s and 1’s in w are divisible by 3 and 4,
(c) S → AB (d) S → BA
respectively} has

A → aaaA|e A → BA|e (a) 12 states (b) 11 states
B → ab |aabb |e B → AAAB|e (c) 10 states (d) 8 states
Chapter 5.indd 277 4/9/2015 9:58:22 AM

22. A = {<M>|Turing machine that accepts only one Which should be the final state(s) of the DFA if it
string}. The language A is should accept strings starting with “a” and ending
with “b”?
(a) context-free
(b) recursively enumerable (a) {q0} (b) {q1}
(c) non-recursively enumerable (c) {q0, q1} (d) {q3}
(d) regular
29. Which of the following strings will be accepted in
23. A push-down automata recognizing anbn has mini- the above case if q0 and q1 are accepting states?
mum of n states where n is
1. ababab
(a) 3 (b) 0 2. babaaa
(c) 1 (d) 2 3. aaaba
24. Given the following statements: (a) 1, 2 (b) 2, 3
S1: Every context-sensitive language L is recursive. (c) 1, 3 (d) None of these
S2: There exists a recursive language that is not
30. Which of the following represents the set of accept-
context-sensitive.
ing states if the language to be accepted con-
Which statement is correct? tains string having the same starting and ending
(a) Only S1 is correct. symbols?
(a) {q0} (b) {q0, q3}
(b) Only S2 is correct.
(c) {q3} (d) {q3, q1}
(c) Both S1 and S2 are not correct.
(d) Both S1 and S2 are correct.
31. If L and ~L are recursive enumerable then L is
25. The Greibach normal form grammar for the lan-
guage L = {anbn + 1 | n ≥ 0} is (a) regular and context-free
(b) only context-free
(a) S → aSB, B → bSB |e
(c) regular and context-sensitive
(b) S → aSB, B → bSB |b
(d) recursive
(c) S → aSB |b, B → b
(d) S → aSb |b, B → a 32. The following grammar
26. Recursive enumerable languages are not closed under G = (N, T, P, S)
N = {S, A, B, C, D, E}
(a) Union (b) Intersection T = (a, b, c}
(c) Reversal (d) Complement P: S → ABCD, BCD → DE, D → aD, D → a,
E → bE, E → c is
27. Consider the following statements:
(a) Type 1
I. Recursive languages are closed under com- (b) Type 1 but not Type 3
plementation. (c) Type 2 but not Type 3
II. Recursively enumerable languages are closed (d) Type 0 but not Type 1
under complementation.
III. Recursively enumerable languages are closed 33. Context-free languages are closed under
under union. (a) Concatenation and union
Which of the above statements are true? (b) Kleene closure and union
(c) Concatenation and union
(a) II only (b) I and III
(d) Concatenation, union, Kleene closure
(c) I and II (d) I, II and III
28. Consider the transition diagram of a DFA as given 34. The number of DFAs with the set of states {1,2,
below …n} (n ≥ 1) over the alphabet {0, 1}, with 1 as
a the initial state is
b
Start a b (a) 2n*2 (b) n2
(c) n22n/2 (d) n2n2n
a
b 35. The problem whether the given CFG is finite or
infinite is
a
a
b (a) Undecidable
b (b) NP hard and decidable
Chapter 5.indd 278 4/9/2015 9:58:23 AM

(c) NP hard (c) S

ome non-regular languages cannot be gener-
(d) NP complete ated by CFG.
(d) Both (b) and (c).
36. Let L ={w:w has 3k + 1 b’s ∀ k ≥ 0}and a mini-
mized finite automata D accepting L is constructed. 43. What can be said about a regular language L over
The number of states in D will be {a} whose minimal finite state automation has two
states?
(a) 2 (b) 3
(c) 5 (d) 0 (a) L must be {an| n is odd}
(b) L must be {an| n is even}
37. Which of the following problems are decidable? (c) L must be {an| n > 0}
1. Does a given program ever produce an output? (d) Either L must be {an| n is odd} or L must be
{an| n is even}
2. If L is a context-free language, then is L′ (com-
plement of L) also context-free? 44. Let L be a language recognizable by a finite auto-
3. If L is a regular language, then is L′ also regular? mation. The language REVERSE (L) = {w such
4. If L is a recursive language, then is L′ also recursive? that w is the reverse of v where v ∈ L} is a
(a) 1, 2, 3, 4 (b) 1, 2, (a) Regular language

(c) 2, 3, 4 (d) 3, 4 (b) Context-free language
(c) Context-sensitive language
38. Number of states of the FSM required simulating (d) Recursively enumerable language
behaviour of a computer with a memory capable of
storing “m” words, each of length “n” 45. Consider the following grammar
S → Ax |By
(a) m × 2n
(b) 2mn
A → By |Cw
(c) 2m + n
B → x |Bw
(d) All of these
39. Which of the following statement is correct? C→v
(a) A = {If anbn| n = 0, 1, 2, 3, …} is a regular

Which of the regular expression describes the same
set of strings as the grammar?
language.
(b) Set B of all strings of equal number of a’s and (a) xw*y + xw*yx + ywx
b’s denies a regular language. (b) xwy + xw*xy + ywx
(c) L(A*B*) ∩ B gives the set A. (c) xw*y + xwxyx + ywx
(d) None of these (d) xwxy + xww*y + ywx
40. P, Q, R are three languages; if P and R are regular

and if PQ = R, then Set 2
(a) Q has to be regular. 1. Let R1 and R2 be regular sets defined over the
(b) Q cannot be regular. alphabet, then
(c) Q need not be regular.
(d) Q cannot be a CFL. (a) R1 ∩ R2 is not regular
(b) R1 ∪ R2 is not regular
41. Which of the following definitions below generate (c) ∑* − R1 is regular
the same language as L, where (d) R1* is not regular
L = {xnyn such that n ≥ 1}? 2. Which of the following definitions generates the
1. E → xEy |xy same language as L?
2. xy |(x+xyy+) Where L = {xnyn| n ≥ 1}
3. x+y+ (a) E → xEy|xy (b) xy|x+xyy+
(a) 1 only (b) 1 and 2 (c) |x+y+ (d) None of these
(c) 2 and 3 (d) 2 only 3. Let L ⊆ ∑* where ∑ = {a, b}, which of the follow-
ing is true?
42. Which of the following statement is correct?
(a) L = {x | x has an equal number of a’s and b’s}
(a) All languages cannot be generated by CFG.
is regular.
(b) Any regular language has an equivalent CFG.
Chapter 5.indd 279 4/9/2015 9:58:23 AM

(b) L = {an bn| n ≥ 1} is regular.

(c) L = {x | x has more a’s than b’s} is regular. 0/01
(d) L = {am bn| m ≥ n ≥ 1} is regular.
A B C
1/01 1/10
4. Which one of the following regular expression over
{0, 1} denotes the set of all strings not containing
1/10
0/01
0/00
100 as a substring?
(a) 0*(1 + 0)* (b) 0*101*0 (a) O utputs the sum of the present and the previ-
(c) 0*1*01* (d) 0*(10 + 1)* ous bits of the input
(b) Outputs 01 whenever the input sequence con-
5. If the regular set A is represented by A = (01 + 1)* tains 11
and the regular set B is represented by B = (c) Outputs 00 whenever the input sequence con-
((01)*1*)*, which of the following is true? tains 10
(a) A ⊂ B (d) None of the above
(b) B ⊂ A
11. The smallest DFA which accepts the language L =
{x | length of x is divisible by 3} has
(c) A and B are incomparable
(d) A = B
(a) 2 states (b) 3 states
6. Let S and T be language over ∑ = {a, b} repre- (c) 4 states (d) 5 states
sented by the regular expressions (a + b*)* and
(a + b)*, respectively, which of the following is 12. A context-free grammar is ambiguous if
true?
(a) The grammar contains useless on-terminals
(a) S ⊂ T (b) T ⊂ S (b) It produces more than one parse tree for some
(c) S = T (d) S ∩ T = ∅ sentence
(c) Some production has two non-terminals side by
7. Consider the following statements: side on the right-hand side
S1: {02n | n ≥ 1} is a regular language. (d) None the above
S2: {0m1n0m + n | m ≥ 1 and n ≥ 1} is a regular
13. Fortran is a
language.
(a) regular language
Which of the following is true about S1 and S2?
(b) context-free language
(a) Only S1 is correct. (c) context-sensitive language
(b) Only S2 is correct. (d) none of these
(c) Both S1 and S2 are correct.
(d) None of S1 and S2 is correct. 14. Context-free language and regular languages are
both closed under the operation(s) of
8. Give an arbitrary non-deterministic finite automa-
(a) union and intersection
tion (NFA) with N states, the maximum number
(b) intersection and complement
of states in an equivalent minimized DFA is
(c) union and concatenation
(a) N2 (b) 2N (d) complementation and concatenation
(c) 2N (d) N!
15. Deterministic context-free languages are
9. Consider a DFA over ∑ = {a, b} accepting all (a) closed under union
strings which have number of a’s divisible by 6 and (b) closed under complementation
number of b’s divisible by 8. What is the minimum (c) closed under intersection
number of states that the DFA will have (d) closed under concatenation
(a) 8 (b) 14
(c) 15 (d) 48 16. Consider the grammar with the following production
S → aab|ba |ab
S → aS |b
10. The finite state machine described by the following

state diagram with A as starting state, where an
arc label is x/y and x stands for 1-bit input and y S → abb|ab
stands for 2-bit output Sa → bdb|b
Chapter 5.indd 280 4/9/2015 9:58:23 AM

The above grammar is 24. Recursive languages are

(a) context-free grammar (a) a proper superset of context-free languages
(b) regular grammar (b) also called 0 languages
(c) context-sensitive grammar (c) recognizable by Turing machines
(d) LR(k) (d) all of the above
25. The productions
17. If L1 and L2 are context-free languages and R a
regular set, one of the languages below is not neces- E → E+E
sarily a context-free language. Which one? E → E−E
E → E *E
(a) L1L2 (b) L1 ∩ L2 E → E/E
(c) L1 ∩ R (d) L1 ∪ L2 E → id
18. Which of the following statement is false? (a) generate an inherently ambiguous language
(b) generate an ambiguous language but not inher-
(a) Every finite subset of a non-regular set is ently so
regular. (c) are unambiguous
(b) Every subset of a regular set is regular. (d) can generate all possible fixed length valid
(c) Every finite subset of a regular set is regular. computation for carrying out addition, subtrac-
(d) The intersection of two regular sets is regular. tion, multiplication and division, which can be
expressed in one expression
19. If two finite state machines are equivalent, they
should have the same number of 26. In which of the cases stated below is the following
statement true?
(a) states (b) edges “For every non-deterministic machine M1 there
(c) states and edges (d) none of these exists an equivalent deterministic machine M2 rec-
ognizing the same language”
20. Context-free languages are closed under
I. M1 is non-deterministic finite automation
(a) union, intersection II. M1 is a non-deterministic PDA
(b) union, Kleene closure III. M1 is a deterministic Turing machine
(c) intersection, complement IV. there is no machine M1 for which the given
(d) eomplement, Kleene closure statement is true
21. Let LD be the set of all languages accepted by a (a) I and II (b) I and III
PDA by final state and LE the set of all languages (c) I, II and III (d) II and IV
accepted by empty stack. Which of the following is 27. Which of the following conversion is not possible
true? (algorithmically)?
(a) LD = LE (b) LD ⊃ LE (a) Regular grammar to context-free grammar
(c) LD ⊂ LE (d) None of these (b) Non-deterministic FSA to deterministic FSA
(c) Non-deterministic PDA to deterministic PDA
22. If L1 is a context-free language and L2 is regular, (d) Non-deterministic Turing machine to determin-
which of the following is false? istic Turing machine
(a) L1 — L2 is not context-free. 28. Which one of the following is not decidable?
(b) L1 ∩ L2 is context-free.
(a) G ive a Turing machine M, a string s and an
(c) ~L1 is not context-free.
integer k, M accepts s within k steps
(d) ~L2 is regular.
(b) Equivalence of two Turing machines
(c) Languages accepted by a given finite state
23. The language accepted by pushdown automation
machine is non-empty
in which the stack is limited to 10 items is best
(d) Languages accepted by a context-free grammar
described as
is non-empty
(a) context-free
29. Which of the following is true?
(b) regular
(c) deterministic context-free (a) T
he complement of a recursive language is
(d) recursive recursive.
Chapter 5.indd 281 4/9/2015 9:58:23 AM

(b) The complement of a recursively enumerable (III) Whether a finite state automation halts on all
language is recursively enumerable. inputs
(c) The complement of a recursive language is (IV) Membership problem for type 0 languages
either recursive or recursively enumerable.
(a) II and IV (b) I and II
(d) The complement of a context-free language is
(c) III and IV (d) I and IV
context-free.
33. For two regular languages
30. The C language is
(a) A context-free language L1 = (a + b)*a and L2 = b(a + b)*
(b) A context-sensitive language
(c) A regular language The intersection of L1 and L2 is given by
(d) Parsable fully only by a Turing machine
(a) (a + b)*ab (b) ab(a + b)*
31. Any string of terminals that can be generated by (c) a(a + b)*b (d) b(a + b)*a
the following CFG is
34. Consider the following problem X.
S → XY
X → aX |bX |a “Given a Turing machine M over the input alpha-
bet ∑, any state q of M and a word ∑*, does the
Y → Ya|Yb|a
computation of M on w visit the state q”
(a) Has at least one “b”
Which of the following statements about X is correct?
(b) Should end in “a”
(c) Has no consecutive a’s or b’s (a) X is decidable.
(d) Has at least two a’s (b) X is undecidable but partially decidable.
32. Which of the following problems are undecidable? (c) X is undecidable and not even partially
(I) Membership problem in context-free languages decidable.
(II) Whether a given context-free language is (d) X is not a decision problem.
regular
Set 1
1. (d) 9. (c) 17. (c) 25. (c) 33. (d) 41. (a)
2. (d) 10. (d) 18. (d) 26. (c) 34. (d) 42. (d)
3. (b) 11. (c) 19. (b) 27. (b) 35. (a) 43. (b)
4. (c) 12. (b) 20. (c) 28. (b) 36. (b) 44. (a)
5. (b) 13. (c) 21. (a) 29. (c) 37. (d) 45. (a)
6. (a) 14. (c) 22. (b) 30. (b) 38. (b)
7. (b) 15. (c) 23. (a) 31. (d) 39. (c)
8. (b) 16. (c) 24. (d) 32. (d) 40. (c)
Set 2
1. (c) Regular languages are closed under union, The given language L produces all strings having
interaction, Kleene closure and complementation. equal number of x and y, where y follows symbol x.
So option (c) is correct and remaining options (a), Only expression ⇒ E → Ey | xy can produce equal
(b) and (d) are false. number of x and y, where y follows x.
2. (a) L = {xnyn |n ≥ 1}
Chapter 5.indd 282 4/9/2015 9:58:23 AM

3. (d) Regular language is accepted by DFA. DFA 16. (c)

does not have memory, so it is not able to do com- S → aab|ba |ab
parison between two values, that they are equal or
not (means equal number of a’s and b’s). So only S → aS |b
option (d) does not have comparison, which means S → abb|ab
it can be accepted by DFA. Sa → bdb|b
4. (a) The given production (Sa → bdb |b) contains termi-
nal on the left-hand side, which means this gram-
a. 0*(1 + 0)* can generate string 0000100, which
mar cannot be context-free grammar and regular.
contains substring 100.
So, this grammar is context-sensitive.
b. 0*101*0 generates strings which does not contain
100 as substring. 17. (b) If L1 and L2 are context-free language then
c. 0*1*01* generates strings which does not contain they are closed under
100 as substring.
d. 0*(10+1)* generates all strings which does not (i) Concatenation L1L2
contain 100 as substring. (ii) Union L1 ∪ L2
(iii) Under regular intersection L1 ∩ R
5. (d) Strings in language generated by both the regu- But they are not closed under intersection L1 ∩ L2.
lar expression is equivalent.
18. (b) For (a) and (c), if a set is finite then it means
So, A = B
we can create DFA for that. It does not matter
6. (c) Language generated by both the regular expres- that it belongs to non-regular or regular. If we are
sion is equivalent, so (a + b*)* = (a + b)*. able to draw DFA for that set then it means that
7. (a) L1 = {02n| n ≥ 1}, L1 is regular language because
it is regular. So every finite set of regular or non-
it produces even number of 0’s. L2 = {0m1n0m + n|
regular is regular.
m ≥ 1 and n ≥ 1}, L2 is context-free language For option (d), regular language is closed under
because there are two comparisons. So, S1 is cor- intersection.
rect but S2 is not correct. Let ∑ = (a, b), ∑* is a regular set.
8. (b) NFA with N states, an equivalent minimized L1 = {anbn| n ≥ 0}
DFA having maximum 2N states.
L2 = {anbn| nm ≥ 0}
9. (d) A DFA over ∑ = {a, b} accepting all strings Here, L1 ⊂ ∑* and L2 ⊂ ∑*
which have number of a’s divisible by m and
number of b’s divisible by n have m*n states. In But L1 is CFL and L2 is regular. So from the above
our example we have 6 × 8 = 48 states. explanation, every subset of regular set need not
be regular.
10. (a) FSM outputs the sum of the present and previ-
ous bits of the input. Let us suppose input is 101, 19. (d) Two finite state machines are said to be equiva-
then output is 0 + 1 (last 2 bits) = 01. When input lent if they can accept the same language.
is 11 then output is 1 + 1 = 10.
20. (b) Context-free languages are closed under union
11. (b) The minimal DFA which accepts the language and Kleene closure and not closed under intersec-
L = {x | length of x is divisible by n} always have n tion and complement.
states.
21. (a) PDA accepts its input by two ways:
12. (b) A context-free grammar is ambiguous if there (a) Accepting by entering into final state
is a word which has two different derivation trees. (b) Accepting by empty stack
The acceptance or recognition power of PDA of
13. (c) Fortran is a context-sensitive languages.
both approaches are same because there exists
14. (c) Regular languages are closed under the opera- algorithm to convert PDA accepting by empty
tion of union, interaction, reversal, concatenation stack to PDA accepting by final state and vice
and complementation. Context-free languages are versa, Therefore LD = LE.
closed under the operation of concatenation and
union. 22. (a) L1 = Context-free language (CFL)
15. (b) Deterministic context-free languages are closed L2 = Regular

under complementation. L1 — L2 = L1 ∩ L2C
Chapter 5.indd 283 4/9/2015 9:58:24 AM

L2C is a regular language because regular is closed 28. (b)

under complement and so L1 ∩ L2C is context-free
(a) G ive a Turing machine M, a string s and an
because languages are closed under regular inter-
integer k, M accepts s within k steps is a decid-
section. CFL is closed under regular intersection,
able problem.
so L1 ∩ L2 is a context-free language. CFL is not
(b) Equivalence of two Turing machines is an
closed under complementation, so ∼L1 is not a
undecidable problem.
context-free language. Regular languages are closed
(c) Emptiness of regular language is a decidable
under complementation, so ∼L1 is regular.
problem.
(d) Emptiness of CFL’s is a decidable problem.
23. (b) A PDA with a stack limited to containing lim-
ited items is equivalent to a DFA, and language 29. (a) Recursive languages are closed under comple-
accepted by DFA is regular. So, regular is the best mentation so complement of a recursive language
word to describe it. is also recursive. As every recursive language is
also recursive enumerable. So the complement of
24. (d) Recursive language is a superset of CFL. recursive languages is also recursive enumerable.
Recursive language is called type 0 or unrestricted So both options (a) and (c) are true but (a) is the
language and accepted by Turing machine. strongest answer.
25. (b) Given grammar will generate ambiguous gram- 30. (b) The C language is a context-sensitive language.
mar because for a single string there can be two 31. (d) Has at least 2 numbers of a’s
parse tree according to a given grammar.
32. (a) Only options (II) and (IV) are undecidable
because there exists no algorithm to check that a
26. (b)
CFL is regular and also for membership problem
(a) P
ower of NFA and DFA is same because there for RE language.
exists an algorithm to convert every NFA to 33. (d) Language L1 = {(a + b)*a} is a set of strings
DFA. possible with “a” and “b” and which ends with
(b) Recognition power of NPDA and DPDA are terminal “a”.
Language L2 = {b(a + b)*} is a set of all strings
not same because there is no algorithm to con-
vert NPDA to DPDA.
possible with “a” and “b” and which starts with
(c) Power of NTM and DTM is same because there
terminal “b”.
exists an algorithm to convert every NTM to
So their intersection will give set of strings
DTM.
which starts with “a” and ends with “b”, that is,
(d) Since for finite automaton and Turing machines
b(a + b)*a.
the above statements are true, so this option is
false. 34. (b) There is algorithm to reduce halting problem
of Turing machine to state entry problem. The
27. (c) Since recognition power of NPDA (non-deter- given problem X is a state-entry problem. As halt-
ministic PDA) and DPDA (deterministic PDA) ing problem of Turing machine is undecidable,
are not same because there is no algorithm to con- so problem X also must be undecidable. The lan-
vert NPDA to DPDA. Means there exists some guage corresponding to state-entry problem is RE
languages such as L = {wwR| w ∈ (a, b*} which are but not REC. So X is undecidable, but partially
accepted by NPDA but not by DPDA. decidable.
Chapter 5.indd 284 4/9/2015 9:58:24 AM

UNIT VI: COMPILER DESIGN
LAST SIX YEAR'S GATE ANALYSIS
4
Number of questions
3
Marks 1
2 Marks 2
0
2015 2014 2013 2012 2011 2010
Year Concepts
2015 Control flow graph, Compiler phases, Parsers, LL(1) and
LR(1) Grammar, Shift reducing parser
2014 Parsing, Static and dynamic memory allocation
2013 Parser, Languages
2012 Parsing, Grammar, Decidability
2011 Decidability, Languages
2010 Grammar, Compiler keywords
Chapter 6.indd 285 4/9/2015 9:59:32 AM

Chapter 6.indd 286 4/9/2015 9:59:32 AM
CHAPTER 6
COMPILER DESIGN
Syllabus: Compiler design: Lexical analysis, Parsing, Syntax-directed translation, Runtime environments, Intermediate
and target code generation, Basics of code optimization.
6.1 INTRODUCTION 6.2 COMPILERS AND INTERPRETERS
Computer understands programs written in machine 6.2.1 Compiler

language. Building program in machine language is
a tedious and an error-prone task for human. So the A compiler is a special program that reads statements
programs are written in high-level languages which are written in a language (called source language) and then
easily understood by human. A compiler is a program converts them into another language (called target lan-
that converts high-level program into low-level machine guage). In other words, a compiler is a program which
language that is understood by machine. This subject translates statements written in high-level language (i.e.
deals with how a compiler is designed and organized. Java, C#, Visual Basic) into machine-level language
While writing a compiler, the compilation process is (Fig. 6.1). In the process of translation, a compiler also
divided into various phases. These phases operate in checks for errors if any.
sequence; each phase takes input from the previous
phase and provides output to the next phase.
Chapter 6.indd 287 4/9/2015 9:59:32 AM

288 Chapter 6: Compiler Design
6.2.3.1 Lexical Analyzer
High level language (HLL)
Lexical analyzer reads the source program character by
Compiler character at a time and unites them into a stream of tokens.
Token is a group of character which represents keywords,
operators and identifiers. Character sequence formed by
Input Low level language Run Output tokens is called “lexeme”. Lexical analyzer is also known
Figure 6.1 | Compiler.

as lexer, tokenizer or scanner. If a lexical analyzer gets an
invalid token, then it will generate an error. The lexical
analyzer assigns an id to each token according to their
6.2.2 Interpreter occurrence. It makes entries of each identifier into the
symbol table. As a lexical analyzer cannot enter all infor-
An interpreter takes a single instruction as input and con- mation regarding an identifier such as type and scope, the
verts it into machine-level language and shows errors in the remaining information is inserted by the other phases of
statement if any (Fig. 6.2). It requires less memory than a the compiler into the symbol table. For detail explanation
compiler, and execution of conditional control statements about lexical analyzer, refer to Section 6.3.
are slower. Debugging is easier in an interpreter. 1. Symbol table: A symbol table is a data structure
which stores identifier information related to its
High level language declaration and appearance in a program such as
identifier id, type, scope, value and sometimes loca-
Input Interpreter Run Output tion. Link list and hashing are the common tech-
Figure 6.2 | Interpreter.

niques which are used to construct symbol tables.
Example 6.1
6.2.3 Phases of a Compiler
Statement written in a source program is
A compiler takes a source program written in high-level A = B + C *25;
language as input and produces an equivalent set of Tokens generated by lexical analyzer
machine instruction as output. The compiler process is
A → Identifier
complex, so it is divided into six sub-processes which are
also known as phases of a compiler. The following are = → Assignment operator
different phases of a compiler (Fig. 6.3): B → Identifier
1. Lexical analyzer + → Add operator
2. Syntax analyzer C → Identifier
3. Semantic analyzer * → Multiplication operator
4. Intermediate code generator 25 → Constant
5. Code optimization
6. Target code
Input 6.2.3.2 Syntax Analyzer or Parser
A syntax analyzer is the second and important phase

Lexical analyzer of a compiler. It receives the tokens from the output
of a lexical analyzer as an input. Parser performs two
Syntax analyzer functions:
Semantic analyzer 1. A parser checks that the input tokens from a lexi-
cal analyzer are valid or not according to the speci-
Intermediate code
generator fied grammar of source language.
2. It generates a parse tree according to the given
Code optimization grammar to the source language. The grammar of
source language is given below:
Target code
S → id = P
P → P + T/T
Output T → T *Q/Q
Figure 6.3 | Phases of a compiler. Q → id/Integer constant
Chapter 6.indd 288 4/9/2015 9:59:33 AM

6.2 COMPILERS AND INTERPRETERS 289
Example 6.2 So, here in our example, implicit conversion took place.
Implicit-type casting is also known as coercion.
Parse tree of a given problem A = B + C *25.
6.2.3.4 Intermediate Code Generator
S
The intermediate code generator phase takes a tree as an
input produced by a semantic analyzer and produces an
(A)id = P
intermediate code. The intermediate code thus generated
has mainly two properties: it should be easy to produce
P + T and easy to translate into target program. An intermedi-
ate code can be represented in variety of forms. One of
the forms is the three-address form, which is very similar
T T ∗ Q to the assembly language in which every memory loca-
tion acts like a register. The intermediate code of our
Q Q 25 example is
Source code:
(B)id (C)id A = B + C * 25
Parse tree of A = B + C *25 Intermediate code:
T1 = C * 25
6.2.3.3 Semantic Analyzer T2 = B + T1
A = T2
When the parse tree is generated by a syntax analyzer
and passed as an input to the semantic analyzer, then
6.2.3.5 Code Optimization
the semantic analyzer computes the additional informa-
tion related to the recognized tokens, such as operator,
The code optimization phase is to reduce the size of the
operand, expression or statement, and inserts that infor-
code and improve the performance of the code gener-
mation into the symbol table. The information stored in
ated by an intermediate code phase. The most impor-
the symbol table is frequently used by the other phases
tant part of optimized code is to minimize the amount
of the compiler. During the semantic analysis, the type
of time taken by the code to execute and less common
of identifier is checked. In our example, let all identifiers
is to minimize the amount of memory used by the code.
be float, and 25 be treated as an integer constant. If
Optimized code of our example is
required, the semantic analyzer will perform an implicit-
type conversion, and if it is not possible, then it will Intermediate code: Optimized code:
throw an error. This can be easily understood by the T1 = C * 25 T1 = C * 25
example given in Fig. 6.4.
T2 = B + T1 A = B + T1
A = T2
Semantic analyzer
6.2.3.6 Target Code
S
The target code is the final phase of the compiler which
Float id = P Float normally converts the input obtained from the code opti-
mization phase into the target code (machine code or
Float P + T Float assembly code). The target code of our example is as
Type
mismatch, follows:
Float T T ∗ Q Integer implicit
Float conversion MOV R1, C
takes place
MUL R1, 25
Float Q Q Float 25 Integer MOV R2, B
ADD R2, R1
Float id id Float
STORE A, R2
Figure 6.4 | Semantic analyzer of A = B + C *25. This is the final output of the compiler.
Chapter 6.indd 289 4/9/2015 9:59:35 AM

6.2.4 Grouping of Phases consist of identifiers, operators and operand. A lexical

analyzer also stores the name and id of identifiers in the
Phases deal with the logical organization of a compiler. symbol table (Fig. 6.6).
In an implementation, activities from more than one
phase are often grouped together. Basically, phases are
grouped into two parts. The first part is known as the Error
handler
front end, which consists of initial four phases (lexical
analyzer, syntax analyzer, semantic analyzer and inter-
mediate code) along with symbol table operations and get token( )
error handling. The second part, also known as back end, Lexical
analyzer Parser
consists of the last two phases (code optimization and send token( )
target code). The back end also includes error handling
and symbol table operation (Fig. 6.5).
Symbol
Lexical table
Figure 6.6 | Lexical analyzer.

Syntax
Semantic Front end
Intermediate
code
6.3.1 Functions of a Lexical Analyzer
Code optimization
Target code Back end A lexical analyzer has the following functions.
1. Lexical analyzer divides the given source code or
Figure 6.5 | Grouping of compiler phases.
program into some meaning full words called tokens.
2. It eliminates the comment lines.
3. It finds integer and floating point constant.
6.2.4.1 Compiler Construction Tools
4. It eliminates white-space character such as blank
space and tab.
The compiler writers use software tools such as debug-
5. It helps in giving error message by providing row
gers, version managers, profilers and so on. The follow-
and column numbers.
ing is a list of some useful compiler construction tools:
6. It identifies identifier, keywords, operators and
1. Parser generators: These produce syntax ana- constants.
lyzer from context-free grammar as input.
2. Scanner generators: These automatically pro-
duce lexical analyzer from a specification based on 6.3.2 Implementation of a Lexical Analyzer
regular expressions.
3. Syntax-directed translation engines: These Method of implementing a lexical analyzer or scan-
produce collection of routines from parse tree, gen- ner is regular expression and finite automaton. Some
erating the intermediate code. background information related to regular expression
4. Automatic code generators: These take col- and finite automaton are given in the following sec-
lection of rules that define the translation of tions which will help to understand how a scanner
each operation of the intermediate language into works.
machine language for the target machine.
5. Data-flow engines: Data flow analysis is req
uired to perform good code optimization and data 6.3.2.1 Regular Expression Review
flow engines facilitates the gathering of informa-
tion about how values are transmitted from one 1. Symbol: Letters, digits and special symbols are
part of a program to another part of that program. examples of a symbol.
2. Alphabet: A finite set of symbols through which
we build large structures. An alphabet is denoted
6.3 LEXICAL ANALYZER by Σ, for example, Σ = {0, 1}.
3. String: A finite set of symbol made up of alpha-
bets, for example, a, b are alphabets and aaab, abba
Lexical analyzer is the starting phase of a compiler. It are the strings.
reads the source program character by character at a 4. Empty string: A string which has zero symbols,
time and unites them into a stream of tokens. Tokens and represented by e.
Chapter 6.indd 290 4/9/2015 9:59:35 AM

6.4 PARSER 291
5. Formal language: A set of all possible strings a lexical analyzer when called by the parser to get
which can be generated from given alphabets, and query for the next token. The positions are shown by
represented by Σ*. the circles called states which are connected by edges.
6. Regular expression: The rules that define Here is a finite automaton which recognizes an integer
the set of words that are valid tokens in a (Fig. 6.7).
formal language. These rules are made by three
operators:
••Alternation x|y (x or y) Digit
••Repetition x* (x is repeated 0 or more times)
••Concatenation xy Start Digit
Not Not
6.3.2.2 Finite Automata Review digit digit
Once we have all type of tokens defined by regular Figure 6.7 | Finite automata for recognizing an integer.
expression, we can create a finite automaton for recog-
nizing them. A finite automaton has the following:
1. A finite set of states, one of which is the start state Problem 6.1: Consider the following C program, find
or initial state, and some (maybe none) of which the number of tokens.
are final states.
float average(int a, int b)
2. An alphabet Σ of possible input symbols. {
3. A finite set of transitions that specifies for each float c;
state and for each symbol of the input alphabet, c = (a + b)/2;
which defines that for an input symbol which will return c;
be the next state to go. }
Solution: Every token individual has been under-

Example 6.3 lined. Counting the number of underlines, we have
the number of tokens as 27.
float average ( int a , int b )
a {
a float c ;
q0 a q1 q3 c = (a + b)/ 2;
b return c ;
}
b b
q2 Problem 6.2: Find the number of tokens in the fol-

a,b lowing C statement.
where q0 is the initial state and q3 is the final state. The printf(“k = %d”, i);
given finite automaton accepts a language which has a
string starting with à’ and ending with `b’. Solution: The number of tokens is 7.
printf(“k = %d”, i);
6.3.2.3 Recognition of Tokens
6.4 PARSER
In this section, we will explain how a token is rec-
ognized by a lexical analyzer. A lexical analyzer uses
finite automaton to recognize a token. Transition dia- A parser is a part of a compiler. It takes sequence of
gram is shown below which consists of stages and arcs. tokens from the lexical analyzer as an input and then
Arcs show the transition from one state to another builds a data structure in the form of a parse tree
state. Transition diagram describes the working of (Fig. 6.8). A parser’s main purpose is to determine if
Chapter 6.indd 291 4/9/2015 9:59:37 AM

the input data may be derived from the start symbol 6.4.2 Derivation Tree or Parse Tree
of the grammar. Depending upon the method how
the parse tree is derived, we have two types of parsers— The string generated by a CFG G = (Vn, T, S, P) is rep-
top-down parser and bottom-up parser (discussed resented by a hierarchical structure called tree. A deriva-
shortly). tion tree or parse tree for a CFG is a tree that satisfies
the following condition:
1. If A → a1, a2, a3, …, an is a production in G,
Grammar then A becomes the father of nodes, labelled a1, a2,
a3, …, an.
V→T
Source
program 2. The root has label S (starting symbol).
3. Every vertex (or node) has a label.
4. Internal nodes should be labels with variables only.
Parser Parse 5. The leaves nodes are labelled with e or terminal symbol.
tree 6. The collection of leaves from left to right yields the
string w.
Error
handler Problem 6.4: Consider the grammar S → S +
S |S *S |a|b. Construct a derivation (or parse) tree for
Figure 6.8 | Parser. the string w = a + b *b.
Solution:
6.4.1 Context-Free Grammar S
A grammar G = (Vn, T, S, P) is said to be a context-free
grammar (CFG) if the production P = {u → v} of G are S + S
of the form u → v and satisfy the following conditions:
1. u → v, where v ∈ (V ∪ T ) * , and V stands for vari-
a S * S
able and T for terminal
2. u → v, where u ∈ Vn
3. u ≤ v (length of u is less than v) b b
4. Only single variable is allowed in left side (means u
has single variable only)
As we know that a CFG has no context either left or right, 6.4.2.1 Leftmost Derivation Tree
this is the reason why it is also known as context-free.
A derivation tree is called a leftmost derivation (LMD)
tree if the ordering of decomposed variable is from left to
Problem 6.3: Consider a grammar G = (Vn, T, S, P) right. Thus, for generating string w = aab from grammar:
having production S → aSa|bSb|x. Check the produc-
tion and find the language generated. S → AB (production 1)
A → aaA (production 2)
Solution: A → e (production 3)
Let P1: S → aSa B → bB (production 4)
P2: S → bSb B → e (production 5)
P3: S → x LMD:
(a, b, x) are terminals and S is a variable. As all the pro- S → AB (by production 1)
duction are of the form A → a, where a ∈ (Vn ∪ S ) * S → aaAB (by production 2)
and A ∈ Vn, hence G is a CFG. And it will produce
S → aaB (by production 3)
context-free language.
S → aabB (by production 4)
Language generated: L(G) = {wxwR : w ∈ (a + b)*}
Chapter 6.indd 292 4/9/2015 9:59:38 AM

6.4 PARSER 293
6.4.2.2 Rightmost Derivation Tree 6.4.3.1 Removal of Ambiguity
A derivation tree is called rightmost derivation tree 1. Removal of left recursion: A production of
(RMD) if the ordering of decomposed variable is from grammar G = (Vn, T, S, P) is said to be left recur-
right to left. Thus, for generating string w = aab from sive grammar if it has one of the productions in the
the above grammar: given form:
RMD: A → Aa, where A is a variable and a ∈ (Vn ∪ S)*
S → AB (by production 1) Elimination of left recursion: Let the variable A
S → AbB (by production 4) have left recursive problem as following:
S → Ab (by production 5) A → Aa1|Aa2|Aa3|…|Aan|b1|b2|b3|…|bm
S → aaAb (by production 2) where b1, b2, …, bm do not begin with A. Then we
replace A production by:
{A → b1A′|b2A′|b3A′|…|bnA′|}
where A′ → a1A′|a2A′|a3A′|…|anA′|e
S
Example 6.4
A B Let grammar S → S + S |S *S |a|b|c
To eliminate left recursion, the grammar S is replaced by
a a A b B S ′ → +SS ′|*SS ′|e
S → aS ′|bS ′|cS ′
e e
Left to right 2. Removal of left factoring: In grammar G, two
or more productions of variable A are said to have
left factoring if the productions are in the form:
6.4.3 Ambiguous Grammar A → ab1|ab2|ab3…|abm
A grammar G is called ambiguous if for some string where {b1|b2|b3…|bm} ∈ (Vn ∪ S)* and does not
w ∈ L(G), there exist two or more derivation tree (two start with a. All these production have common
or more LMD or two or more RMD tree). Let us consider left factor a.
a CFG grammar having production: Elimination of left factoring: Let variable A have
S → S + S |S *S |a|b, for string w = a + a *a have more (left factoring) production as follows:
than one LMD tree. A → ab1|ab2|ab3…|abm|γ1|γ2|γ3|…|γm
where γ1, γ2, γ3, …, γm and {b1, b2, b3…bm} do not
S
S contain a as a prefix, then we replace this produc-
tion by
S + S S * S A → aA′|γ1|γ2|γ3|…|γm
A′ → b1|b2|b3…|bm
S + S a
a S * S
Example 6.5
a a
a a Let grammar A → abc|abd|abe
To remove left factoring, we have
A → abA′
Note: A language (L) is called ambiguous if and only
if every grammar which generates it is ambiguous. The
only known ambiguous language is {anbmcn} ∪ {anbmcm}. A′ → c|d|e
Left recursion and left factoring is the major cause for

a grammar to be ambiguous. But presence of these in 6.4.4 Top-Down Parser
a grammar does not mean that grammar is ambiguous;
and similarly absence of these does not mean that the In top-down parser, parse tree construction starts from
grammar is unambiguous. the root and proceeds to the leaf. Top-down parser uses
Chapter 6.indd 293 4/9/2015 9:59:39 AM

LMD for constructing the parse tree. When a variable Problem with Recursive Descent Parser
contains more than one choice, choosing the correct
1. Recursion is used to generate parse tree.
production is always going to be difficult. In top-down
2. More time is wasted in backtracking.
parsing, no left recursion and no left factoring exist.
3. Recursive descent parser can enter into an infinite
loop if the given grammar contains left recursion.
Problem 6.5: String w = abbcde, grammar G is given 4. Time complexity of RDP is O(2n).
below: 6.4.4.2 Top-Down Parser without
S→aABe Backtracking
A→Abc/b
A predictive parser is a special class of recursive decent
B→D
parser. The goal of predictive parsing is to construct
Draw a parse tree for the production of a string w a top-down parser that does not require backtracking.
with the help of grammar G. Predictive parsing technique can only be used for class
of LL(k) grammars, where k is some integer. In LL(k)
Solution: grammar, by seeing k tokens a recursive decent parser
decides which production should be examined so that
S the LL(k) grammars are able to exclude all grammars
that are ambiguous and having left recursion (Fig. 6.9).
a A B e
Input
A b c d
LL() Parser Output
b $
Stack
Parse tree for string w = abbcde.
LL() Action LL() Goto
table table
Top-down parsing can be performed by the following Parsing table
Figure 6.9 | Predictive parser.

two methods:
1. Top-down parsing with backtracking
2. Top-down parsing without backtracking
6.4.4.3 Algorithm of Predictive Parser
6.4.4.1 Top-Down Parsing with Let x be the top of stack and à’ be the look-ahead
Backtracking symbol. Then
One of the most straightforward forms of parsing is recur- 1. If (x = = a = = $), then successful complete.
sive descent parsing (RDP). This is a top-down process in 2. If (x = = a = = $), then pop the stack element and
which the parser attempts to verify that the syntax of the increment input pointer.
input stream is correct as it is read from left to right. The 3. If (x is a variable), then see the LL(1) action pars-
pseudocode of recursive descent parser is given as follows: ing table M.
If M[x1, a] = x → uvw, then replace x by uvw in the
RDP(S) reverse order.
{ 4. If M[x1, a] = blank, then parsing error.
Choose a production S →x1,x2,x3,x4
for(i=1 to n) Problem 6.6: Draw a top-down predictive parser
{ using the following parser table.
if(xi is variable)
RDP(i)
Grammar of given language
else if(xi==lookahead) S → (L)/a
increment i/p pointer
L → SL′
else
error(choose another production) L′ → e/,SL′
}
String w = (a, a, a)
}
Chapter 6.indd 294 4/9/2015 9:59:40 AM

6.4 PARSER 295
Parsing Table 6.4.4.4 LL(1) Parsing Table Construction
A , ( ) $ 1. First set: First(A) gives a set of all terminals that

S→a S → (L)
may begin in a string derived from A. To get first
S
of A, start finding all strings generated by that
L L → SL′ L → SL′ variable in a language and then pick the first ele-
L′ L′ →, SL′ L′ → e ment of every string.
Rules:
Solution: ••If S → ab/cd/ef then
First(S) = {a, c, e}
S. No. Input Stack Production ••If S → ab then
1. (a, a, a)$ $S S → (L) First(S) = {a}
First(e) = (e)
First(a) = {a}, where à’ is terminal
$)L(
2. a, a, a)$ $)L L → SL′ ••If S → AB
$)L′S S→a A→a
B→b
then First(s) = {a}
$)L′a
3. , a, a)$ $)L′ L′ →, SL′ ••If S → aS/b then
First(S) = {a, b}
$)L′S, ••If S → AB
4. a, a)$ $)L′S S→a A → a/c/d/e
B→b
then First(S) = {a, c, d, b}
$)L′a
5. , a)$ $)L′S, L′ →,SL′ 2. Follow set: Follow(A) gives a set of all terminals
that may follow immediately to the right of A. e
6. a)$ $)L′S S→a
can never be a part of any follow set.
$)L′a
Rules:
7. )$ $)L′ L′ → e ••If S is the starting symbol, then $ is also one of
the element of Follow(S).
$) ••If S → aBCDE, then
8. $ $
Follow(C) = First(DE) = First(D)
••If S → AB or S → ABC and C → e, then
By following these steps we can generate a parse tree
for the string w = (a, a, a). Follow(B) = Follow(S)
Problem 6.7: Find the first and the follow sets for the
S
given grammar:
E → TE ′
C L J
E ′ → e/ + TE ′
T → FT ′
S L T ′ → e/*FT ′
F → id/(E)
,
Solution:
a S L
Variable First( ) Follow( )
a , S L′ E id, ( $, )
E ′ +, e $, )
a ¨ T id, ( +, $, )
T ′ *, e +, $, )
Parse tree for the string w = (a, a, a). F id, ( *, +, $, )
Chapter 6.indd 295 4/9/2015 9:59:40 AM

Problem 6.8: Find the first and the follow sets for the Number of cells in parsing table = {V *(T + 1)}
given grammar: So, table entries = {5*(5 + 1)} = 30
S → (L)/a Parsing table constructed with the help of the first
L → SL′ and follow sets of given grammar:
L′ → e/,SL′
Id + * ( ) $
Solution:
E E → TE ′ E → TE ′
Variable First( ) Follow( ) E ′ E ′ → +TE ′ E ′ → e E ′ → e
S (, a ,, ), $ T T → FT ′ T → FT ′
L (, a )
T ′ T ′ → e T ′ → *FT ′ T′ → e T ′ → e
L′ ,, e ) F F → id F → (E)
6.4.4.5 Algorithm for Constructing a How to Check a Given Grammar (G) Is LL(1) or Not?
Parsing Table
The following are two ways to identify that a given
If a grammar G is given, the steps to construct a parsing grammar is LL(1) or not:
table is as follows:
1. By constructing parsing table, if parsing table does
For each production A → a, do the following not have more than one production entry in any of
1. Add A → a under M[A, b], where b ∈ First(a). the cell then only it is LL(1).
2. If First(a) contains e, then add A → a under 2. By checking the following two conditions:
M[A, c], where c ∈ Follow(A). ••If a given grammar G does not contain null(e)
production:
Number of parsing table entries = {V *(T + 1)}, where Let grammar G be
V is the number of variables in the given grammar and T
is the number of terminals in the given grammar. A → a1/a2/a3
Then a1, a2, a3 should be pair wise disjoint.
That is,
Problem 6.9: Construct the parsing table from the
following grammar: First(a1) ∩ First(a2) = ∅
E → TE ′ First(a2) ∩ First(a3) = ∅
E′ → e/ + TE ′ First(a3) ∩ First(a1) = ∅
T → FT ′ ••If grammar G contains null(e) production:
T′ → e/*FT ′ Let grammar G be
F → id/(E) A → a1/a2/e
Solution: Find First(a1), First(a2) and Follow(A), they
The first and the follow sets for the given grammar should be pair wise disjoint.
are as follows:
First(a1) ∩ Follow(A) = ∅
Variable First( ) Follow( ) First(a2) ∩ Follow(A) = ∅
E id, ( $, )
E ′ +, e $, ) Problem 6.10: Check whether the following grammar
T id, ( +, $, ) is LL(1) or not.
T ′ *, e +, $, ) S → E/a
F id, ( *, +, $, )
E→a
Chapter 6.indd 296 4/9/2015 9:59:41 AM

6.4 PARSER 297
Solution: In the given grammar there is no null pro- Draw a parse tree for reduction of a string w with the
duction, so help of the grammar G.
First(S) = a Solution: The parse tree for reduction of the string
w = abbcde.
First(E) = a
According to the given condition 1, First(S) ∩ S

First(A) ≠ ∅. So the given grammar is not LL(1).
a A B e
Problem 6.11: Check the following grammar is LL(1)
or not.
a A d e
S → aABb
A → a/e
a A b c e
B → d/e
d
Solution: In the given grammar there is null pro- a b b c d e

duction, so find first and follow of those productions
which have null production.
 First(A) = a 
Follow(A) = d, b  No common element 6.4.5.1 Classification of Bottom-Up Parser
 
 First(B) = d  The bottom up parser is classified as shown in Fig. 6.10.
  No common element
Follow(B) = b 
Bottom-up parser
Important Points
LR Parser Operator precedence
1. Every regular grammar need not be LL(1) because Unambigious parser
that grammar may contain left factoring. grammar Operator grammar
2. Any ambiguous grammar cannot be LL(1).
3. If any grammar contains left factoring, then it
cannot be LL(1) grammar. LR(0) SLR(1) CLR(1) LALR(1)
4. If a given grammar contains left recursion, then it
cannot be LL(1) grammar. Figure 6.10 | Classification of bottom-up parser.
LR Parser
6.4.5 Bottom-Up Parser
In computer science, an LR parser is a type of bottom-
In bottom-up parser, parse tree construction starts from up parser that efficiently handles context-free languages in
children and proceeds to root. Bottom-up parser uses guaranteed linear time. An LR parser reads input from left
RMD for constructing the parse tree. A substring which to right and produces an RMD. The canonical LR (CLR),
will give reduction of string is called handle. look-ahead LR (LALR) and simple LR (SLR) parsers are
common variants of LR parsers. For all types of LR parsers,
The difficulty with bottom-up parser is identifying the
parsing algorithm is same but parsing tables are different.
right handle which will give one required variable so that
we will go to start symbol. LR Parsing Algorithm
Let S be the state on top of the stack and à’ be the look-
Problem 6.12: String w = abbcde, grammar G is ahead symbol, then
given below:
1. If action [S, a] = Si, then shift à’ and ì’, and also
S → aABe increment the i/p pointer.
A → Abc/b 2. If action [S, a] = rj and rj is a → b, then pop 2|b|
symbols and replace by a. If Sm−1 is the state below
B→D
a, then push Goto [Sm−1, a].
Chapter 6.indd 297 4/9/2015 9:59:42 AM

3. If action [S, a] = accepted, then successful parsing.

Problem 6.13: Consider the following grammar and
4. If action [S, a] = blank, then parsing error.
construct the LR parsing table.
Stack Operations S → AA

(production number 1)
1. Shift: Shifts the next input symbol onto the top A → aA (production number 2)
of the stack. A → b (production number 3)
2. Reduce: Replaces a set of grammar symbol or
handle on the top of the stack with the LHS of a Solution:
production rule. Step 1: Construct the augmented grammar.
3. Accept: Declares successful completion of the S′ → S
S → AA
parsing.
LR(0) Parser A → aA
A→b
1. LR(0) parsing table construction: For con-
structing LR(0) parsing table we have to follow the Step 2: Find closure.
given steps: I1
••Construct augmented grammar. S′ → S.
Let given grammar G be S
S → AA
I2
S′ → .S S → A.A I5
A → aA/b
A
S → .A
A A → aA S → AA.
Augmented grammar G′ will be A → .aA a A → .b
S ′ → S A → .b I3 a
A → a.A
b I6
S → AA A → aA
I0 aA
A →aA.
A → .b
A → aA/b
I4
S ′ → S is called the augmented production. A → b. b
b
••Find closure {I0 = closure (S ′ → S )}. ·
Step 3: Construct DFA.

To find closure of a production I, follow these
steps: I1
In closure of I, add I also to closure (I ).
S
If I is A → B CD and C → EF is in the given
·
grammar G, then add C → EF to closure (I ). ·

A I2 A I5
I0
Repeat these steps for every newly added
LR(0) item. a a
This can be understood clearly by the given exam- b b A
ple; it uses the above grammar to find closure. I3 I6
Closure (S ′ → S ) = {S ′ → S |S → AA|A → aA|
· · · ·
A → b} · I b
a
Closure (S → AA) = {S → AA|A → aA|A → b}
· · · ·
Closure (A → aA) = {A → aA}

· · Step 4: Construct LR(0) parsing table.
••Using I0 construct deterministic finite automata KK Si denotes the shift move in the above DFA, where
(DFA). ì’ denotes the state number.
••Reduce DFA into LR(0) parsing table. KK ri represents reduction move in the above DFA,
••Find Goto (I, X). where ì’ denotes the production number.
Goto (I, X) function is used to fill the second
part of the parsing table, which have entries Action Goto
related to move after X. State A b $ S A
I0 S3 S4 1 2
Example 6.6 I1 Accepted
I2 S3 S4 5
Consider the above grammar for the Goto function. I3 S3 S4 6
KK Goto (S′ → S, S) = S′ → S.
· I4 r3 r3 r3
KK Goto (S′ → AA, A) = S → S |A → aA|A → b
· · · ·
I5 r1 r1 r1
KK Goto (A → aA, a) = A → a A|A → aA|A → b
· · · ·
I6 r2 r2 r2
Chapter 6.indd 298 4/9/2015 9:59:43 AM

6.4 PARSER 299
Important Points
Problem 6.14: Check whether the following grammar
is LR(0) or not. 1. If there are two reductions on any state of DFA,
then in table entry we have to put both in that cell.
S → (L) (production number 1)
In that situation parser will not be able to decide
S → a (production number 2) for which reduction it has to parse, so this problem
L → L, S (production number 3) is called reduce-reduce (R-R) problem.
L → S (production number 4) 2. If in any state one production is reduced and
another is shifted then also parser is not able to
Solution: parse it. This problem is called shift-reduce (S-R)
Step 1: Construct the augmented grammar. problem.
S ′ → S 3. Due to DFA no shift-shift (S-S) problem occurred.
4. For conflict, there should be two productions and
S → (L) at least one of them should be reduced.
S→a 5. Goto section does not participate in conflict.
L → L, S
L→S
Step 2: Find closure (I0 = S′ → S). Problem 6.15: Check if the following grammar is
I1 LR(0) or not.
S S→ S.
I0
E → T + E …….(production number 1)
I4 I6
I2 )
S′→ .S L S→ (L.) S→ (L).
S→ .(L) ( S→ (.L)
L→ .L,S
L→ L.,S E → T …….........(production number 2)
S→ .a I7 I8
L → .S
(
T → id ……........(production number 3)
,
L→ L,.S
I 5
a S→ .(L) S L→ L,S.
L → S. S→ .(L)
I3 S→ .a S→ .a
S→ a. a
( Solution:
a Step 1: Construct an augmented grammar.
Step 3: Construct DFA. E′ → E
I1 E→T+E
I6
E→T
S )
T → id
I
I0 ( L ,
I2 ( I7
Step 2: Find closure.
a S I
a S I1
I3 a I8 I0 E E′ → E.
I4
E′ → .E I2
E → T+.E I5
E → .T+E E → T.+E → .T+E E → T+E.
T + E E
Step 4: Construct LR parsing table.
E → .T E → T. E → .T
T → .id
T T → .id
Action Goto I7
id
State A ( ) , $ S L T → id. id
I0 S3 S2 1
I1 Accepted Step 3: Construct DFA.
I2 S3 S2 5 4
I3 r2 r2 r2 r2 r2 I1
I4 S6 S7 E
I5 r4 r4 r4 r4 r4
I6 r1 r1 r1 r1 r1 T + E
I0 I2 I4 I5
I7 S3 S2 8 T
I8 r3 r3 r3 r3 r3
id
The given grammar is LR(0) because there are no I3 id
multiple entries in the same cell.
Chapter 6.indd 299 4/9/2015 9:59:44 AM

Step 4: Construct LR() parsing table. ••Also add I to closure(I ).

••Repeat above steps for every newly added items.
Action Goto 2. Goto function of CLR:
+ Id $ E T Goto(I, X)
••Add I to Goto(I, X) and also move dot( ) after X. ·
••Apply closure to the result obtained in the pre-

I0 S3 1 2
I1 Accepted
vious step.
I2 S4/r2 r2 r2
I3 r3 r3 r3
I4 S3 5 2 Problem 6.16: Construct parsing table for the follow-
I5 r1 r1 r1 ing grammar and also check that the given grammar
is CLR(1) or not?
The given grammar is not LR(0) because there are S → AA
multiple entries (S4/r2) in the same cell. This conflict A → aA
is called as SR (shift-reduce) conflict.
A→b
SLR(1) Parser Solution:
SLR(1) parser is one of the variants of LR parser. An SLR Step 1: Construct an augmented grammar.
parser is efficient at finding the single correct bottom- S′ → S
up parser in a single scan without backtracking. SLR(1) S → AA
parser table has only one difference from LR(0) pars- A → aA
ing table, reduction entries are made only in the specific
location. Let production E → T be reduced, then entry A→b
of the reduced production will be in those column which Step 2: Find closure.
comes in Follow(T). If there are multiple entries in the I1
S′→ S. ,$
I5
same cell, then the given grammar will not be SLR(1).
S S′→ AA.,$
From Problem 6.15, we will construct parse table I2 A
I1
I7
only for SLR(1) (Table 6.1). State 2 has shift and reduce A S′→ A. A,$ b
S′→ S,$ A→ a.A,$ A→ b. ,$
entry, but we have to find that where we have to put the S→ .AA,$ A→ .b,$
reduce entry. The production which is reduced is E → T, a b
A→ .aA,a/b, I6 I9
so we will put r2 at places which come in Follow(T). A→ .b,a/b, a
I3
A→ a.A,$
A A→ aA.,$
A→ a.A,a/b A→ .aA,$ a
Follow (T) = {+, $} b A→ .aA,a/b
a
A→ .b,$
A→ .b, a/b A
Table 6.1 | Parsing table for SLR(1)
I8
I4 b A→ aA.,a/b
Action Goto A′→ b., a/b
+ Id $ E T
Step 3: Construct DFA.
I0 S3 1 2
I1 Accepted I1
I2 S4/r2 r2 S I5
A
I3 r3 r3 r3 a
I4 S3 5 2 A a A
I2 I6 I9
I5 r1 r1 r1 I0
a b b
I7
There are two entries in the same cell, so this is not b I3 A
SLR(1) grammar. b I8
Canonical LR Parser I4
Canonical LR parser is a simplified version of an LR parser
and is also known as CLR parser. CLR parser is the most Step 4: Construct parsing table.
powerful parser of the LR parser family. Closure and Goto
function are processed differently than in SLR parser. Action Goto
1. Closure function of CLR: State A B $ S A
Closure(I) I0 S3 S4 1 2
••If I is A → B CD, $ and C → EF is in G, then
· ·
I1 Accepted
add C → EF, {First(D), $}.
·
(Continued)
Chapter 6.indd 300 4/9/2015 9:59:45 AM

6.4 PARSER 301
Continued Problem 6.18: Convert the following grammar into

Action Goto operator grammar.
State A B $ S A P → SR|S
I2 S6 S7 5 R → bSR|bS
I3 S3 S4 8 S → WbS|W
I4 r3 r3 W → L*W|L
I5 r1 L → id
I6 S6 S7 9
I7 r3 Solution:
I8 r2 r2
The operator grammars for the given grammars are:
I9 r2
P → SbP |SbS|S
Here, in our example we have three productions which R → bP|bS
are reduced.
S → WbS|W
S → AA production 1 (represented by r1)
W → L*W|L
A → aA production 2 (represented by r2)
L → id
A → b production 3 (represented by r3)
To know in which place r3 will get entry, check what
the look-ahead terminals in the closure are of the given Important Points
production of the grammar. Production A → b has
{a, b as look-ahead terminals. So reduction r3 will get 1. The relation between CLR(1), LALR(1), SLR(1),
entry under a and b column. Given grammar is CLR(1) LR(0) and LL(1) is shown in Fig. 6.11.
grammar because there are two entries in the same cell.
Unambiguous
Note: CLR(1) parser is more powerful than other variants CLR(1)
of LR parser, but it is costlier than the other due to more LALR(1)
number of states in the DFA of CLR(1), even though two
states are exactly the same other than look-ahead symbol. SLR(1)
So by merging those two states together, CLR(1) can be LR(0) LL(1)
minimized and the minimized CLR(1) is called LALR(1).
Operator Precedence Parser: It is based upon
bottom-up parsing technique that follow shift-reduce
parsing method. Operator precedence parser interprets
an operator precedence grammar. Operator precedence
parser is capable of parsing all LR(1) grammars.
Figure 6.11 | Relation between CLR(1), LALR(1),
Operator grammar: A grammar G is said to be opera-
SLR(1), LR(0) and LL(1).
tor grammar iff
1. G does not have null production. 2. The relation between CLR(1), LALR(1), SLR(1),
2. G does not have two adjacent variables on the LR(0) and operator precedence parser is shown in
right-hand side of the production. Fig. 6.12.
Problem 6.17: Which is/are operator grammar in the Unambiguous

given set? CLR(1)or LR(1)
LALR(1)
(a) E → E + E |E *E |id SLR(1)
(b) E → AB, A → a, B → b Ambiguous LR(0)
(c) E → E + E |E *E |id|e
Solution: Operator
precedence
(a) Operator grammar parser
Figure 6.12 | Relation between CLR(1), LALR(1),
(b) Not operator grammar
(c) Not operator grammar
SLR(1), LR(0) and operator precedence parser.
Chapter 6.indd 301 4/9/2015 9:59:46 AM

3. If grammar G is LL(1), then Table 6.2 | Characteristics of S-attribute and

••It may be LR(0). L-attribute definitions
••It may be SLR(1).
••It will surely be LALR(1). S Attribute Definition L Attribute Definition
4. Every LL(1) grammar is surely LALR(1), but if S-attribute definition L-attribute definition
any grammar is LALR(1) then it may or may not uses only synthesized uses both synthesized and
be LL(1). attributes. inherited attributes.
5. If any grammar is LALR(1), then it will surely be Semantic rules should be Semantic rules can be placed
CLR(1). placed at the rightmost anywhere on the right-hand
6. CLR(1) is also known as LR(1). place of the right-hand side. side of its production.
7. Every LL(1) grammar will surely be CLR(1). Evaluated by a bottom-up Evaluated by a bottom-up
8. All LL(k) parsers are subset of LR(k) parser. or post-order traversal of a or pre-order traversal of a
9. If the number of states in LR(0), SLR(1), CLR(1) parse tree. parse tree.
and LALR(1) is n1, n2, n3, n4, respectively, then
the relation between them is
6.5.3 Applications of SDT
n1 = n2 = n4 ≤ n3
1. Evaluating arithmetic expression
2. Creating syntax tree
3. Converting infix to postfix
6.5 SYNTAX-DIRECTED 4. Converting infix to prefix
TRANSLATION 5. Generating intermediate code
6. Converting binary to decimal
7. Storing type information into symbol table
Syntax-directed translation (SDT) is a method of com-
piler implementation which attaches semantic rules with Problem 6.19: Construct SDT to evaluate the given
every production of a CFG while translating a string arithmetic expression:
into a sequence of actions. An SDT can be implemented
by first constructing a parse tree and then performing Input: 2 + 3*4
the actions in a pre-order traversal. Output: 14
Solution:
Example 6.7 +Grammar Semantic Rules
Let a production be E → E1 + E2, then SDT will be
E→E+T .val = E.val + T.val
E
E ⋅ val = E1 ⋅ val + E2 ⋅ val
Print (E.val);
E → E1 + E2   E→T
Pr int (E ⋅ val);
E.val = T.val
 
T → T *F E.val = T.val * F.val
T→F T.val = F.val
F → id F.val = id
6.5.1 Attributes of Syntax-Directed Translation Parse Tree:
E E.val = 12+2 =14
The following are two types of attributes supported by
the SDT: E.val = 2
1. Synthesized attribute: An attribute is said E + T T.val = 3 × 4 = 12
to be synthesized only if its value is calculated in
terms of its children in the parse tree. T ∗ F T.val = 4
2. Inherited attribute: An attribute is said to be T.val = 2T T.val = 3
inherited attribute only if its value is calculated
in terms of its parent or children or both in the
parse tree. F.val = 2F F.val = 3F 4 id
6.5.2 Types of Syntax-Directed Translation 2 id 3 id
An SDT is basically divided into two types: S-attribute Parse tree for given arithmetic expression
definition and L-attribute definition (Table 6.2). 2 + 3*4.
Chapter 6.indd 302 4/9/2015 9:59:46 AM

6.6 RUNTIME ENVIRONMENT 303
Problem 6.20: Consider the following SDT: Problem 6.22: Construct an SDT to convert infix
S → TR
expression to prefix expression.
R → +T {print (+);} R

Input: a + b*c
R→e
Output: + a*b c
T → Num {print (num);} Solution:
For input string (9 + 5 + 2), what will be the output? The SDT is constructed as follows:
E → {print (+)} E + T
Solution:
E→T
We first draw the parse tree and then determine the
output. T → {print (*)} T*F
S T→F
F → id {print (id)}
T R
Num(9) +
6.6 RUNTIME ENVIRONMENT
T Print(+) R
To generate the target code, there is a need to know

Num(5) + T Print(+) R
about the environment where the source program will
execute. It consists mapping of names and objects in the
Num(2) ε memory, procedure activation, storage allocation, library
routines and exception handling, scopes and extents of
Parse tree for given arithmetic expression declarations and symbol table organization.
(9 + 5 + 2).
Output = 95 + 2+
6.6.1 Storage Organization
Problem 6.21: Construct an SDT to convert infix Runtime memory needs to be subdivided as follows to
expression to postfix expression. hold the different components of an executing program
(Fig. 6.13):
Input: a + b*c
Output: abc* +
Program code
Solution:
Static global data
The SDT is constructed as follows:
E → E + T {print (+)} Stack
E→T
T → T *F {print (*)}
T→F
F → id {print (id)}
E
Heap
Figure 6.13 | Storage organization.

E + T
1. Program code: Refers to static area used by the

T T * F generated target code which is fixed at compile
time.
2. Static global data: Refers to storage space for
F F id(c) data, which does not change during the execution
of the program.
3. Stack: Manages activation of procedures at run-
id(a) id(b) time. The area usually grows towards lower address.
Parse tree for given infix expression a + b*c. 4. Heap: Holds variable created at runtime.
Chapter 6.indd 303 4/9/2015 9:59:47 AM

There are two different approaches for runtime storage Solution: The activation takes place as follows:
allocation, as given in Table 6.3.
System starts main
Table 6.3 | Static and dynamic allocation enter f(5)
Static Allocation Dynamic Allocation enter f(4)
enter f(3)
Allocates all needed space Allocates space when it is enter f(2)
when program starts. needed. exit f(2)
Deallocates all space Deallocates space when it enter f(1)
when program terminates. is no longer needed. exit f(1)
exit f(3)
enter f(2)
6.6.2 Activation Record and Activation Trees exit f(2)
exit f(4)
enter f(3)
Activation is the function in execution mode. The func-
enter f(2)
tion code is the static part whereas execution is the exit f(2)
counterpart. The storage associated with an activation enter f(1)
of a procedure is called activation record. exit f(1)
1. Activation record content: exit f(3)
exit f(5)
••Temporary values: Values generated as a result
of the expression evaluations which cannot be The activation tree for f (5) is as follows:
put in registers.
••Local data: Local data to belong to the procedure.
main ( )
••Saved machine: Keeps the context or machine
f(5)
status (register, PC, etc.).
••Access link: Points to non-local data in other AR.
••Control link: Points to the caller’s activation f(4) f(3)
record the return value space of the called func-
tion, if any.
••Actual parameters: Used by the calling proce- f(3) f(2) f(2) f(1)
dure to pass parameters to called procedures;
registers are used to pass these information. f(1) f(1)
For a recursive procedure or function, several activations Activation tree of f (5).
may be alive simultaneously. Activation tree shows the
path through which control enters and leaves activa- The following observations are made based upon the
tions for single run of a program. Each node represents activation tree:
an activation of a function, if an arrow is facing a child
node from the parent node that means the child func- 1. Order of activation corresponds to the pre-order
tion is called by the parent function. Sibling à’ is left traversal of the tree.
to `b’ means that function à’ is called before function b. 2. Order of deactivation corresponds to the post-order
traversal of the tree.
3. If an activation of p calls q, then p will not termi-
nate before q.
Problem 6.23: Draw an activation tree of the given
example.
6.6.3 Procedure Call Return Model
int a[10];
int main(){ Every machine’s architecture and every language are
int i; slightly different from each other. The basic steps fol-
for (i=0; i<10; i++){ lowed by a function call are as follows:
a[i] = f(i);
} 1. Before a function call, the calling routine:
} ••Saves any necessary registers
int f (int n) { ••Pushes the arguments onto the stack for the tar-
if (n<3) return 1; get call
return f(n-1)+f(n-2); ••Sets up the static link (if appropriate)
main ends ••Pushes the return address onto the stack
} ••Jumps to the target
Chapter 6.indd 304 4/9/2015 9:59:48 AM

6.7 INTERMEDIATE CODE GENERATION 305
2. During a function call, the target routine: Symbol table can be implemented by:
••Saves any necessary registers
1. Ordered list
••Sets up the new frame pointer
2. Unordered list
••Makes space for any local variables
3. Hash table
••Does its work
4. Tree
••Tears down frame pointer and static link
••Restores any saved registers Possible entries in a symbol table are as follows:
••Jumps to saved return address 1. Name
3. After a function call, the calling routine: 2. Data type
••Removes return address and parameters from 3. Size
the stack 4. ID
••Restores any saved registers 5. Scope information
••Continues executing 6. Storage allocation
6.6.4 Lexical Versus Dynamic Scoping

6.7 INTERMEDIATE CODE
Lexical (or Static Dynamic Scoping GENERATION
Scoping)
Binding of variable to Binding of variable to
The source program can be directly converted into
declarations is done at declarations is done at
the target language. But there are many benefits of
compile time. compile time.
having an intermediate code or machine-independent
Lexical scope rules Dynamic scoping can be code such as increased abstraction, clear separation
specify the association achieved by copying a between front end and back end, and easily retarget
of variables with function verbatim at the and code optimization technique can also be applied
declaration based on place of call. (Fig. 6.14).
the textual order of the
source code.
Innermost enclosing Most recent occurrence Front end
block rule. rule.
Scanner
Intermediate
Parser
Name resolution is Name resolution depends

independent of caller. on caller. Static Intermediate code Back
checker code
generator end
Used by most modern Formerly used in some
languages: C/C++, interpreted languages
Pascal, etc. (older versions of LISP).
Figure 6.14 | Intermediate code generation.
6.6.5 Symbol Table
It is a data structure used by compiler to store all the 6.7.1 Intermediate Representations
information of tokens generated by lexical analyzer. After
the lexical analysis phase, semantic analyzer performs The commonly used representations of intermediate
type checking on the input code. During type check- code are as follows:
ing, semantic analyzer checks whether the use of names 1. Syntax tree
(such as type names, variables, functions) is consistent 2. Postfix notation
with their definition in the source code. Consequently, 3. Three-address code
if there are any inconsistencies or misuses found during
type checking then it is shown as an error. This is the
task of a symbol table. 6.7.1.1 Syntax Tree
Operations that can be performed on a symbol table A syntax tree is the tree representation of the abstract
are as follows: syntactic structure of source code. Due to its abstract
1. Insert nature it is also called as abstract syntax tree. Each of
2. Delete the tree node denotes a construct occurring in source
3. Search code.
Chapter 6.indd 305 4/9/2015 9:59:48 AM

Example 6.8 1. Quadruples: A quadruple is a record structure

having four fields, namely op, arg1, arg2 and result.
Statement written in source language: The op field contains an internal code for opera-
a = b*c + d *c tor. To translate binary expression `x operator y’
into three-address code, operator is placed in òp’
column; x in àrg1’ column, y in àrg2’ column and
= a new temporary variable stores their calculated
result in the result column.
a +
Example 6.10
The quadruples for the assignment a = b*c + d *e gener-
ates the following three-address code:
* *
Op arg1 arg2 Result
(0) = C t1
b c d c (1) * B t1 t2
(2) = D t3
Syntax tree of a = b*c + d*c. (3) * E t3 t4
(4) + t2 t4 t5
(5) = t5 A
6.7.1.2 Postfix Notation
Postfix notation is also known as `reverse polish’ nota- 2. Triples: Temporary variables or memory names
tion. Any expression can be written unambiguously. in the symbol table can be avoided by the position
Interpreters can be built easily for postfix notation by number of the statement that computes it. If we
using the stack data structure. In the postfix notation, used this method then three address statements can
an operator follows the operand. be represented by records with three columns oper-
ation, arg1 and arg2. The column arg1 and arg2,
for the arguments of operation, are either point-
Example 6.9 ers to the symbol table or pointers into the triple
structure. As numbers of used fields are three, so
The source language statement: this format is known as triples.
a = b*c + d*c
Can be rewritten in postfix notation as: Example 6.11
abc*dc* + =
Operation Arg1 Arg2
(0) = c
(1) * b (0)
6.7.1.3 Three-Address Statement
(2) = c
Three-address statement is a linearized representation (3) * b (2)
of an abstract syntax, in which names of the temporar-
(4) + (1) (3)
ies correspond to the nodes. The intermediate values
name allows three-address code to be easily rearranged, (5) = a (4)
which is convenient for optimization technique. The
reason to call `three-address code’ is that each state-
ment generally contains three addresses, two for the 3. Indirect triples: Triples are very difficult to opti-
operands and one for the result. Representations of the mize because for optimization it required moving
three-address statement are quadruples, triples and of intermediate code and other triples connected
indirect triples. to it also have to be updated. So, a new variant of
Chapter 6.indd 306 4/9/2015 9:59:49 AM

6.8 CODE OPTIMIZATION 307
triples is called indirect triples is being used which 6.8.2 Primary Source of Optimization
is easier to optimized. Indirect triples perform list-
ing pointers to triples, rather than listing the tri- Optimization can be done by using different ways.
ples themselves. Some of the primary source of optimization is discussed
below:
Example 6.12
6.8.2.1 Dead-Code Elimination
# Stmt # Op Arg1 Arg2
Dead-code is a section of program code which is executed
(0) (14) → (14) C
but the result produced by that section is never used.
→
(1) (15) (15) * B (14) So, it can be removed from the program code because it
does not have any effect on the functionality of program.
(2) (16) → (16) C
Let we have temp1 = temp2 − temp3, and temp1 is
→
(3) (17) (17) * B (16) never used further in the program then we can eliminate
this whole instruction.
(4) (18) → (18) + (15) (17)
(5) (19) → (19) = A (18) 6.8.2.2 Constant Propagation
Program Triple container Constant propagation is the process of substituting a
constant value by the subsequent uses of a variable while
there is no intervening changes in the value of that vari-
able. Consider the given example below:
6.8 CODE OPTIMIZATION
int a = 10;
int b = 12 – a / 2;
return b = b * (35/a +2);
Code optimization phase is to reduce the size of the
code and improve the performance of the code gener- Variable à' propagating the constant value 10 in the
ated by intermediate code phase. The most important given code.
part of optimized code is to minimize the amount of
time taken by the code to execute and less common is 6.8.2.3 Constant Folding
to minimize the amount of memory used by the code.
The basic requirement optimization methods should Constant folding is the process of evaluating the
comply with is that an optimized program must have constant expression at compile time. Constant fold-
the same output and side effects as its non-optimized ing process improves run-time performance and also
version. reduces code size by evaluating constant at compile
time. In program code below, the expression (4−2) can
be evaluated at compile time and replaced with the
6.8.1 Types and Levels of Optimization constant 2.
Optimization can be performed by automatic optimizers int sub( )

{
or programmers. An optimizer is a built-in unit of a com-
Return (4-2);
piler. Modern processors have the capability to optimize }
the execution order of code instructions. Optimizations
are classified mainly into two types, one is high-level
and the other is low-level optimizations. High-level opti- 6.8.2.4 Elimination of Common
mizations are generally performed by the programmer Sub-Expression
who handles source code of the programs such as classes,
functions, control statements, and procedures. Low-level If two or more operations are similar and produce same
optimizations are performed at the stage when source results then it is efficient way that to compute at once
code is compiled into a set of machine instructions, and and refer that results further rather than re-evaluate
it is at this stage that automated optimization is usually it. In the given program code below, temp 3 and temp
employed. 6 produce similar results and temp 4 and temp 5 also
Chapter 6.indd 307 4/9/2015 9:59:49 AM

produce similar results, so we can compute them once tmp2 = -x;

and refer results further below in the code. x = 21 * tmp2;
tmp3 = x * x;
main() tmp4 = x/y;
{ y = tmp3 + tmp4;
int x, y, z; tmp5 = x/y;
z = tmp5/tmp3;
x = (1+20)* -x; y = z;
y = x*x+(x/y);
y = z = (x/y)/(x*x);
} 6.8.2.5 Copy Propagation
straight translation: Copy propagation is the different way of optimization,

in which assignment a=b for some variable à’ and `b’,
tmp1 = 1 + 20; we can replace later uses of à’ with use of `b’ (it is to
tmp2 = -x; be assumed that there is no change to either variable in-
x = tmp1 * tmp2; between). The code on the left side makes a copy of tmp1
tmp3 = x * x; in tmp2 and a copy of tmp3 in tmp4. When we do optimi-
tmp4 = x/y; zation then on the right, we eliminated those unnecessary
y = tmp3 + tmp4; copies and propagated the original variable into later uses.
tmp5 = x/y;
tmp6 = x * x;
tmp2 = tmp1;
z = tmp5/tmp6;
tmp3 = tmp2 * tmp1;
y = z;
tmp4 = tmp3;
tmp5 = tmp3 * tmp2;
What sub-expressions can be eliminated? How can valid c = tmp5 + tmp4;
common sub-expressions (live ones) be determined? Here tmp3 = tmp1 * tmp1;
is an optimized version, after constant folding and prop- tmp5 = tmp3 * tmp1;
agation and elimination of common sub-expressions: c = tmp5 + tmp3;
IMPORTANT FORMULAS
1. Time complexity of RDP is O(2n). 7. Every LL(1) grammar will surely be LALR(1), but
if any grammar is LALR(1) then it may or may not
2. Every regular grammar need not be LL(1), because
be LL(1).
that grammar may contain left factoring.
8. If any grammar is LALR(1), then it will surely be
3. Any ambiguous grammar cannot be LL(1).
CLR(1).
4. If any grammar contains left factoring, then it
9. CLR(1) is also known as LR(1).
can’t be LL(1) grammar.
10. Every LL(1) grammar will surely be CLR(1).
5. If given grammar contains left recursion, then it
can’t be LL(1) grammar. 11. All LL(k) parsers are subset of LR(k) parser.
6. If grammar G is LL(1), then 12. If the number of states in LR(0), SLR(1), CLR(1)
•• It may be LR(0). and LALR(1) is n1, n2, n3, n4, respectively, then
•• It may be SLR(1). relation between them is
•• It will surely be LALR(1). n 1 = n2 = n4 ≤ n3
Chapter 6.indd 308 4/9/2015 9:59:50 AM

SOLVED EXAMPLES 309
SOLVED EXAMPLES
1. Which of the following derivations does a top-down Solution: The tokens in C tokens include identi-
parser use while parsing an input string? The input fiers, keywords, constants, operators, string literals
is assumed to be scanned in left to right order. and other separators. Thus, there are 10 tokens in
the given printf statement.
(a) Leftmost derivation
Ans. (c)
(b) Topmost derivation
(c) Rightmost derivation 6. The identifications of common sub-expression and
(d) Leftmost derivation traced out in reverse replacement of run-time computations by compile-
time computations is
Solution: Top—Down parser uses leftmost deriva-
tion for constructing parse tree. (a) local optimization
Ans. (a) (b) global optimization
(c) constant folding
2. The process of assigning load addresses to vari-
(d) common fielding
ous parts of the program and adjusting the code
and data in the program to reflect the assigned Solution: Constant folding is the process of evalu-
addresses is called ating the constant expression at compile time.
Ans. (c)
(a) address assembly (b) parsing
(c) relocation (d) indexing 7. In a compiler the module that checks every charac-
ter of the source text is called
Solution: The process adjusting the code and data
in the program to reflect the assigned addresses is (a) the machine dependant optimizer
called relocation. (b) the code checker
Ans. (c) (c) the lexical analyzer
(d) the syntax analyzer
3. Which of the following statements is true?
Solution: Lexical analyzer reads the source program
(a) An unambiguous grammar need not have same
character by character at a time and unites them
leftmost and rightmost derivation.
into a stream of tokens.
(b) An LL(1) parser is not a top-down parser.
Ans. (c)
(c) LALR is less powerful than SLR.
(d) An ambiguous grammar can be LR(k) for any k. 8. Match the following.
Solution: Only option (a) is true. LL(1) parser is Group 1 Group 2
a top—down parser; LALR is more powerful.
A. Lexical analysis P. Directed acyclic graphs
Ans. (a)
B. Code optimization Q. Syntax trees
4. Given the following expression grammar: C. Code generation R. PDA
E → E *F |F + E |F
D. Abelian groups S. Finite automaton
F → F − F |id (a) A-S, B-P, C-R, D-Q

(b) A-Q, B-R, C-S, D-P
Which of the following is true? (c) A-S, B-Q, C-R, D-P
(a) * has higher precedence than + (d) A-P, B-S, C-R, D-Q
(b) − has higher precedence than * Solution: A-S, B-P, C-R, D-Q
(c) + and − have same precedence Ans. (a)
(d) + has higher precedence than *
9. Which of the following strings can definitely be
Solution: id has higher precedence than any other said to be token without looking at the next input
symbol. character while compiling a Pascal program?
Ans. (b)
1. Begin
5. The number of tokens in the following C statement is 2. Goto
printf(“i=%d, &i =%x”,i&i); 3. <>
(a) 13 (b) 6 (a) 1 (b) 2
(c) 10 (d) 11 (c) 3 (d) All of the above
Chapter 6.indd 309 4/9/2015 9:59:50 AM

Solution: Begin and Goto cannot be properly (a) i only (b) i and ii
grouped as keywords without considering the (c) i and iii (d) i and iv
delimiter. But <> can be considered as Boolean
Solution: Operator grammar has two properties:
operator, not matter what follows the symbol.
First, there should not be any null production. Second,
Ans. (c)
there should not be any production which has two con-
10. Let us suppose, in a programming language an secutive variables in the right-hand side of production.
identifier is permitted to be a letter followed Ans. (d)
by any number of letters or digits. If L and D 13. Consider the following grammar:
denote the sets of letters and digits, respec-
S → (L)/a
tively, which of the following expressions defines
L → SL′
an identifier?
L′ → e/+SL′
(a) (L.D)* (b) L(L + D) In the predictive parser table M of the grammar,
(c) L + (L.D)* (d) L(L.D)* the entries M[S, ( ] and M[L′,)], respectively,
Solution: An identifier expression cannot have (a) S → (L) and L′ →,SL′
character such as *. (b) S → a and L′ → e
Ans. (b) (c) S → (L) and L′ → e
(d) S → a and L′ →,SL′
11. Consider LR(1) collections of grammar (G), in
state I 0 (i.e. first state) what is the follow set of Solution:
LR(1) item, T → .TF? Parsing table
(a) $| + (b) $| + |a|b A , ( ) $
(c) $| + |a (d) + |a|b
S S→a S → (L)
Solution: L L → SL′ L → SL′
I0: L′ L′ →,SL′ L′ → e
E′→ .E, $
E → .E + T, $| +
Ans. (c)
E → .T, $| + 14. Which is equivalent CFG without useless symbol
T → .TF, $| + |a|b for the given grammar:
T → .F, $| + |a|b S → PQ|p, P → p
F → .F *, $| + |a|b|*
F → .a, $| + |a|b|* (a) S → PQ (b) S → p
F → .b, $| + |a|b|* (c) S → P, P → p (d) S → P|p, P → p
Ans. (b) Solution: P and Q are the useless symbol because
12. Which grammar rules violate the requirement of Q does not produce anything and P produces p
the operator grammar? A, B, C are variables and terminal through P → p, which can directly be
a, b, c are terminals. generated by the starting symbol. So, option (b)
is correct, because the given grammar generates
i. A → BC ii. A → CcBb terminal p only, which can be generated by S → p.
iii. A → BaC iv. A → e Ans. (b)
1. Which of the following suffices to convert an arbi- 2. Assume that the SLR parser for a grammar G has
trary CFG to an LL(1) grammar? n1 states and the LALR parser for G has n2 states.
The relationship between nl and n2 is
(a) Removing left recursion alone
(b) Factoring the grammar alone (a) n1 is necessarily less than n2.
(c) Removing left recursion and factoring the grammar (b) n1 is necessarily equal to n2.
(d) None of the above (c) n1 is necessarily greater than n2.
(GATE 2003: 1 Mark) (d) None of the above.
(GATE 2003: 1 Mark)
Solution: To convert a CFG to an LL(1) grammar,
left recursion and left factoring must be removed. Solution: States for SLR parser and LALR parser
Otherwise, the grammar becomes ambiguous. are equal.
Ans. (c) Ans. (b)
Chapter 6.indd 310 4/9/2015 9:59:51 AM

3. In a bottom-up evaluation of a syntax-directed 6. Consider the grammar shown below:

S → CC
definition, inherited attributes can
(a) always be evaluated. C → cC |d
(b) be evaluated only if the definition is L attributed.
This grammar is
(c) be evaluated only if the definition has synthe-
sized attributes. (a) LL(1)
(d) never be evaluated. (b) SLR(1) but not LL(1)
(GATE 2003: 1 Mark) (c) LALR(1) but not SLR(1)
(d) LR(l) but not LALR(1)
Solution: In a bottom-up evaluation of a syntax- (GATE 2003: 2 Marks)
directed definition, inherited attributes can be
evaluated using synthesized attributes. Solution: The given grammar is LR(1). Every
Ans. (c) SLR(1) grammar is LR(1) but not every LALR(1)
grammar is LR(1).
4. Which of the following statements is FALSE? Ans. (d)
(a) In statically typed languages, each variable in a 7. Consider the translation scheme shown below:
program has a fixed type.
(b) In untyped languages, values do not have any S → TR
types. R → +T {print(`+’);} R|e
(c) In dynamically typed languages, variables have T → num {print(num.val);}
no types. Here num is a token that represents an integer
(d) In all statically typed languages, each variable and num.val represents the corresponding integer
in a program is associated with values of only a value. For an input string `9 + 5 + 2’, this transla-
single type during the execution of the program tion scheme will print
(GATE 2003: 1 Mark) (a) 9 + 5 + 2 (b) 9 5 + 2 +
(c) 9 5 2 + + (d) + + 9 5 2
Solution: In dynamically typed languages such as
C, variables have no types. (GATE 2003: 2 Marks)
Ans: (c) Solution: Tree for translation scheme
5. Consider the grammar shown below: S
S → iEtSS ′|a
S′ → eS |e T
E→b
R
In the predictive parse table, M, of this grammar, 9 {Print 9} + {Print+} R

the entries M[S ′,e] and M[S ′,$], respectively, are
T
R
(a) {S ′ → eS } and {S ′e} 9 {Print 5} + T {Print+}
(b) {S ′ → eS } and { } ε
(c) {S ′ → e} and {S ′ → e} 2 {Print 2}
(d) {S ′ → eS, S ′ → e} and {S ′ → e}
Thus, the translational scheme will print 9 5 + 2 +
Solution: Predictive parser table is 8. Consider the syntax-directed definition shown
below.
Non-
Terminal A b E i t $ S → id: = E {gen(id.place = E.place;);}
S S→a S → iEiSS ′
E → E1 + E2 {t = newtemp( );
S ′ → e
S ′ S ′ → e
S ′ → eS gen(t = E1.place + E2.place;);
E E→b
M[S ′, e] = {S ′ → e, S ′ → eS } E.place = t;}
M[S ′, $] = S ′ → e
Ans. (d) E → id {E.place = id.place;}
Chapter 6.indd 311 4/9/2015 9:59:51 AM

Here, gen is a function that generates the output operation requires the first operand to be in the
code, and newtemp is a function that returns the register. If E1 and E2 do not have any common sub-
name of a new temporary variable on every call. expression, in order to get the shortest possible code
Assume that ti’s are the temporary variable names
(a) E1 should be evaluated first.
generated by newtemp. For the statement `X: =
(b) E2 should be evaluated first.
Y + Z’, the 3-address code sequence generated by
(c) evaluation of E1 and E2 should necessarily be
this definition is
interleaved.
(a) X = Y + Z (d) order to evaluation of E1 and E2 is of no
(b) t1 = Y + Z; X = t1 consequence.
(c) t1 = Y; t2 = t1 + Z; X = t2 (GATE 2004: 1 Mark)
(d) t1 = Y; t2 = Z; t3 = t1 + t2; X = t3
Solution: If we evaluate E2 first then we can save
(GATE 2003: 2 Marks) this to memory location and evaluation of E1 is
kept in registers only. But if E1 is solved first, it
Solution: Three-address code for X = Y + Z is should be saved to memory. Then evaluate E2 and
t1 = Y save to memory. Both need to be fetched again. So,
t2 = Z option (b) is the optimized answer.
t3 = t1 + t2 Ans. (b)
X = t3 12. Consider the grammar with the following transla-
Ans. (d) tion rules and E as the start symbol.
9. Consider a program P that consists of two source E → E1 #T {E.value = E1.value * T.value}
modules M1 and M2 contained in two different files. |T {E.value = T.value}
If M1 contains a reference to a function defined in T → T1 & F {T.value = T1.value + F.value}
M2, the reference will be resolved at |F {T.value = F.value}
F → num {F.value = num.value}
(a) edit time. (b) compile time.
(c) link time. (d) load time. Compute E.value for the root of the parse tree for
the expression: 2 # 3 & 5 # 6 & 4.
(GATE 2004: 1 Mark)
(a) 200 (b) 180
Solution: Modules can be compiled separately, but (c) 160 (d) 40
they are linked to resolve references at link time. (GATE 2004: 2 Marks)
Ans. (c)
Solution: The parse tree is
10. Which of the following grammar rules violate the E = 160
requirements of an operator grammar? P, Q, R are
non-terminals, and r, s, t are terminals. # T1 = 10
E1 = 16
(i) P → QR
(ii) P → QsR E2 = 2 T2 = 8 T5 = 6 & F = 4
(iii) P → e #
(iv) P → QtRr T3 = 2 T4 = 3
&
T2 = 5 F=6
(a) (i) only (b) (i) and (iii) only

(c) (ii) and (iii) only (d) (iii) and (iv) only F=2 F=3
(GATE 2004: 1 Mark) num = 2 num = 3 num = 5 num = 6 num = 4

Solution: The rule for operator grammar is that Ans. (c)
no production drive e and it should not contain two
non-terminals. 13. The grammar A → AA|(A)|e is not suitable for
predictive parsing because the grammar is
P → QR and P → e violate the rules.
Ans. (b) (a) ambiguous.
(b) left recursive.
11. Consider the grammar rule E − E1 — E2 for arithme- (c) right recursive.
tic expressions. The code generated is targeted to a (d) an operator grammar.
CPU having a single-user register. The subtraction (GATE 2005: 1 Mark)
Chapter 6.indd 312 4/9/2015 9:59:52 AM

Solution: The given grammar is ambiguous because Solution:

it have more than one parse tree for sting “( )”. E → number |E `+’ E |E `×’ E
Ans. (a)
yacc compiler detects shift-reduce conflict, and resolves
14. Consider the grammar: the conflict in favour of a shift over a reduce action.
E → E + n|E × n|n
Ans. (c)
For a sentence n + n × n, the handles in the right-

17. Assume the conflicts in Part (a) of this question
are resolved and an LALR(1) parser is generated
sentential form of the reduction are
for parsing arithmetic expressions as per the given
(a) n, E + n and E + n × n grammar. Consider an expression 3 × 2 + 1. What
(b) n, E + n and E + E × n precedence and associativity properties does the
(c) n, n + n and n + n × n generated parser realize?
(d) n, E + n and E × n
(a) E qual precedence and left associativity; expres-
sion is evaluated to 7.
Solution: (b) Equal precedence and right associativity;
n+n×n expression is evaluated to 9.
Using E → n E+n×n (c) Precedence of `×’ is higher than that of `+’, and
Using E → E + n E×n both operators are left associative; expression is
Using E → E × n E evaluated to 7.
Handles are n, E + n, E × n. (d) Precedence of `+’ is higher than that of `×’, and
Ans. (d) both operators are left associative; expression is
evaluated to 9.
15. Consider the grammar:
S → (S)|a
Solution: X has higher precedence than +. So 3 ×
Let the number of states in SLR(1), LR(1) and 2 will be solved first. Hence, 3 × 2 = 6 + 1 =7.
LALR(1) parsers for the grammar be n1, n2 and n3, Ans. (c)
respectively. The following relationship holds good:
18. Consider the following grammar:
(a) n1 < n2 < n3 (b) n1 = n3 < n2
S → S*E
(d) n1 ≥ n3 ≥ n2
S→E
(c) n1 = n2 = n3
E→F+E
Solution: Number of states: SLR = LALR < LR. E→F
Ans. (b) F → id
Linked Answer Questions 16 and 17: Consider the Consider the following LR(0) items corresponding
following expression grammar. The semantic rules to the grammar above.
for expression evaluation are stated next to each (i) S → S*.E
grammar production. (ii) S → F. + E
E → number E.val = number.val (iii) S → F + .E
|E `+’ E E(1).val = E(2).val + E(3).val Given the items above, which two of them will
|E `×’ E E(1).val = E(2).val × E(3).val appear in the same set in the canonical sets of
16. The above grammar and the semantic rules are fed items for the grammar?
to a yacc tool (which is an LALR(1) parser genera- (a) (i) and (ii) (b) (ii) and (iii)
tor) for parsing and evaluating arithmetic expres- (c) (i) and (iii) (d) None of the above
sions. Which one of the following is true about the (GATE 2006: 1 Mark)
action of yacc for the given grammar?
Solution: If S → S*.E is in LR(0) then E →
(a) It detects recursion and eliminates recursion. F +.E will also be there because both of them
(b) It detects reduce-reduce conflict, and resolves. have `.’ (dot) before E.
(c) It detects shift-reduce conflict, and resolves the Ans. (c)
conflict in favour of a shift over a reduce action.
(d) It detects shift-reduce conflict, and resolves the 19. Consider the following translation scheme:
conflict in favour of a reduce over a shift action. S → ER
(GATE 2005: 2 Marks) R → *E{print(`*’);}R|e
Chapter 6.indd 313 4/9/2015 9:59:52 AM

E → F + E{print(+);}|F 21. Which one of the following is a top-down parser?

F → (S)id{print(id.value);} (a) Recursive descent parser
Here id is a token that represents an integer and (b) Operator precedence parser
id.value represents the corresponding integer value. (c) An LR(k) parser
For an input 2*3 + 4, this translation scheme prints (d) An LALR(k) parser
(GATE 2007: 1 Mark)
(a) 2*3 + 4 (b) 2* + 34
(c) 23*4 (d) 234 + * Solution: Recursive decent parsing is top-down
(GATE 2006: 2 Marks) parsing.
Ans. (a)
Solution:
22. Consider the grammar with non-terminals N = {S,
Parse tree is:
C, S1}, terminals T = {a, b, i, t, e}, with S as the
S
start symbol, and the following set of rules:
E R S → iCtSS1|a
S1 → eS|e
R
∗ C→b
Print ∗
F E ε The grammar is NOT LL(1) because
(a) it is left recursive. (b) it is right recursive.
id Print F + E Print + (c) it is ambiguous. (d) it is not context-free.
id value (GATE 2007: 2 Marks)
2
F Solution: Two parse trees are possible for the above
id
Print grammar. So, the given grammar is ambiguous.
3 Ans. (c)
id Print
234+∗ 4
23. Consider the following two statements:
Ans. (d)
A. Every regular grammar is LL(1).
20. Consider the following grammar: B. Every regular set has an LR(1) grammar.
S → FR Which of the following is TRUE?
R → *S |e (a) Both A and B are true.
F → id (b) A is true and B is false.
In the predictive parser table, M, of the grammar, (c) A is false and B is true.
the entries M[S, id] and M[R, $], respectively, (d) Both A and B are false.
(a) {S → FR} and {R → S}
(b) {S → FR} and { } Solution: A left recursive grammar is not LL(1).
(c) {S → FR} and {R → *S} Ans. (c)
(d) {F → id} and {R → e}
(GATE 2006: 2 Marks) 24. Which of the following describes a handle (as appli-
cable to LR parsing) appropriately?
Solution:
Parsing table (a) It is the position in a sentential form where the
next shift or reduce operation will occur.
Non-Terminal id * $ (b) It is non-terminal whose production will be
used for reduction in the next step.
S S → FR
(c) It is a production that may be used for reduc-
R R → *S R → e tion in a future step along with a position in the
F F → id sentential form where the next shift or reduce
operation will occur.
From this table, we find option (a) is correct. (d) It is the production p that will be used for
Ans. (a) reduction in the next step along with a position
Chapter 6.indd 314 4/9/2015 9:59:53 AM

in the sentential form where the right-hand implemented with a stack-based storage alloca-
side of the production may be found. tion scheme for activation records.
(GATE 2008: 1 Mark) (a) II and V only (b) I, III and IV only
(c) I, II and V only (d) II, III and V only
Solution: Fact (GATE 2008: 2 Marks)
Ans. (d)
Solution:
25. Some code optimizations are carried out on the
intermediate code because I: Multi-level access link (or display) arrangement
is needed to arrange activation records only if the
(a) They enhance the portability of the compiler to programming language being implemented has
other target processors. nesting of procedures/functions is true.
(b) Program analysis is more accurate on interme- V: Programming languages which permit a func-
diate code than on machine code. tion to return a function as its result cannot be
(c) The information from dataflow analysis cannot implemented with a stack-based storage allocation
otherwise be used for optimization. scheme for activation records is true.
(d) The information from the front end cannot oth- Ans. (a)
erwise be used for optimization.
28. Match all items in Group 1 with correct options
(GATE 2008: 1 Mark) from those given in Group 2.
Solution: Fact Group 1 Group 2

Ans. (b) P. Regular expression 1. Syntax analysis
26. An LALR(1) parser for a grammar G can have Q. Pushdown automata 2. Code generation
shift-reduce (S-R) conflicts if and only if R. Dataflow analysis 3. Lexical analysis
S. Register allocation 4. Code optimization
(a) The SLR(1) parser for G has S-R conflicts.
(b) The LR(1) parser for G has S-R conflicts. (a) P-4, Q-1, R-2, S-3 (b) P-3, Q-1, R-4, S-2
(c) The LR(0) parser for G has S-R conflicts. (c) P-3, Q-4, R-1, S-2 (d) P-2, Q-1, R-4, S-3
(d) The LALR(1) parser for G has reduce-reduce (GATE 2009: 1 Mark)
conflicts.
Solution: Regular expressions: Lexical analysis
Pushdown automata: Syntax analysis
Solution: The LR(1) parser for G has Shift-Reduce Dataflow analysis: Code optimization
conflicts. Register allocation: Code optimization
Ans. (b) Ans. (b)
27. Which of the following are true? 29. Which of the following statements are TRUE?
I. A programming language which does not A. There exist parsing algorithms for some pro-
permit global variables of any kind and has no gramming languages whose complexities are less
nesting of procedures/functions, but permits than Q(n3).
recursion can be implemented with static stor- B. A programming language which allows recur-
age allocation. sion can be implemented with static storage
II. Multi-level access link (or display) arrangement allocation.
is needed to arrange activation records only if C. No L-attributed definition can be evaluated in
the programming language being implemented the framework of bottom-up parsing.
has nesting of procedures/functions. D. Code improving transformations can be per-
III. R ecursion in programming languages cannot be formed at both source language and intermedi-
implemented with dynamic storage allocation. ate code level.
IV. Nesting procedures/functions and recursion (a) A and B (b) A and D
require a dynamic heap allocation scheme and (c) C and D (d) A, C and D
cannot be implemented with a stack-based (GATE 2009: 2 Marks)
allocation scheme for activation records.
V. Programming languages which permit a func- Solution: Statements A and D are true.
tion to return a function as its result cannot be Ans. (b)
Chapter 6.indd 315 4/9/2015 9:59:53 AM

30. Which data structure in a compiler is used for Program main;

managing information about variables and their Var ...
attributes? Procedure A1;
Var ...
(a) Abstract syntax tree (b) Symbol table Call A2;
(c) Semantic stack (d) Parse table End A1
Procedure A2;
(GATE 2010: 1 Mark) Var ...
Solution: Symbol table is a data structure used by Procedure A21;
a compiler to store variables and their attributes. Var ...
Ans. (b) Call A1;
31. The grammar S → aSa|bS|c is End A21
(a) LL(1) but not LR(1) Call A21;

(b) LR(1) but not LR(1) End A2
(c) Both LL(1) and LR(1)
(d) Neither LL(1) nor LR(1) Call A1;
End main.
Consider the calling chain: Main → A1 →A2 →
Solution: The given grammar is regular and A21 → A1
LR(1). It is also LL(1) as multiple entries for one The correct set of activation records along with
iteration are present. their access links is given by
Ans. (c)
(a) (b)
32. In a compiler, keywords of a language are recog-
nized during Main Main
(a) parsing of the program. A1 A1

(b) the code generation.
(c) the lexical analysis of the program. A2 A2
(d) dataflow analysis. A21 A21
(GATE 2011: 1 Mark)
Frame A1 Access Frame A1
Access
Solution: Tokens are recognized in lexical analy- pointer links pointer links
sis phase of compiler.
Ans. (c) (c) (d)
33. The lexical analysis for a modern computer lan- Main Main
guage such as Java needs the power of which one
of the following machine models in a necessary and Frame A1 A1
sufficient sense? pointer A2 A2
(a) Finite state automata
A21 A21
(b) Deterministic pushdown automata
Access
(c) Non-deterministic pushdown automata A1
links Frame Access
(d) Turing machine
pointer links
(GATE 2011: 1 Mark)
Solution: Lexical analyzer is implemented using Solution: Access link is defined as link to acti-
finite state automata. vation record of closest lexically enclosing block
Ans. (a) in program text, so the closest enclosing blocks
34. Consider the program given below, in a block- representing for A1, A2 and A21 are main, main
structured pseudo-language with lexical scoping and A2.
and nesting of procedures permitted. Ans. (d)
Chapter 6.indd 316 4/9/2015 9:59:53 AM

Linked Answer Questions 35 and 36: For the (i) E3:B → S

grammar below, a partial LL(1) parsing table is (j) E1:A → S, S → e
also presented along with the grammar. Entries (k) E2:B → S, S → e
that need to be filled are indicated as E1, E2, (l) E3:B → S
and E3. e is the empty string, $ indicates end of (GATE 2012: 2 Marks)
input, and, | separates alternate right-hand sides
of productions. Solution:
S → aAbB|bAaB|e First(S) = {a, b, e} and Follow(S) = {a, b, $}
A→S First(B) = {a, b, e} and Follow(B) = {a, b, $}
B→S So,
a B e
A B $
S → aAbB S → bAaB S→e
S→e
S E1 E2 S
S→e S→e
A A→S A→S Error
B→S B→S B→S
B→S B→S E3 B
B
Ans. (c)
35. The FIRST and FOLLOW set for the non-terminals
A and B are 37. What is the maximum number of reduce moves that
can be taken by a bottom-up parser for a grammar
(a) FIRST(A) = {a, b, e} = FIRST(B) with no epsilon- and unit-production (i.e., of type
(b) FOLLOW(A) = {a, b} A → e and A → a) to parse a string with n tokens?
(c) FOLLOW(B) = {a, b, $}
(d) FIRST(A) = {a, b, $} (a) n/2 (b) n − 1
(e) FIRST(B) = {a, b, e} (c) 2n − 1 (d) 2n
(f) FOLLOW(A) = {a, b} (GATE 2013: 1 Mark)
(g) FOLLOW(B) = {$} Solution: n − 1 moves can be taken by bottom
(h) FIRST(A) = {a, b, e} = FIRST(B) up parser.
(i) FOLLOW(A) = {a, b} Ans. (b)
(j) FOLLOW(B) = ∅
(k) FIRST(A) = {a, b, e} = FIRST(B) 38. Consider the following two sets of LR(1) items of
(l) FOLLOW(A) = {a, b} an LR(1) grammar:
(m) FOLLOW(B) = {a, b}
X → c.X, c/d
(GATE 2012: 2 Marks) X → .cX, c/d
X → .d, c/d
Solution: First(A) gives set of all terminals that X → c.X, $
begin in string derived from A, and Follow(A) X → .cX, $
means it will consist set of all terminals that may X → .d, $
follow immediately to the right of A.
Which of the following statements related to merg-
First(A) = {a, b, e} ing of the two sets in the corresponding LALR
First(B) = {a, b, e} parser is/are FALSE?
Follow(A) = {a, b}
Follow(B) = {a, b, $} A. Cannot be merged since look aheads are
Ans. (a) different.
B. Can be merged but will result in S-R conflict.
36. The appropriate entries for E1, E2 and E3 are C. Can be merged but will result in R-R conflict.
(a) E1:S → aAbB, A → S D. Cannot be merged since goto on c will lead to
(b) E2:S → bAaB, B → S two different sets.
(c) E3:B → S (a) A only (b) B only
(d) E1:S → aAbB, S → e (c) A and D only (d) A, B, C and D
(e) E2:S → bAaB, S → e (GATE 2013: 1 Mark)
(f) E3:S → e
(g) E1:S → aAbB, S → e Solution: All the statements are false.
(h) E2:S → bAaB, S → e Ans. (d)
Chapter 6.indd 317 4/9/2015 9:59:54 AM

PRACTICE EXERCISES
Set 1
1. Type checking is normally done during X→c

Y→c
(a) parsing
(b) syntax analysis The grammar is
(c) syntax-directed translation
(a) LR(1) but not LALR(1)
(d) semantic-based optimization
(b) LALR(1) but not LL(1)
2. Consider the following grammar: (c) SLR(1) but not LR(0)
S → m|mn|mno (d) None of the above
Choose the correct statement from the following: 7. Which of the following statements is true?
(a) The grammar is LL(1). (a) A
n unambiguous grammar does not have the
(b) The grammar is LL(2). same leftmost and rightmost derivation.
(c) The grammar is LL(3). (b) An LL(1) parser is a left-right parser.
(d) None of the above. (c) LALR is more as powerful as SLR.
3. A shift-reduce parser carries out the actions speci- (d) An ambiguous grammar can be defined as
fied within braces immediately after reducing with LR(k) for any k.
the corresponding rule of grammar.
8. Generation of intermediate code based on an
S → xx W {print“1”} abstract machine model is useful in compilers
S → y {print“2”} because
W → Sz {print“3”}
(a) It makes implementation of lexical analysis and
What is the translation of xxxxyzz using the syn- syntax analysis easier.
tax-directed translation scheme described by the (b) Syntax translations are easier for intermediate
above rules? code generation.
(a) 23131 (b) 11233 (c) It enhances the portability of the compiler
(c) 11231 (d) 33211 system program.
(d) It is difficult to generate executable code from
4. The pass numbers for each of the following high-level language programs.
activities
9. A linker is given object modules for a set of pro-
1. Object code generation
grams that were compiled separately. What
2. Literals added to literal table
information needs to be included in an object
3. Listing printed
module?
4. Address resolution of local symbols that occur in
a two-pass assembler, respectively, are (a) Version of the compiler being used
(b) Relocation code
(a) 2, 2, 1, 2 (b) 2, 1, 2, 1
(c) Symbol table and various macros
(c) 2, 1, 1, 2 (d) 1, 2, 2, 2
(d) Absolute addresses of internal symbols
5. Which of the following statements is true?
10. In the following grammar
(a) LALR parser is more powerful than canonical
LR parser. X : : = X Θ Y/Y
(b) LALR parser cannot be compared with LALR Y : : = Z O Y/Z
parser. Z : : = id
(c) Canonical LR parser is more powerful than Which of the following is true?
LALR parser.
(d) The parsers canonical LR and LALR have the (a) `Θ’ is left associative while Ò’ is right
same power. associative.
(b) It cannot be defined as the rules defined are
6. Consider the given grammar: few.
S’ → S (c) `Θ’ is right associative while Ò’ is left associative.
S → aXd |bYd |aYe |bXe (d) Both `Θ’ and Ò’ are right associative.
Chapter 6.indd 318 4/9/2015 9:59:54 AM

11. Which of the following statements are true/false, (c) N

o L-attributed definition can be evaluated in
map them appropriately, with respect to syntax- the framework of bottom-up parsing.
directed definitions? (d) Some undefined symbols are also defined in the
symbol table.
A. The terminals are defined for inherited attributes.
B. Value of attributes of terminals is generally sup- 17. Which of the following describes a handle (as appli-
plied by lexical analyzer. cable to LR parsing) appropriately?
C. The start symbol does not have an inherited
attribute. (a) It is the production p in a sentential form
D. Attribute grammar is used to define the special where the sentence begins and reduce operation
cases of the attributes. occurs.
(b) It is the production p used for reduction in the
(a) TTFT (b) TFTF next step and the next shift operation will occur.
(c) FFTF (d) FTTF (c) It is a production p that may be used for reduc-
tion in a future step along with a position in the
12. Which of the following statements are true?
sentential form where the next shift or reduce
A. All unambiguous grammars are regular. operation will occur.
B. Every LL(K) and LR(K) grammar is (d) It is the production p that will be used for
unambiguous. reduction in the next step along with a position
C. A regular language can never be inherently in the sentential form where the right hand side
ambiguous. of the production may be found.
D. A regular grammar contains special string for
syntax-directed definitions. 18. An LALR(1) parser for a grammar G can have
shift-reduce (S-R) conflicts if and only if
(a) A and D only (b) B and C only
(c) A and B only (d) C and D only (a) The LR(1) parser for G has no S-R conflicts.
(b) The LR(1) parser for G has S-R conflicts.
13. What is the maximum number of reduce moves that (c) The SLR(1) parser for G has S-R conflicts.
can be taken by a bottom-up parser for a grammar (d) The LALR(1) parser for G has reduce-reduce
with no epsilon and unit production (i.e., of type conflicts.
A → e and A → a) to parse a string with n tokens?
19. Which one of the following is a top-down parser?
(a) n + 1/2 (b) n − 1
(a) Recursive descent parser
(c) n (d) n
(b) Shift left associative parser
14. The grammar S → aSa|bS|c can be defined as (c) SLR(k) parser
(d) LR(k) parser
(a) It is only LR(1).
(b) It is only LL(1). 20. Which of the following is true?
(c) Both LL(1)and LR(1).
(a) E very regular grammar is not LL(1) and
(d) Neither LL(1) nor LR(1).
every regular set does not have an LR(1)
15. Match all items in Group 1 with correct options grammar.
from those given in Group 2. (b) Every regular grammar is LL(1) and every reg-
ular set does not have an LR(1) grammar.
Group 1 Group 2 (c) Every regular grammar is not LL(1) and every
P. Regular expression 1. Lexical analysis regular set has an LR(1) grammar.
Q. Register allocation 2. Code generation (d) Every regular grammar is LL(1) and every reg-
R. Dataflow analysis 3. Syntax analysis ular set has an LR(1) grammar.
S. Pushdown automata 4. Code optimization
21. Some functions of code optimizations are carried out
(a) P-1, Q-2, R-3, S-4 (b) P-1, Q-2, R-4, S-3
on the intermediate code of the compiler because
(c) P-3, Q-1, R-2, S-4 (d) P-4, Q-3, R-2, S-1
(a) T
hey enhance the portability of the compiler to
16. Which of the following statements are true? other target computer.
(a) A programming language which allows loops can (b) Program analysis is more accurate and runs
be implemented with static storage allocation. faster.
(b) There exists parsing algorithms for some pro- (c) It saves time in performing dataflow analysis.
gramming languages whose complexities are (d) Provides a better understanding of the object
less than O(n3). file.
Chapter 6.indd 319 4/9/2015 9:59:54 AM

22. Which one of the following is FALSE? 27. The compiler type that converts all operands to
the type of the largest operand is called
(a) A simple basic block enters the sequence at the
beginning and exits at the end. (a) Type promotion (b) Type evaluation
(b) Expression analysis can be used for common (c) Type bootstrap (d) Type operand
sub-expression elimination. testing
(c) Dead-code elimination can be done by live vari-
able analysis. 28. High-level knowledge which relates to the use of
(d) x = 4 ∗ 5 ⇒ x = 20 is an example of common sentences in different contexts and how the context
sub-expression elimination. affect the meaning of the sentences is called
23. One of the purposes of using intermediate code in (a) Pedagogy (b) Syntactic
compilers is to (c) Morphological (d) Pragmatic
(a) Improve memory and stack management. 29. The process adjusting the code and data in the
(b) Improve detection of syntax and semantic program to reflect the assigned addresses is
errors. called ________.
(c) Improve the reusing of the machine-independent (a) Assembly (b) Linking
code optimizer. (c) Absolute addressing (d) Relocation
(d) Improve the register allocation and cache
management. 30. A top-down parser while parsing an input string
from left to right makes use of:
24. Which of the following statement(s) is/are true
about LALR(1) parsers? (a) Leftmost derivation
(b) Topmost derivation
S1: LALR(1) parsers have same number of states
(c) Rightmost derivation traced out in reverse
as SLR(1) parsers (core LR(0) items are the same).
(d) Bottom-up derivation traced in reverse
S2: LALR(1) is derived from LR(1) with no shift-
reduce conflict will also have no shift-reduce 31. Which of the following statement(s) is/are true?
conflict.
S3: LALR(1) may create reduce-reduce conflict S1: If G is unambiguous, then every right-sentential
which was not in LR(1) from which LALR(1) is form has a unique handle.
derived. S2: Shift-reduce parsers use a stack and input
buffer.
(a) None of the statement is true
(b) S1 and S3 (a) Neither S1 nor S2 (b) S1 only
(c) S1 and S2 (c) Both S1 and S2 (d) S2 only
(d) S1, S2 and S3
32. Which of the following is/are type(s) of tools used
25. Which of the following suffices to convert an arbi- in various phases of compiler?
trary CFG to an LL(1) grammar? (a) Control flow graph, binary sort, cyclic graphs
(a) Removing left recursion alone and right asso- (b) Three-address code, bootstrap loader, cyclic
ciative productions graphs
(b) Removing the grammar and the productions (c) Directed acyclic graph, assembler, binary linker
but not the start symbol (d) Control flow graph, three-address code, directed
(c) Factoring the grammar and removing the left acyclic graph
associative productions
(a) If a grammar is LL(k), it is also ~(LR(k − 1)).
26. The major limitation of a shift reduce parser is to (b) If a grammar is not LR(k), it is ~(SLR(k)).
overcome (c) If a grammar is CLR(k), it is also SLR(k).
(a) Representation of productions in top-down (d) If a grammar is not LALR(k), it is not SLR(k).
parser
(b) Reduce-reduce conflict only
(c) Both shift reduce conflict and reduce-reduce E → AB
conflict A → +BA/e
(d) Shift conflicts only B → B * A/id
Chapter 6.indd 320 4/9/2015 9:59:54 AM

Then which of the following is true? (c) S is SLR(1).

(a) Both +, * are left associative. (d) S gives a higher precedence to + over *.
(b) Only * is left associative.
(c) Only + is left associative. 41. What is the maximum number of reduce moves
(d) Both +, * are not left associative. that can be taken by a bottom-up parser for a
grammar with no epsilon and unit production (i.e.
35. The type of conflicts that can rise in LR(0) tech- of type A →? and A → a) to parse a string with n
niques are tokens?
(a) Shift-right only (a) n + 1 (b) n − 1 (c) n (d) 2n
(b) Reduce-reduce only
(c) Shift-left only 42. The following grammar is
(d) Both shift-reduce and reduce-reduce S → Aa|bAc|dc|bda
A→d
36. Shift-reduce parsers are
(a) Not SLR(1) and not CLR(1)
(a) Shift parsers (b) Bottom-up parsers (b) CLR(1) but not SLR(1)
(c) Predictive parsing (d) Reducing parsers (c) SLR(1) but not CLR(1)
(d) SLR(1) and CLR(1)
S → Aa|bAc|Bc|bBa
43. Write true (T)/false (F) for each of the following
A → x, B → x
statements:
A. Symbol table is used only in first three phases
(a) The grammar is LR(1) but not LALR(1)
of a compiler.
(b) LALR(1) but not LL(1)
B. A pretty printer analyses a program and prints
(c) SLR(1) but not LR(0)
it in such a way that the structure of the pro-
gram becomes clearly visible.
38. Which of the following statements is false? C. YACC build up LALR parsing table.
D. If a grammar is LALR(1), it is not necessarily
(a) An unambiguous grammar has same leftmost
SLR(1).
and rightmost derivation.
(b) An LL(1) parser is a top-down parser. (a) TTFT (b) TTTF (c) FTTF (d) FTTT
44. Consider two binary operators * and ∧ with prece-

(c) LALR is more powerful than SLR.
dence of operator * being lower than of ∧. Operator *
(d) An ambiguous grammar can never be LR(k)
is left associative while operator ∧ is right associative.
for any k.
39. Consider the grammar given below: The value of 3*2∧3*4 is –––.
A → AA|a|e
String à aaa’ has
S → aB|aAb
A → bAb|a
(a) A unique derivation tree in the grammar
B → aB|e
(b) Four leftmost derivations in the grammar
(c) Four rightmost derivations in the grammar
(d) An infinite number of leftmost and rightmost How many back tracks are required to generate the
derivations string àab’ from the above grammar?
40. Consider the grammar R: (a) Nil (b) 2 (c) 1 (d) 3
E → E + E |E *E|(E)|id 46. Which of the following statements are TRUE?
Consider the other grammar S: (a) T here does not exist any parsing algorithm
E → E + T |T which can parse in parallel.
T → T *F |F (b) A programming language which does not use
F(E)|id loops makes use of static storage allocation.
(c) No L-attributed definition can be evaluated in
Choose the incorrect statement.
the framework of bottom-up parsing.
(a) S generates the same language as R. (d) Code improving transformations can be per-
(b) R is ambiguous but not LL(1) as it is left formed at both the source language as well as
recursive. the intermediate code level.
Chapter 6.indd 321 4/9/2015 9:59:54 AM

47. Choose the correct one from the following X → yXz|yz

(a) An SLR(1) grammar is always LL(0). Y → wYz|wZz
(b) A grammar that is LR(0) is necessarily LL(1). Z → xZy|xy
(c) A grammar that is LR(0) may not be LL(1). Which language (L) is generated by the grammar?
(d) None of the above.
(a) {wnxnynzn: n ≥ 1}
48. Consider the grammar shown below: (b) {wnxnymzm: n > 1, m > 1}
T → PP (c) {wnxnymzm: n ≥ 1, m ≥ 1} ∪ {wnxmymzn: n ≥ 1,
P → sP/r m ≥ 1}
(d) {wmxnyozp: m, n, o, p > 1}
The grammar is
4. How many parse trees can be generated for the
(a) LL(1) string `wwxxyyzz’ from the grammar given in the
(b) SLR(1) but not LL(1) previous question?
(c) LALR(1) but not SLR(1)
(d) LR(1) but not LALR(1) (a) 4 (b) 3 (c) 2 (d) 1
49. Which statement is false about parsers? 5. The grammar E → EE, E → (E), E → ( ), E → t is
(a) Top-down parsing algorithms use a leftmost (a) Not LL(1) as it is left recursive
derivation. (b) Not LL(1) as it is ambiguous
(b) Bottom-up parsing algorithms use the reverse (c) Not LR(1) as it is left recursive
of a rightmost derivation. (d) None of the above
(c) An LR(1) parser may sometimes exist for some 6. Which one of the following string can definitely be said
ambiguous grammars. to be a token without looking at the next input char?
(d) An LL(1) parser is a top-down parser.
(a) < (b) > (c) + (d) *
7. Consider the grammar:
L → eLsL|sLeL|e
E → TE ′
Choose the correct statement: E ′ → e/+TE ′
(a) The string èses’ has two parse trees T → FT ′
(b) The grammar is unambiguous. T ′ → e/*FT ′
(c) The grammar is LL(1) and LR(1). F → a/[E]
(d) None of the above. 7a. Find out the First( ) of variable T ?
Set 2 (a) a, [ (b) a, [, ]

(c) +, e (d) *, e
1. The number of tokens in the following C statements: 7b. Find out First( ) of variable E ′?
printf (“a = %d, b = %d”, a, b); (a) a, [ (b) a, [, ]
(a) 8 (b) 9 (c) 10 (d) 11 (c) +, e (d) *, e
2. Consider the grammar with the following transla- 7c. What will be the Follow( ) of E?
tion rules and S as the start symbol.
(a) +, * (b) a, [
S → S @ T {S.value = S.value * T.value} (c) $, ] (d) $, [
|T {S.value = T.value}
T → T $ F {T.value = T1.value + F.value}
7d. What will be the Follow( ) of T ?
|F {T.value = F.value} (a) +, *, a (b) +, a, [
F → num {F.value = num.value} (c) *, $, ] (d) +, $, ]
Compute S.value for the root of the parse tree for 8. Which of the following statements is incorrect?
the expression: 2 @ 8 $ 5 @ 6 $ 4.
(a) LALR is more powerful than SLR.
(a) 50 (b) 180 (c) 260 (d) 340 (b) An LL(1) parser uses top-down parsing approach.
(c) An unambiguous grammar has the same left-
most and rightmost derivation.
S → WX|Y (d) An ambiguous grammar can never be LR(k)
W → wWx|wx for any k.
Chapter 6.indd 322 4/9/2015 9:59:55 AM

9. Assume that the LALR parser for a grammar G1 Which of the following strings is generated by the
has n1 states and the CLR parser for G2 has n2 grammar?
states. The relationship between n1 and n2 is
(a) aabbab (b) abbbab
(a) n1 > n2 (b) n1 ≥ n2 (c) bbaaa (d) aabbb
(c) n1 < n2 (d) n1 ≤ n2
14. For the correct answer string to the above question, how
10. Which of the following grammar (G) is an operator many derivation trees are possible?
grammar? (a) 1 (b) 2
(a) {S → SS|Sa|Sb|a} (c) 3 (d) 4
(b) {S → AB|e, A → a, B → b}
(c) {S → aA, A → a|b|e}
15. To manage and store information about variables
(d) {S → AaS|b, A → a}
and their attributes, compiler uses
(a) Parse table (b) Semantic stack
11. Consider the following grammar (G): (c) Symbol table (d) Tree
E → E + T |T
T → TF |F
16. Consider the grammar
F → F *|a|b S → AbBaCc|e
A → aAb|ba
B → bBc|cb
The above given grammar is
(a) Not LR(0) due to S-R problem C → cCa|ac
(b) Not LR(0) due to R-R problem
(c) LR(0) 16a. Find the First( ) of S ?
(d) None of the above (a) {e} (b) {a, e}
(c) {a, b, e} (d) {a, b, c, e}
12. Consider the above grammar. If a grammar is
LL(1) then 16b. Find the Follow( ) of B?
(a) It will be LR(0). (b) It will be SLR(1). (a) {a} (b) {b}
(c) It will be LALR(1). (d) None of the above. (c) {a, b} (d) {a, c}
13. Consider the CFG given below with {S, X, Y} 16c. Consider the grammar (G) = {A → A + A|a|e},
non-terminal and {a, b} terminals. S is the starting which one is the correct after removing left recursion?
(a) {A′ → AA′, A′ → |a|e}
symbol and set of production rules are
S → bX |aB (b) {A′ → A + A′|e, A′ → |a}
A → a|aS |bAA (c) {A′ → A + A′|a|e}
B → b|bS |aBB (d) {A′ → +AA′|e, A → aA′}
Set 1
1. (a) 10. (b) 19. (d) 28. (c) 37. (c) 46. (c)
2. (c) 11. (d) 20. (c) 29. (d) 38. (b) 47. (b)
3. (a) 12. (d) 21. (a) 30. (a) 39. (b) 48. (a)
4. (c) 13. (a) 22. (d) 31. (c) 40. (d) 49. (c)
5. (d) 14. (a) 23. (d) 32. (a) 41. (a) 50. (c)
6. (b) 15. (b) 24. (a) 33. (b) 42. (d)
7. (c) 16. (d) 25. (c) 34. (d) 43. (a)
8. (d) 17. (b) 26. (d) 35. (b) 44. (12338)
9. (b) 18. (d) 27. (d) 36. (b) 45. (a)
Chapter 6.indd 323 4/9/2015 9:59:55 AM

Set 2
1. (b) printf ( “ a = %d , b = %d” , a , b ) ; 7b. (c) First( ) of E′ = {a, [ }

Number of tokens in the above statements is 9, are 7c. (c) Follow(E) consists of three things:
underlined separately.
S (260) i. The terminal which follow variable E in the
2. (c)
right side of any production.
a [∗] [26 ∗ 10 = 260]
26 S T 10
ii. If any variable follow E in the right side of pro-
a [∗] 4 F [6 + 4 = 10] duction, then first of that variable will be in
[2 ∗ 13 = 26] 2S T 13 6T $ [+]
follow of E.
8 T $[+] 5 F 6F (4)Num iii. If E is the starting symbol, then add $ sign into
2T
the follow of E.
2F 8F (5)Num (6)Num So, Follow(E) ={$, ]}
7d. (d) Refer to solution of Q9.
(2)Num (8)Num
8. (c) Since an unambiguous grammar has leftmost
3. (c) Grammar
derivation and rightmost derivations but both of
1. S → WX |Y them are not same.
2. W → wWx |wx 9. (d) In LALR, all the states of CLR which are differ
3. X → yXz |yz only by look-ahead symbol are merged. So, the
4. Y → wYz |wZz number of states in LALR may be less than or
5. Z → xZy |xy
equal to the number of states in CLR.

10. (d) A grammar (G) is said to be an operator gram-
From the second and third production we can
get (wnxnymzm: n ≥ 1, m ≥ 1) and from the fourth
mar if and only if
and fifth production we can get {wnxmymzn: n ≥ 1, i. Grammar G does not have any null production.
m ≥ 1}, so option (c) is correct. ii. G does not have any two adjacent variables on
the right-hand side of the production.
4. (c)
11. (a) Draw parsing table for the grammar, for details
S S procedure of parsing table refer text.
w x y 12. (c) Check the relationship diagram between LL(1)
and LR(0), SLR(1), LALR(1) (refer text).
w z
w w x y x z y 13. (a) S → aB → aaBB → aabB → aabbS → aabbaB
→ aabbab
w z
z 14. (b) Two parse trees are possible.
w x y z
15. (c) Symbol table is used to store all the informa-
x z y
tion about a variable such as variable name, id and
type.
x y 16a. (c)
5. (b) The given grammar is ambiguous, so it is not 16b. (d)
LL(1), and hence not LR(1).
16c. (d) Steps to remove left recursion:
6. (d) The multiplication string itself is a complete string
which does not require any further character. A′ → +AA′|e
A → aA′
7a. (a) First(T) consists the terminal which is the first in
the production of T. So, First(T) = First(F) = {a, [ }
Chapter 6.indd 324 4/9/2015 9:59:56 AM

UNIT VII: OPERATING SYSTEM
6
Number of questions
4 Marks 1
Marks 2
3
2
0
2015 2014 2013 2012 2011 2010
Year Concept
2015 Shared variable, SCAN algorithm, Page replacement, Processes, Process synchronization,
Deadlock, Process scheduling, TLB, Best fit algorithm, Page table
2014 Page Frames, Deadlock, Translation Lookaside Buffer (TLB), Disk Scheduling Algorithm
2013 Scheduling, Disk Management, Semaphores, Concurrent Processes, Memory Management
2012 System Calls (Fork), Scheduling Algorithms, Concurrent Processes
2011 DMA Concept, Interrupts, Lightweight Processes, User–Kernel Mode, Scheduling Algorithms
2010 Deadlock, Critical Section, Page table, Concurrent Process Concept
Chapter 7.indd 325 4/9/2015 10:01:07 AM

Chapter 7.indd 326 4/9/2015 10:01:07 AM
CHAPTER 7
OPERATING SYSTEM
Syllabus: Operating system: Processes, Threads, Inter-process communication, Concurrency, Synchronization, Deadlock,
CPU scheduling, Memory management and virtual memory, File systems, I/O systems, Protection and security
7.1 INTRODUCTION The function of the operating systems is to provide

an environment in which a user can execute his/her
commands in a convenient and efficient manner.
An operating sysem (OS) can be defined as an inter- Operating system is also called resource allocator.
mediate program between the user and the computer The placement of operating systems in a computing
hardware. On the user end, it handles application pro- environment is shown in Fig. 7.1.
grams and at the other end it makes use of system call(s)
to instruct the hardware to perform certain task as
User
instructed by the user. There are different types of oper-
ating systems to accomplish various tasks. Mainframe
operating systems are designed to optimize hardware User application
utilization. The real-time operating systems are designed
for timeliness, the time-sharing systems are designed for
Operating system
efficient usage of the system and other resources. Some of
the examples of OS are: UNIX, Linux (different flavors,
such as, Redhat, Suse, Fedora, Debian etc.), MacOS, Physical device (hardware)
Microsoft Windows, etc. Nowadays, the mobiles that
Figure 7.1 | Location of an operating system in a

we use are also equipped with operating system. Some
of the well-known OS for mobiles are – Symbian and
Android. computing environment.
Chapter 7.indd 327 4/9/2015 10:01:07 AM

328 Chapter 7: Operating System
System Call is a request made by a user program to 7.2.2 Multiprogramming Operating System
get the service of an operating system.
To consider the “inside view” of the system, the OS is A single user cannot keep either the CPU or the I/O
written as a collection of procedures, each of which can devices busy at all times. Multiprogramming increases
call any other, whenever required. Each parameter has a CPU utilization by organizing jobs so that the CPU is
well-defined interface with input parameters and output not idle at any time (Fig. 7.4). Jobs to be executed are
results. A simplified layered structure of Dijkstra’s maintained in the memory simultaneously and the OS
“THE” operating system is shown in Fig. 7.2: switches among these jobs for their execution. When
one job waits for some input (WAIT state), the CPU
switches to another job. This process is followed for all
LAYER FUNCTION jobs in memory. When the wait for the first job is over,
the CPU is back to serve it. The CPU will be busy till all
5 The operator (user) the jobs have been executed. Thus, increasing the CPU
utilization and throughput.
4 User programs
3 Input/output management
Job 2 Job 1
2 Operator-process communication
CPU
1 Memory management Job 4 Job 3 JOB 3
0 Process allocation and multiprogramming Job 5 JOB 1
Figure 7.2 | Simplified structure of Dijkstra’s OS I/O

THE operating system. Memory
Figure 7.4 | A multiprogramming operating system.
7.2 TYPES OF OPERATING SYSTEM 7.2.3 Multitasking Operating System
Multitasking system is the extension of multiprogram-

7.2.1 Batch Operating System ming system. In this system, jobs will be executed in the
time-sharing mode. It uses CPU scheduling and multi-
In a batch operating system, tasks are submitted to a
programming techniques to provide a small portion of a
pool, and only after the completion of one job, the next
timeslot to each user. Each user has at least one separate
job is executed (Fig. 7.3). In this system, the CPU idle
program in memory.
time (free time) is high and throughput of the system
is low. Throughput is the number of jobs executed per
unit time. 7.2.4 Multiprocessor Operating System
Multiprocessor operating system consists of more than

Job 1 one processor for execution (Fig. 7.5). It improves the
throughput of the system and it is also more reliable
Job 2 than the single processor system because if one processor
CPU breaks down then the other can share the load of that
Job 3
processor (Table 7.1).
Job 4
....... Batch
....... CPU 1 CPU 2 CPU 3 CPU 4
I/O
Job n
OS
Memory Memory
Figure 7.3 | A batch operating system. Figure 7.5 | A multiprocessor operating system.
Chapter 7.indd 328 4/9/2015 10:01:09 AM

7.3 PROCESS MANAGEMENT 329
Table 7.1 | Comparison between Batch Processing processor is called a process. A process has various attri-
and Timesharing OS butes, states and operation.
Batch Timesharing
Processing 7.3.1.1 Attributes of a Process
AIM Maximize Minimize A process is distinct from another based upon the follow-
processor response time ing attributes:
utilization
1. Process id: It is a unique identification number
COMMANDS Job Control As per the which is assigned by the OS at the time of process
Language (JCL) commands creation.
provided along provided at 2. Process state: It contains the state of the process
with the job the terminal where the process is currently residing. The state
may be new, ready, running, waiting, halted and
7.2.5 Real-Time Operating System so on.
3. Program counter: It is having the address of the
A real-time operating system (RTOS) is the one which next instruction to be executed.
deals with real-time application requests (Fig. 7.6). 4. Priority: It is also a parameter assigned by the
The input data must be processed without any delays. OS at the time of process creation.
The processing time is shorter and the waiting time is 5. General purpose register
minimum. If an RTOS meets a deadline approximately, 6. List of open files
it is known as a soft RTOS, but if deadline has to be met 7. List of closed files
deterministically (hard deadline), it is a hard RTOS. All the attributes of a process are stored in the process
The events that an RTOS may have to respond to can be control block (PCB) (Fig. 7.7).
either periodic (occurring at regular intervals) or aperiodic
(occurring at irregular intervals). Further, if there are n peri- Process Id Process State
odic events and an event i occurs with period Pi and requires Process counter Priority
Ti seconds of CPU time, then the load can be handled if
List of open files List of close files
n
∑ (Ti / Pi ) ≤ 1 General purpose register
information
Accounting information
i =1
An RTOS that meets this criteria is said to be schedulable. I/O status information CPU scheduling
information
Memory management Context data memory
Real time OS
information pointers
Figure 7.7 | A process control block.
Soft real time OS Hard real time OS

7.3.1.2 Process States
Figure 7.6 | Classification of a real-time operating system. Any process is defined by a unique characteristic of its
An operating system is concerned with the management state, which informs about the current activity of the
functions of the following resources: (a) process man- process. A process may be in one of the following states
agement, (b) memory management, (c) file management during its execution:
and (d) device management. 1. New: Whenever a new process is created.
2. Running: Whenever the process is able to get
7.3 PROCESS MANAGEMENT attention from the busy CPU, that is, the CPU is
actually processing the instruction.
3. Waiting: The process waits for its turn as it
7.3.1 Process requires some resource(s) which is/are held by
another process(es).
A program under execution is called a process. In other 4. Ready: A process has been allocated all the
words, a program is in the main memory and the occupied resources required and is waiting for the attention
Chapter 7.indd 329 4/9/2015 10:01:09 AM

from CPU because the CPU is busy in executing state and scheduling that process into the run-
some other task. ning state.
5. Terminated: A process is in terminated state 3. Mid-term scheduler: It is responsible for sus-
when the processor has finished the execution of a pending and resuming the processor.
particular task.
6. Suspended ready: If processes are more than 7.3.3.1 Dispatcher
the capacity of the memory, then they will be sus-
pended and go to suspended ready state, which is It is responsible for saving and loading the context of a
in the secondary memory. process. The context switching is done by a dispatcher.
7. Suspended wait: If processes are more than the
capacity of memory, then they will be suspended 7.3.3.2 Degree of Multiprogramming
and go to suspended waiting state, which is in the
secondary memory. The number of jobs present in the memory is called as
the degree of multiprogramming. Long-term scheduler
controls the degree of multiprogramming.
7.3.2 Schedulers Note: The best data structure to implement ready
queue is double linked list.
An operating system has three types of scheduler
(Fig. 7.8): 7.3.3.3 Time Periods
1. Long-term scheduler: It is responsible for
A process has various time periods as follows:
bringing a newly created process into the system.
2. Short-term scheduler: It is responsible for 1. Arrival time (AT): When a process arrives into
selecting one of the processes from the ready ready state, it is called as arrival time.
Schedule/Dispatch Terminated
leted
Created Comp
New Ready Run
d Priority
e
nd
e
e
um
sp
Su
es
I/O Operation
R
I/O
Suspended Block Suspend Suspended
Complete
ready wait wait
Resume
Complete I/O, suspended due to lack of memory
Short-term scheduler works here
Mid-term scheduler works here
Long-term scheduler works here
Figure 7.8 | Lifecycle of a scheduler.
Chapter 7.indd 330 4/9/2015 10:01:10 AM

7.4 CPU SCHEDULING 331
2. Burst time (BT): The time required by a pro- Example 7.1
cess for its execution is called as burst time.
3. Completion time (CT): The time when a pro- Consider the following workload table of three processes:
cess completes its execution is called as completion
time. Process Burst Time (ms)
4. Turn-around time (TAT): The time difference P1 15
between completion time and arrival time is called
as turnaround time. P2 4
TAT = CT − AT P3 2
5. Waiting time (WT): Waiting time for a pro- Let us suppose that the process arrives at time = 0 in
cess is the time duration which is spent in waiting the order P1, P2, P3. By applying the FCFS scheduling
queue by that process. algorithm, we get the result shown in the following
WT = TAT − BT Gantt chart:
6. Response time: The first scheduled time for a P1 P2 P3
process is called response time.
0 15 19 21
Waiting Time:
7.4 CPU SCHEDULING Process Waiting Time (ms)
WT = (TAT − BT)
CPU scheduling is the technique to put a process from P1 0
ready queue to running state. The objective of CPU
scheduling is to minimize the turnaround time and aver- P2 15
age waiting time of a process. The short-term scheduler
P3 19
is used to apply the CPU scheduling in the ready state
to select the process to schedule on to the processor (run-
Average waiting time = (Total waiting time)/3
ning state).
= (0 + 15 + 19)/3 = 34/3 = 11.33 ms
Turnaround Time:
7.4.1 Scheduling Algorithms Process Turnaround Time (ms)
TAT = (CT − AT)
There are many CPU scheduling algorithms as follows:
P1 15
1. First come, first serve (FCFS)
2. Shortest job first (SJF) P2 19
3. Shortest remaining time first (SRTF)
P3 21
4. Round robin
5. Priority based Average turnaround time = (15 + 19 + 21)/3
6. Highest response ratio next (HRRN) = 55/3 = 18.33 ms.
7. Multilevel queue
8. Multilevel feedback queue
7.4.1.2 Shortest-Job-First Scheduling
7.4.1.1 First-Come, First-Serve Scheduling In SJF scheduling, the job that requires CPU atten-
tion for lesser time is chosen first for CPU attention.
In this simple scheme, the allocation of CPU is on the The data structure used is a queue in which the jobs
basis of first come, first serve (FCFS). A queue data are arranged on the basis of CPU time required by the
structure is maintained for FCFS scheduling. When process (in ascending order). If there are two processes
the wait of the process is over and it enters into the that have the same requirement of CPU time, then the
ready state, its PCB is linked onto the tail of the queue. order in which they arrived (FCFS scheduling) is used to
The process at the head of the queue is allocated to the break the tie. This algorithm can be either pre-emptive
processor and the running process is removed from the or non-preemptive. Pre-emptive SJF scheduling is some-
queue. The FCFS scheduling is non-preemptive. times called shortest-time-first scheduling (STF).
Chapter 7.indd 331 4/9/2015 10:01:10 AM

Continued
Example 7.2
Consider the same Example 7.1: Process Arrival Time Burst Time
(ms) (ms)
Process Burst Time (ms) P3 2 8
P1 15 P4 3 3
P2 4
P3 2 1. Now, P1 is executed first, because it is the only
process available at that time (t = 0).
Assuming that all these processes arrive at the same 2. At t = 1, we have a new process P2 which requires
time, using SJF scheduling, we would schedule these only 2 ms whereas the already executing process P1
processes according to the following Gantt chart: still requires 4 ms to complete. Therefore, the CPU
will start executing P2 because it has the shortest run-
P3 P2 P1
ning time out of the two available processes P1 and P2.
0 2 6 21 3. At t = 2, we have three processes, namely P1, P2
Waiting Time: and P3, P2 has the shortest remaining time, there-
fore it will continue to execute.
Process Waiting Time (ms) 4. At t = 3, process P2 terminates and a new process
WT = (TAT − BT) P4 enters the ready queue. Now, we have three
P1 6 processes P1, P3 and P4 waiting for CPU attention.
P4 has the shortest remaining time (3 ms) as com-
P2 2 pared to P1 (4 ms) and P3 (8 ms).
P3 0 The Gantt chart is shown below:
Average waiting time = (6 + 2 + 0)/3 = 8/3 = 2.66 ms
P1 P2 P4 P1 P3
Turnaround Time:
0 1 3 6 10 18
Process Turnaround Time (ms) Turnaround Time:
TAT = (CT − AT)
P1 21 Process Turnaround Time (ms)
TAT = (CT − AT)
P2 6
P3 2 P1 10 − 0 = 10
Average turnaround time = (21 + 6 + 2)/3 = 29/3

P2 3−1=2
= 9.66 ms P3 18 − 2 = 16
P4 6−3=3
7.4.1.3 Shortest Remaining Time First
Average turnaround time = (10 + 2 + 16 + 3)/4
SRTF is a pre-emptive version of the SJF. It permits pro- = 7.75 ms
cesses that enter the ready list to pre-empt the running Waiting Time:
process if the time for the new process (or for its next
Process Waiting Time (ms)
WT = (TAT − BT)
burst) is less than the remaining time for the running
process (or for its current burst).
P1 10 − 5 = 5
Example 7.3 P2 2−2=0
Consider the following example. We have the workload P3 18 − 8 = 10
table of four processes. P4 3−3=0
Process Arrival Time Burst Time Average waiting time = (5 + 0 + 8 + 10)/4 = 5.75 ms
(ms) (ms)
P1 0 5 7.4.1.4 Round-Robin Scheduling
P2 1 2
Round-robin scheduling is basically similar to FCFS
(Continued) scheduling, but pre-emption is added to switch between
Chapter 7.indd 332 4/9/2015 10:01:10 AM

7.4 CPU SCHEDULING 333
processes. A small unit of time, generally from 10 to 7.4.1.5 Priority Scheduling

100 ms, called a time quantum (or time slice), is defined.
The ready queue is treated as a circular queue. The In priority scheduling, each process is assigned a priority
round-robin (RR) scheduling algorithm is designed espe- and the highest priority process is allocated to the CPU
cially for time-sharing systems. first. The data structure used is a priority queue, that
is, the process with the highest priority is in the front of
Example 7.4 the queue. If there are two processes that have the same
priority, then their order of arrival (FCFS scheduling) is
Consider the following example. We have the workload used to break the tie. The major drawback of this algo-
table of four processes, and time quantum of 4 ms is rithm is starvation, that is, a low-priority process have
used. to wait for an infinite time because the high priority
processes gets executed first and after some time another
Process Burst Time (ms) high priority process comes and the low priority process
P1 21 is again left out.
P2 3 Example 7.5
P3 3 Consider the following example. We have the workload
table of five processes.
P4 5
Process Burst Time (ms) Priority
Using RR scheduling, we would schedule these processes
according to the following Gantt chart: P1 15 3
P2 1 1
P1 P2 P3 P4 P1 P4 P1 P1 P1 P1 P3 5 4
0 4 7 10 14 18 19 23 27 31 35 P4 1 5
P5 10 2
Waiting Time:
Using priority scheduling, we would schedule these pro-
Process Waiting Time (ms) cesses according to the following Gantt chart:
WT = (TAT − BT)
P2 P5 P1 P3 P4
P1 35 − 24 = 11
0 1 11 26 31 32
P2 7−3=4
Waiting Time:
P3 10 − 3 = 7
P4 19 − 5 = 14 P1 11
P2 0
Average waiting time = (11 + 4 + 7 + 14)/4 = 9 ms
P3 26
Turnaround Time:
P4 31
Process Turnaround Time (ms) P5 1
TAT = (CT − AT)
Average waiting time = (11 + 0 + 26 + 31 + 1)/5
P1 35 = 69/5 = 13.8 ms
P2 7 Turnaround Time:
P3 10 Process Turnaround Time (ms)
P4 19 P1 26
P2 1
Average turnaround time = (35 + 7 + 10 + 19)/4
= 17.75 ms P3 31
(Continued)
Chapter 7.indd 333 4/9/2015 10:01:11 AM

Continued 4. At t = 9, we have three processes P3, P4 and P5. To

select which process will be executed next, we have
Process Turnaround Time (ms)
to calculate the response ratio for all of the three
P4 32 processes. The process having the highest response
ratio will be executed first.
P5 11
At t = 9,
Response ratio of P3 = (w + s)/s = (5 + 4)/4
Average turnaround time = (26 + 1 + 31 + 32 + 11)/5 = 2.25
= 20.2 ms Response ratio of P4 = (w + s)/s = (3 + 5)/5
= 1.60
Response ratio of P5 = (w + s)/s = (1 + 2)/2 = 1.5
7.4.1.6 Highest Response Ratio Next Process response ration of P3 is high, so it will be
(HRRN) executed next.
5. Repeat the above steps for all the processes.
In HRRN algorithm, a processor is assigned to a process
on the basis of response ratio. Response ratio is calcu- Waiting Time:
lated by the following formula:
w+s
Response ratio = WT = TAT − BT
s
where w = waiting time and s = service time or burst P1 0
time. P2 1
The HRRN algorithm favours the shorter jobs and P3 5
limits the waiting time of larger jobs. It is non-preemptive
scheduling algorithm. P4 9
P5 5
Example 7.6
Average waiting time = (0 + 1 + 5 + 9 + 5)/5 = 20/5
Consider the following example. We have the workload
= 4 ms
table of five processes.
Turnaround Time:
Process Burst Time (ms) Arrival Time
P1 3 0 Process Turnaround Time (ms)
P2 6 2 P1 3
P3 4 4 P2 7
P4 5 6 P3 9
P5 2 8 P4 14
P5 7
Using HRRN scheduling, we would schedule these pro-
cesses according to the following Gantt chart: Average turnaround time = (3 + 7 + 9 + 14 + 7)/5
= 8 ms
P1 P2 P3 P5 P4
0 3 9 13 15 20
7.4.1.7 Multilevel Queue Scheduling
1. Now P1 is executed first, because it is the only
process available at that time (t = 0). In multilevel queue scheduling, the ready queue is par-
2. At t = 2, we have a new process P2 which requires titioned into several separate queues such that each
only 6 ms whereas the already executing process process belongs to some queue. Various properties for
P1 still requires 1 ms to complete. Therefore, the allocating jobs to different queues may depend upon
CPU will continue to execute P1 because it is non- memory size, process priority or process type, or any
preemptive scheduling. other parameter(s). Each queue may follow different
3. At t = 3, we have only one process P2, so P2 will scheduling algorithms. Figure 7.9 depicts the multilevel
get executed. queue scheduling.
Chapter 7.indd 334 4/9/2015 10:01:11 AM

7.5 PROCESS SYNCHRONIZATION 335
Highest priority of attention will be given to user. For example, if

System processes there are n users, then we can say that each user
will get about 1/n of the CPU time. This seems to
Interactive processes be a good scheduling algorithm theoretically but is
harder to implement due to overheads involved in
switching of processes.
Interactive edit processes 2. Lottery scheduling: This is similar to lottery.
Whenever a scheduling decision has to be made,
any process is selected at random and is given more
Batch processes
amount of CPU time as compared to other pro-
Lowest priority cesses similar to the lottery winner.
User processes
There are two different categories of schedulers—Dynamic
Figure 7.9 | Multilevel queue scheduling.
and Static. A dynamic scheduling can be defined as a
scheduling defined at run-time; whereas in static the
scheduling decision is taken before the system starts run-
7.4.1.8 Multilevel Feedback Queue ning. There are some popular dynamic scheduling algo-
Scheduling rithms such as:
1. Rate monotonic algorithm: A priority value
In multilevel feedback queue scheduling, the multilevel is assigned to each process depending upon the
queue scheduling approach has been enhanced. Here, the frequency of the occurrence. A process with high
process is not permanently assigned to a queue but a frequency occurrence gets a high priority. So, at
process is allowed to move between queues. The major run-time the highest priority process gets the
advantage of this approach is that it prevents starvation, CPU.
because a process which is a low-priority process can be 2. Earliest deadline first: The processes are
moved to a higher-priority queue after some time. sorted according to the deadline (say, in RTOS)
Figure 7.10 depicts a three-level feedback queue and the CPU time is allocated to the first process
scheduler. The scheduler first executes all processes in in the list, that is, which has to meet the deadline
the first queue. When this queue is empty, it will execute urgently.
processes in the next queue (having a higher quantum 3. Least laxity algorithm: A process that has the
or time-slice). Similarly, processes in subsequent queue least spare-time (laxity) is given to the CPU.
will be executed only if the previous queue(s) are empty.
Further, a process that arrives for lower queue (having
lower quantum) will pre-empt a process in queue having 7.5 PROCESS SYNCHRONIZATION
a higher quantum.
7.5.1 Types of Processes

QUANTUM = N
A process consists of two major elements namely, the
program code and a data set. The processes can be clas-
sified as follows:
1. Independent process: This is a process which
QUANTUM = 2N
cannot affect or get affected by execution of another
process.
2. Cooperating process: This is a process that
affects and gets affected by execution of other
processes.
FIFO
3. Concurrent processes: These processes are
Figure 7.10 | Multilevel feedback queue scheduling.
independent, interacting processes which are exe-
cuted simultaneously over a period of time.
Note: Apart from the above well-known scheduling
When two processes want to use or access a single
algorithms, there are two other approaches also.
global variable, then the problem of inconsistency
1. Guaranteed scheduling: This scheduling algo- occurs. The problem can be solved using the process
rithm guarantees the user that particular amount synchronization.
Chapter 7.indd 335 4/9/2015 10:01:12 AM

7.5.2 Interprocess Communication data from file or some want to insert or change the con-
tents in the file. According to Bernstein constraint, many
Inter-process communication is a technique to exchange readers may access the file at the same time, but when
data among multiple processes or threads. IPC allows the writer is accessing the file, no one should be allowed
programmers to schedule activities to different processes. to access it. This can be ensured by following methods:
One process controls the whole communication. There
are different IPC techniques available: 1. Assign readers more priority than writers. If a
writer has no permission to access the file, any
• Message passing reader can get access to it. But this may cause
• Shared memory infinite waiting for the writer.
2. Another method can be used to assign writers large
• Remote procedure calls
priority over readers. When writer requests for file,
• Synchronization all the current readers finish their access and new
readers have to wait until the writer finishes its
7.5.3 Classical Synchronization Problems task. This may cause infinite waiting for reader.
Classical synchronization problems are used to imple-

ment synchronization mechanisms and critical sections. 7.5.3.3 The Dining Philosopher’s Problem
Some of the interesting problems are: The Producer-
Consumer Problem, The Readers-Writers Another classic problem of synchronization is the dining
Problem, The Dining Philosophers Problem, philosopher’s problem. Here five philosophers sit around
Sleeping Barber, Cigarette Smoker Problem, a circular table with five chairs. There are five plates and
etc. The following sections discuss only the first three forks placed in a circular manner (Fig. 7.11). When the
of them. philosopher wants to eat, s/he picks up forks from left
and right of his plate. When finishes eating, s/he keeps
both the forks back. This ensures that no philosopher
7.5.3.1 The Producer-Consumer Problem starves. Two problems are faced in this:
To further understand the concept, let us take an 1. Each philosopher picks up one fork and none of
example of producer–consumer problem. A producer them gets a second fork.
produces data that is consumed by a consumer (e.g., 2. No philosopher should starve due to another.
spooler and printer). A buffer holds the produced data,
which is not yet consumed. There exist several produc-
ers and consumers, and initially, we assume the buffer
is unbounded.
Let count be a global variable. Count = 5
Producer: Register1 = Count [5]
Producer: Register1 + = 1 [6]
Consumer: Register2 = Count [5]
Consumer: Register2 = Register2 -1 [4]
Producer: Count = Register 1 [6]
Consumer: Count = Register 2 [4]
Figure 7.11 | The Dining Philosopher’s Table.

As we can see, output will be either 4 or 6. But it should
be 5.
Here, output depends on the order of execution of
statements. Synchronization is required, otherwise 7.5.4 Race Condition
output will be inconsistent.
When several processes have access to shared data, then
7.5.3.2 The Readers-Writers Problem the final value of this shared data will eventually depend
upon the action performed by the last process. This is a
In this problem, there are many concurrent processes that situation in which the result depends upon which process
try to access the same file. Some processes try to read is executed first. Such a situation or condition is known
Chapter 7.indd 336 4/9/2015 10:01:13 AM

7.5 PROCESS SYNCHRONIZATION 337
as “race condition”. The producer–consumer example For example the process is checking repeatedly whether
discussed in the previous section is one such example. a lock is available or not. This technique can be used
The part of the program where the shared memory is to generate the random time delays. This consumes so
accessed is known as critical section. The following are many CPU cycles.
the four conditions that are used to provide a good solu-
do
tion to this problem:
1. No assumption has to be made about the number {

of CPUs and their speed. wait(s);
2. No two processes may be executing their critical // critical section
section at the same time. signal(s);
3. No process has to wait indefinitely to enter its criti- // remainder section
cal region. } while(1);
4. No process executing outside its critical section can
block other processes.
Processes keep on checking for semaphore value to
be zero. This continuous looping problem is busy
7.5.5 Critical Section
waiting. When semaphore does busy waiting, it is called
`spinlock’.
Critical section is a part of program global variables
which are manipulated by the processes.
7.5.5.2 Solutions to Critical Section
7.5.5.1 Critical Section Problem
Problem
Critical section problem occurs when n number of pro-
1. Mutual exclusion: If process Pi is executing in
cesses are competing for critical section. The problem is
its critical section, then no other process can exe-
to ensure that if one process is in the critical section, no
cute their critical sections.
other process is allowed to enter into critical section at
2. Progress: If no process is executing in its critical
the same time.
section and there exist some process that wishes
do { to enter into critical section, then the selection of
process that will enter into critical section cannot
Entry section
be postponed indefinitely.
Critical Section 3. Bounded waiting: A bound must exist on the
number of times that other processes are allowed
Remainder section to enter into critical section after a process has
made the request for critical section and it is not
Remainder Section granted.
} while(1);
7.5.6 Peterson’s Algorithm for Two Processes

Busy Waiting
Spinning or busy waiting is a mechanism in which the Peterson’s algorithm is considered the most correct algo-
process keeps on checking for the condition to be true. rithm for process synchronization.
Pi Pj
do{ do{ do{ do{

flag[i]
{
flag[i]
= true;
turn =turn
j; = j;
while(flag[j]
= true;
while(flag[j]
and turn
and =turn
j); = j);
{flag[j]
flag[j]
= true;
turn =turn
i; = i;
while(flag[i]
= true;
while(flag[i]
and turn
and =turn
i); = i);
Critical
Critical
Section
Section Critical
Critical
Section
Section
flag[i]
flag[i]
= false
= false flag[j]
flag[j]
= false
= false
Remainder
Remainder
Section
Section Remainder
Remainder
Section
Section
}while(1);
}while(1); }while(1);
}while(1);
Chapter 7.indd 337 4/9/2015 10:01:13 AM

The above code ensures mutual exclusion, progress and while S ≤ 0

bounded wait. Do no_operation;
S--;
7.5.7 Semaphores 2. Signal: It increases the value of semaphore by 1.
This can be denoted by V(S).
Semaphores are synchronization tools. These are the
Signal(S):
shared variables, which can be implemented as inte-
ger variables. These can be accessed via two indivisible S++;
atomic operations that are wait and signal. 3. Disadvantages:
7.5.7.1 Integer Implementation of • We must keep the value of S = 1
Semaphores • Busy waiting which consumes many CPU cycles
The operations (wait and signal) are realized by making

use of integer values. 7.5.7.2 Binary Semaphores
1. Wait: It reduces the value of semaphore by 1 unit. The value of binary semaphores can range between 0
It can be denoted by P(S). and 1 only. Binary semaphores are also known as mutex
Wait(S): locks as they ensure mutual exclusion.
Problem 7.1: Each process Pi = 1, 2, … 11 is coded as follows:
P (mutex)
Critical Section
V (mutex)
The code for P10 and P11 is identical except it uses V in the place of P and P in the place of V.
How many processes will be in critical section?
Solution:
Let S = 1
P1 P2 P3 P4 P5 P6 P7 P8 P9 P10 P11
P P P P P P P P P P V
V V V V V V V V V V P
Say P1 enters in the critical section, S1 → 0.

Now P11 process can increment S → 1 and will have access to critical section.
Now any process can enter into critical section, because S is incremented by P11.
Three processes will enter into critical section.
Note: 1. With change in one sequence, three processes will enter in the critical section.
2. With change in two sequences, five processes will enter into the critical section and so on.
7.5.7.3 Counting Semaphores whenever the instance of that resource is allocated to a

process. So, if the count value of a resource is zero, it means
This semaphore makes use of a `count’ variable. This vari- that all instances of that resource have been allocated and
able is initialized by the numeric value of the finite number no process can enter into critical section. Further, when a
of instances of a resource. The count value is decremented process releases a resource this count value is incremented.
Chapter 7.indd 338 4/9/2015 10:01:13 AM

7.6 DEADLOCKS 339
Implementation of Counting Semaphore Table 7.2 | Continued
1. Wait operation 5. Event Flag It is a memory word (such as

Wait (S): accumulator flags), which is used
{ for synchronization.
C--; 6. Condition A particular data type is used to
if(C < 0) Variable block a given process until some
{ specified condition is not met.
Block the process and put
7. Mailboxes The two processes can exchange
in blocked queue
} information through mailbox
} and also can take its help for
synchronizing their activities.
2. Signal operation
8. Spinlocks A process executes in an infinite
Signal (S):
loop until the locked variable is
{ available.
C++;
if(C ≤ 0)
{ 7.5.8 Deadlock and Starvation
Remove process from block queue
and keep in ready queue
} P0 P1
} wait(S); wait(Q);
wait(Q); wait(S);
Example 7.7
. .
Counting semaphore = 7 . .
P = 20 [decrement semaphore] . .
V = 15 [increment semaphore] . .
Value after implementing P and V: C − P + V signal(S); signal(Q);
7 − 20 + 15 = 2
signal(Q); signal(S);
Table 7.2 summarizes the common concurrency In the above code, let S = 1, Q = 1
mechanisms
Process P0 will make S → 0
Table 7.2 | Common concurrency mechanisms
Process P1 will make Q → 0
1. Semaphore An integer value is used for
Both will not be able to move further. This is the dead-
signalling between different
lock state.
processes. Only three operations
are performed — initialize,
increment and decrement. Also 7.6 DEADLOCKS
known as counting semaphore.
2. Binary Only two values 0 and 1 are used. A deadlock is a situation where computer programs sharing
Semaphore the same resources are permanently blocked. As a result
3. Mutex Similar to binary semaphore (lock of which, each program prevents the other from acquiring
and unlock values are used) but the the resources required for its completion. Let us consider
major difference is that the process an example, there are two friends (processes) and they
which has locked the mutex is the intend to write a letter. To accomplish this task they need
only one that can unlock it. two pens and two pages (resources). Now, one of them has
4. Monitor It is a line of code that contains pages (papers) and the other a pen. A possible solution
access procedures, initialization would have been that one person can give his resource to
code and variables. The access another so that the other person completes the job (write
procedures are critical sections. the letter) and after his job is finished this person can also
complete the job. But none of them is willing to part the
(Continued) resources s/he has. This is a simple case of deadlock.
Chapter 7.indd 339 4/9/2015 10:01:14 AM

7.6.1 Resource Types 7.6.3 Resource Allocation Graph
There are two types of resources, namely, pre-emptable A resource allocation graph is a graph which has two dif-
and non-preemptable resources. ferent types of nodes, the process nodes (represented by
circle) and resource nodes (represented by rectangles).
1. Pre-emptable: These resources can be taken
For different instances of a resource, there is a dot in
away from the process with no ill effects, for exam-
the resources nodes (rectangle). For example, if there
ple, CPU, memory.
are two identical drives, we will depict it with two dots.
2. Non-preemptable: These resources can be
taken away from the process (without causing The directed edges among these nodes represent
any ill effect). For example, CD, USB drive, plot- resource allocation and release. An edge from the resource
ter, printer, etc. We cannot give such resources to to the process node denotes that it has been acquired by
other processes in the middle of their jobs. the process, whereas an edge from the process node to
the resource node denotes that the process is request-
ing for the resource. For example, consider the resource
7.6.2 Characteristics of a Deadlock allocation graph in Fig. 7.13 and we can easily conclude
that there is a deadlock.
The following four conditions must exist simultaneously
for a deadlock to occur:
1. Mutual exclusion condition: The resources Disk

involved are non-shareable. Each resource must be
assigned to only one single process at a time.
2. Hold and wait condition: The process currently
holds a resource and is waiting to acquire another
requested resource held by other process(es). P1 P2
3. Non-preemptive condition: Resource(s) that
have already been acquired by a process cannot
be taken back forcibly unless and until they are
released by the holding process.
4. Circular wait condition: The processes form a
Printer
circular list or chain, where each process in the list
Figure 7.13 | Resource allocation graph.

is waiting for a resource held by another process.
To understand these four concepts clearly, let us take an
example of the road as a resource, shown in Fig. 7.12. By making use of resource allocation graph we can
check for deadlock (i.e., the graph has a circle) and if
there is no circle, it indicates the lack of circular-wait
condition, hence deadlock will not occur. This is depicted
in Fig. 7.14.
R1
P2
Figure 7.12 | A traffic deadlock.
P1
1. Mutual exclusion condition: Only one vehicle

can be moved.
2. Hold and wait condition: Each vehicle occu-
pies some part of the road and is waiting to move
to the other side of road. R2
3. Non-preemptive condition: Road space occu- Figure 7.14 | A resource allocation graph having a
pied by a vehicle cannot be taken away. cycle and no deadlock. (The figure also
4. Circular wait condition: Each vehicle on the depicts an alternative way of representing
road is waiting for the next vehicle to move. resource allocation graphs.)
Chapter 7.indd 340 4/9/2015 10:01:15 AM

7.6 DEADLOCKS 341
The following conditions are depicted in the resource 1. Mutual exclusion condition: Let us take the
allocation graph shown above: example of tossing a coin, getting a head or tail is
1. There are three sets of process (P), resource mutually exclusive, which means that we cannot
(R) and edges (E): have both head and tail. The mutual exclusion
P = {P1, P2} condition means that there are some resources
(such as CD Drive, Printer, etc.) which once allo-
R = {R1, R2} cated to a process cannot be taken back unless and
E = {R1 → P1, P1 → R2, R2 → P2, P2 → R1} until the process releases it (i.e., non-preemptable
2. Resource instances: resources). So, such resources should be allocated
Two instances of R1 to only one process at a time. They cannot be
shared. Although, this does not mean that none of
One instance of R2 the resources can be shared. Certain pre-emptable
3. Process states: Process P1 is holding an instance resources (such as memory, CPU) can be shared,
of resource type R1 and is waiting for an instance because they will never result in a deadlock. So,
of resource type R2. Further, process P2 is holding those resources which cannot be shared are mutu-
an instance of resource type R2 and is waiting for ally exclusive and this condition needs to be taken
an instance of resource type R1. care by OS.
2. Hold and wait condition: In order to ensure
The following tips help us to check the resource alloca-
this condition never occurs in the system, we have
tion graph easily and predict the presence of cycles.
to devise a strategy such that a process requests
1. If there is no cycle in the resource allocation graph, all its required resources at one time and blocks
there is no deadlock. the process until all requests can be granted. This
2. If there is a cycle in the graph and each resource approach has its drawback that a process cannot
has only one instance, then there is a deadlock. In know what resources would be required before it
such a case, the presence of a cycle is the necessary actually starts execution, and sometimes it may
condition in order for deadlock to occur (as there is have to wait for a long time in order to get all
only one instance of the resource). its required resources; whereas it could have
3. If there is a cycle in the graph, and each resource used some of the already allocated resources and
has more than one instance, there may or may not released them back to the pool to be used by other
be a deadlock. Therefore, we conclude that a cycle processes. For example, a program may require five
in the resource allocation graph is a necessary but CDs, there should be five free drives before the
not a sufficient condition for a deadlock in the case execution begins, while the program might be using
of multiple instances of a resource. only one drive to begin with. Thus, this approach
leads to serious waste of resources.
The following table summarizes the above discussion 3. Non-Preemption condition: The non-
preemption condition can be overcome by the
Possibility of Dead- Existence of Dead- following protocol. If a process is holding some
lock (03 conditions) lock (04 conditions) resources and requests another resource that cannot
be immediately allocated to it (which means, the
Mutual Exclusion, No Mutual Exclusion, No process must wait), then all resources currently
Preemption & Hold and Preemption, Hold and being held are pre-empted (released). These pre-
wait wait & Circular Wait empted resources are added to the list of resources.
For example, when a process releases resources,
A deadlock can be dealt by following any of the three the process may lose all its work to that point.
strategies: The major disadvantage of this strategy is the pos-
sibility of indefinite postponement (starvation).
1. Deadlock prevention
However, this protocol is often applied to resource
2. Deadlock avoidance
whose state can be easily saved and restored later,
3. Deadlock detection and recovery
such as CPU registers and memory space. It cannot
generally be applied to printers and drives.
7.6.4 Deadlock Prevention 4. Circular wait condition: A linear ordering
of different resources can be used to allocate the
To prevent a deadlock, we must ensure that any one of resources to a process, which means that if a pro-
the following four conditions listed below (Section 7.5.2) cess has been allocated a resource, then it can
should not hold. Let us consider them one by one: only be allocated a resource in the linear order.
Chapter 7.indd 341 4/9/2015 10:01:15 AM

Another way to implement it could be use some 1. Basic principle: It is similar to a banking system,
kind of indexing or a subscript. Such a strategy is that is, to ensure that the bank never allocates its
helpful in avoiding cycle(s) in resource allocation available cash such that it can no longer satisfy the
graph and results in overcoming the circular wait needs of all its customers.
condition. 2. Working: When a new process enters the system,
it must declare the maximum number of instances
of each resources type it may need. This number
7.6.5 Deadlock Avoidance should not exceed the total number of resources in
the system.
The side effects of preventing deadlocks are similar to
the side effects of walking carefully so that you donot Now, when a user requests a set of resources, the system
fall as compared to walking freely. These are as follows: must determine whether the allocation of these resources
1. Low device utilization will leave the system in a safe or in unsafe state. If in
2. Reduced system throughput safe state, then the resources requested are allocated,
otherwise the process must wait until some other process
Another approach is to anticipate a deadlock before it releases the requested resource.
actually occurs. This method differs from deadlock pre-
We need the following data structures:
vention, which guarantees that deadlock cannot occur
by denying any of the four necessary conditions of the 1. Available: A vector of length m indicates the
deadlock, whereas a `Deadlock Avoidance’ algorithm number of available resources of each type. If
dynamically examines the resource-allocation state to Available[j] = k, there are k instances available of
ensure that there can never be a circular wait condition. resource type Rj.
There are three basic pillars to attain this objective: 2. Max: An n × m matrix defines the maximum
1. Safe state algorithm demand of each process, if Max[i, j] = k, then pro-
2. Resource allocation graph algorithm cess Pi may request at most `k’ instances of resource
3. Banker’s algorithm type Rj.
3. Allocation: An n × m matrix defines the number
of resources of each type currently allocated to
7.6.5.1 Safe State each process.
•• If Allocation[i, j] = k
A safe state can be defined as a state in which there is •• Then, process Pi is currently allocated `k’
no deadlock. This can be achieved if instances of resource type Rj.
1. All the requested resources are allocated to the 4. Need: An n × m matrix indicates the remaining
process. resources needed for each process.
2. If a process needs an unavailable resource, it may •• If Need[i,j] = k
wait until the same has been released by a process •• Then, process Pi may need `k’ more instances of
to which it has already been allocated. If such a resource type Rj to complete its task.
sequence does not exist, it is known as unsafe state. •• Further, Need[i, j] = Max[i, j] − Allocation[i, j]
The Banker’s algorithm consists of the following two
Taking care of these two rules will never result in a
parts:
deadlock.
1. Safety algorithm: The algorithm for finding out
7.6.5.2 Resource Allocation Graph whether a system is in a safe state can be defined
as follows:
Algorithm
•• Let `Work’ and `Finish’ be vectors of lengths m
The resource allocation graphs can also be used for dead- and n, respectively.
lock avoidance. The major limitation of this strategy is Work: = Available
that it is not applicable to a resource allocation system Finish [i]: = False, for i = 1,2,3, …, n
with multiple instance of each resource type. •• Find an ì’, such that
Finish[i]: = False
Need i ≤ work, if no such i exists, go to step 4.
7.6.5.3 Banker’s Algorithm (Proposed by •• Work: = Work + Allocation
Dijkstra’s) Finish[i]: = true
Go to step 2.
Banker’s algorithm is used to avoid deadlocks, is dis- •• If Finish[i]: = true for all i, then the system is
cussed in the following text: in a safe state.
Chapter 7.indd 342 4/9/2015 10:01:15 AM

7.6 DEADLOCKS 343
2. Resource request algorithm does, the request is granted; otherwise, it is postponed
for some time later. This algorithm has a time complex-
If Requesti[j] = k, then process Pi wants k instances
ity of O(n2), where n is the number of processes.
of resources type Rj. When a request for resources is
made by process Pi, the following actions are taken: Limitations of the Banker’s Algorithm
•• If Requesti ≤ Needi, go to step 2. The following are some limitations of the banker’s
Otherwise, raise an error condition because the algorithms:
process has exceeded its maximum claim.
•• If Requesti ≤ Available, go to step 3. 1. A fixed number of processes and resources are
Otherwise, Pi must wait, as the resources are required. Further, all the processes should know
not available. their requirement of resources before starting.
•• Let the system pretend to have allocated the 2. No other process can start until the algorithm has
requested resources to process Pi by modifying been completely executed.
the state as follows: 3. It relies on the input information, the resources
Available: = Available − Requesti; may get underutilized if inaccurate information is
Allocationi: = Allocationi + Requesti; provided.
Needi = Needi − Requesti 4. This algorithm itself takes good amount of time
to execute, more information about processes and
Now, if the resulting transactions are safe, process Pi is resources will result in more time to execute.
allocated its resources. 5. The requests granted should be in a finite time.
6. If a resource becomes unavailable (e.g., a CD drive
Time Complexity
stops functioning), it can result in an unsafe state.
Banker’s algorithm considers each request as it occurs, 7. It is difficult to have a priori information about the
and checks whether it leads to a safe state or not. If it need of resources.
Problem 7.2: Consider a system having five processes and three resources types A, B and C. Resource type A has
nine instances, B has four instances and C has seven instances. Initially, the snapshot of the given system is as
follows:
Process Allocated Maximum Available

ABC ABC ABC
P0 0,1,0 7,5,3 2,2,2
P1 2,0,0 3,2,2
P2 3,0,2 9,0,2
P3 2,1,1 2,2,2
P4 0,0,2 4,3,3
Using Banker’s algorithm, how can we avoid the deadlock?

Solution: We will start with the first part of the Banker’s algorithm, as follows:
Step 1.1: We will first calculate the Need matrix
Need[i] = Max[i] − Allocation [i]
Process Need
ABC
P0 7,4,3
P1 1,2,2
P2 6,0,0
P3 0,1,1
P4 4,3,1
Chapter 7.indd 343 4/9/2015 10:01:15 AM

Step 1.2: Now, we will decide which process should execute first, that is, what are the processes that need resources
which are less than available.
Process Need Available

ABC ABC
P0 7,4,3 2,2,2
P1 1,2,2
So, we will start with P1 (because its resource need is less than the available resource) and after its completion all
the resources will be free; hence, after completion of P1 the available matrix contents will be
Available
222
+ 2 0 0 (resources held by P1)

422

Step 1.3: Again let us see the status

ABC ABC
P0 7,4,3 4,2,2
P1 Complete
P2 6,0,0
P3 0,1,1
P4 4,3,1
Now, we observe that P3 can be executed, and is considered to be executed.

Step 1.4: Again, let us see the status

ABC ABC
P0 7,4,3 6,3,3
P1 Complete
P2 6,0,0
P3 Complete
P4 4,3,1
So, P4 executes and similarly we can say that P2 and then P0 will be executed, and as a result, the available matrix
after execution of each process would be:
Available 6,3,3
P4 executes 0,0,2 (+)

6,3,5

After the execution of P2
Available 6,3,5
P2 executes 3,0,2

9,3,7

Chapter 7.indd 344 4/9/2015 10:01:15 AM

7.6 DEADLOCKS 345
After the execution of P0

Available 9,3,7
P0 executes 0,1,0

9,4,7

Thus, the final status of the system under consideration will be:

ABC ABC
P0 Complete 9,4,7
P1 Complete
P2 Complete
P3 Complete
P4 Complete
Therefore, the proposed sequence of execution is [P1, P3, P4, P2, P0]. Now, we will find whether this sequence of
execution is a `safe sequence’. We will now consider the resource request algorithm (the second portion) of the
Banker’s algorithm. In the beginning, we have:
Step 2.1:
Work = [2,2,2] (i.e., avail at the start)
Finish = [F, F, F, F, F] (the finish value of all processes is F meaning False)
So, P1 is executed first, and we observe that
Needi < Work: [1,2,2] < [2,2,2] holds good, and as a result we have
Work = [4,2,2]
Finish = [F,T,F,F,F] (the finish value of P1 is T meaning True)
Step 2.2:
Next, we have P3 executed and in this case the condition Needi < Work holds good, and as a result we have,
Work = [6,3,3]
Finish = [F,T,F,T,F]
After the execution of P4, we have
Step 2.3:
Work = [6,3,5]
Finish = [F,T,F,T,T]
Step 2.4:
Work = [9,3,7]
Finish = [F,T,T,T,T]
Step 2.5:
Work = [9,4,7]
Finish = [T,T,T,T,T]
In the whole process, we have never encountered a situation where the condition Needi < Work does not hold good.
Therefore, we can execute our processes in the order [P1, P3, P4, P2, P0] without encountering the deadlock.
Chapter 7.indd 345 4/9/2015 10:01:15 AM

7.6.6 Deadlock Detection and Recovery P1

Deadlock detection is a process of detecting whether a
deadlock has occurred. Deadlock occurs if neither avoid-
ance nor prevention has been employed. To detect a R1 R2 R3
deadlock, we simulate, in a recursive manner, the most P1
favoured execution of each unblocked process.
1. Any unblocked process can acquire the resources
needed for its execution. P3 P3
P2
2. All the acquired resources are released.
P2
3. These released resources can be used by the previ-
ously blocked processes, which are in need. R4
Figure 7.15 | A resource allocation graph and the

Repeat the steps 1 to 3 above. If any blocked process
remains, they are in a deadlock state.
corresponding wait-for graph.
7.6.6.1 Deadlock Detection
• Available: A vector of length m indicates the
To deal with this concept further, we can discuss the number of available resources of each type.
different types of resources available, namely, reusable • Allocation: An n × m matrix defines the num-
resources and consumable resources (Table 7.3). ber of resources of each type currently allocated to
Table 7.3 | Reusable vs. consumable resources

each process.
• Request: An n × m matrix defines the current
Reusable Resources Consumable Resources request of each process.
Number of units are Number of units may vary The algorithm is as follows:
constant • Let Work and Finish be vectors of length m and n,
Unit is either free or Process may create new respectively. Initially,
allocated; no sharing units Work = Available.
Process requests, Process may consume the • Find an index i such that both the following condi-
acquires and releases units tions are false:
the units
Finish[i] = False
Example: memory, Example: messages, signals Requesti ≤ Work. If no such i exists, go to step 4.
devices, files, tables
• Work = Work + Allocationi
Finish [i] = True.
We have already discussed about resource allocation
graphs (Section 7.6.3) and now we will discuss about Go to step 2.
deadlock detection in two different scenarios: • If Finish[i] = False, for some value of i, 1≤i≤n, then
1. Single instance of each resource type: If all the system is in deadlock. Moreover, if Finish[i] =
the resources have only a single instance, then we False, then it states that the process Pi is dead-
use a variant of resource allocation graph known locked.
as wait-for graph (Fig. 7.15). A deadlock exists
in the system if and only if the wait-for graph 7.6.6.2 Recovery from Deadlock
contains a cycle. To detect deadlocks, the system
maintains the wait-for graph and an algorithm is If at all a deadlock has been detected, any one of the
invoked, periodically in order to check for cycle in following techniques are used by the OS to recover from
the graph. the deadlock.
2. Single instance of each resource type: 1. Recovery by Process Termination: Once the
The wait-for graph scheme is not applicable to a deadlock has occurred, our approach is to overcome
resource system with multiple instances. The fol- the deadlock. This can be achieved if the processes
lowing algorithm is used for deadlock detection in that have been allocated the resources submit
such type of systems (somewhat similar to Banker’s them back to the resource pool. This submission
algorithm). is done in a one-by-one fashion until the deadlock
Chapter 7.indd 346 4/9/2015 10:01:16 AM

7.7 THREADS 347
is removed. Now, the question is that which pro-

cess should submit first or the order of submission. Code Date Files Code Data Files
Some of the possible parameters could be:
• The priority of a process: A process having
Registers Stack Registers Registers Registers
a low priority could be the first one to give up
the resources. Stack Stack Stack
• Number of resources held: A process holding
most of the resources could be the victim for ter-
mination or a process having the least number of
allocated resources be the first.
• Remaining time for completion: In this
approach, a process which needs more time for Single threaded process Multi-threaded process
Figure 7.16 | Threads and their types.

completion is terminated.
• Process type: Whether the processes are
interactive (which means that if we terminate
one process, all interactive processes will be ter- 7.7.2 Types of Threads
minated because it might be possible that one
process in execution needs some input from the A process can be composed of a single thread or multi-
process that was terminated, hence this will not threads. A thread (also known as lightweight process)
serve our purpose) or Batch processes (in this includes the program counter, stack pointer and its
case entire batch has to be taken care of). own data area. A thread executes sequentially and is
2. Recovery by checkpointing and rollback also interruptable. Whereas a process is composed of
(resource pre-emption): Another way to thread(s) and other system resources (data files, devices,
recover from deadlock is making use of checkpoints RAM etc). The concept of multithreading is used to
and rollback. These two strategies involve saving perform a number of independent tasks. There are two
the states of a running process at certain inter- types of threads, namely,
vals, so that they can be terminated and start their 1. User-level threads
execution from any of the saved states. The major 2. Kernel-level threads
overhead is extra memory and complexity to store
the intermediate states. User-Level Threads Kernel-Level Threads
Deadlock recovery is a high cost process, so is used when These are created by Kernel threads are
it is very rare and their might be a need for process user or programmer. created by operating
migration. Such a concept is more popular in databases. system.
OS cannot recognize These are recognized by
7.7 THREADS user threads. OS.
If thread is performing If thread is performing
I/O system call, then I/O system call, then
A process can have more than one thread. A light-weight
the entire process will other threads will
process is called thread. A process is generally known
be blocked. continue to execute.
as a single-threaded process. Multithreaded process has
more than one thread. User threads are Kernel threads are
dependent threads. independent threads.
A process has code, data and files. A thread shares code,
data and files with all other threads. A thread has its own It has less context- It has more context-
registers and stacks which are non-sharable (Fig. 7.16). switching time. switching time.
User threads do not Kernel threads require
7.7.1 Benefits of Threads
require hardware hardware support.
support.
1. Improved responsiveness
2. Faster context switching (context-switching time
for thread < context-switching time for process)
3. Resource sharing 7.7.2.1 fork() Function
4. Enhanced throughput of system
5. Effective utilization of multiprocessor systems The fork is a system call which is used to create the child
6. Economical to implement process. The fork() function returns a value of 0 (zero)
Chapter 7.indd 347 4/9/2015 10:01:16 AM

to the newly created child process. The fork() returns is allocated and when the process is complete it
a positive integer (process id of child process) to the releases the particular partition to be re-used by
parent process. The fork() returns a negative value if a another process.
child process creation is unsuccessful.
Note: If a program contains n fork() calls, then the
number of created child processes are (2n − 1). 0 70 K
1 40 K
2 110 K
Problem 7.3: Consider the following code and find
the total number of processes. 3 105 K
main() 4 250 K
{
fork (); 5 30 K
fork (); 6 80 K
fork ();
printf (“hello”); 7 120 K
}
Figure 7.17 | A fixed memory partition scheme.
Solution:
Number of child process = (2n − 1) = 23 − 1 = 7 = 7 2. Variable partition scheme: Variable-sized partition
is also known as dynamic partitioning (Fig. 7.18).
Total number of process = Parent process + Child
In this scheme, initially the memory will be full
process = (1 + 7) = 8
contiguous free block. Whenever it gets a request
from a process, accordingly partition is made.
The processes are accommodated in the memory
7.8 MEMORY MANAGEMENT according to their requirements and thus internal
fragmentation is overcome. But external fragmen-
tation exists.
The OS manages both the primary memory (RAM)
and the secondary memory (hard disk). In the follow- P1
ing section, we discuss various primary management P2
algorithms followed by secondary memory management
algorithms (hard disk scheduling algorithms). The OS P3
manages the memory in an efficient and orderly manner. P4
The OS has the following main responsibilities related to
the memory–Isolation of a process (non-interference), P5
Allocation and management (based upon an algorithm,
the user does not decide but knows where the content P6
is), Protection and access control (security) and reliable P7
long-term storage.
P8
7.8.1 Partitioning Intially After allocating

memory to processes
Figure 7.18 | A variable memory partition scheme.

It is a technique to divide memory into several parts
for process use. Basically, two partitioning methods are
used:
1. Fixed partition scheme: Memory is divided into 7.8.2 Partition Allocation Policies
several fixed-sized partitions (also known as static
partitioning) (Fig. 7.17). The partition can contain Whenever a process requests this resource, the allocation
only one process, therefore, the number of parti- of free memory (partitions or holes) is done based upon
tions is directly equal to the number of processes different strategies. There are four different strategies:
(concept of multiprogramming). So, whenever 1. First Fit: Allocation is made to the first free par-
a process needs this resource any free partition tition that is capable of meeting the requirements
Chapter 7.indd 348 4/9/2015 10:01:17 AM

7.8 MEMORY MANAGEMENT 349
of the process. The search can start from either end 4. Worst Fit: It is the exact opposite of best fit.
(up or down). The resource is allocated a partition that results in
2. Next Fit: A variation of first fit, the allocation higher free space. The idea behind this strategy is
is started from the previous partition that was that, a small process when allocated with a bigger
allocated. partition will be able to make it free in a lesser
3. Best Fit: The entire list of free partitions is time, this free partition may be used at a later
scanned and the partition that fits best, that is, stage by a process which might require a large par-
which results in a lesser free space after allocation tition; so that larger process does not have to wait
is chosen. for resource allocation.
Problem 7.4: Request from the processes is 300K, 25K, 125K and 50K, respectively.
150 K 350 K
The above request could be satisfied with? (Assume variable partition scheme)
(a) Best fit but not first fit
(b) First fit but not best fit
(c) Both best and first fit
(d) Neither best nor first fit
Solution:
We have to check for only two algorithms best fit and first fit (as given in question). Let us discuss by considering
one algorithm at a time, as follows:
(a) In Best Fit:
1. First of all, the 300K jobs will use 350K slot. Free space in the slot will be = 350 − 300 = 50K.
2. Next process of 25K will use remaining 50K slot (available as free from the operation above), as per the best
fit algorithm. So, the free space left = 50 − 25 = 25K.
3. Third process of 125K will be accommodated by 150K space, so we are left with 150 − 125 = 25K.
4. Fourth process requires 50K. Now, we have two slots of 25K left but our last process is of 50K size.
Therefore, the need of all the processes is not satisfied.
(b) In First Fit:
1. First process of 300K will use 350K slot. Remaining free space will be = 350 − 300 = 50K.
2. Next process of 25K will use remaining 150K slot (as per first fit algorithm). So, free space left = 150 − 25
= 125K slot.
3. Third process of 125K will be accommodating by 125K slot which is left free after the allocation of the
memory to second process.
4. Now, we have only one slot of 50K left and our process requires 50K memory.
Observing, the execution of both the algorithms, we can say that by using the first fit algorithm, all processes can
be executed. Therefore, option (b) is the correct answer.
7.8.3 Loading and Linking 7.8.3.1 Loading
A program may consist of different modules which are Loading is copying a program from secondary storage
to be `linked’ and then `loaded’ in the Main Memory for into main memory so it is ready to run. In some cases,
execution. An analogy to understand it could be that a loading just involves copying the data from disk to
program in C is first linked (as it makes use of different memory, in others it involves allocating storage, setting
header files) and then loaded in memory for execution. protection bits or arranging for virtual memory to map
It is hereby worth-mentioning that the image of the pro- virtual addresses to disk pages. Loading can be done
gram gets loaded into memory. through the following two ways:
Chapter 7.indd 349 4/9/2015 10:01:18 AM

1. Static Loading: It is the technique in which the 1. Compiler Time: If the compiler takes responsibility
entire program loads into memory before the start of associating a program and data to the actual physi-
of execution of the program. Some points of static cal memory location, it is called compile-time address
loading are as follows: binding. The compiler needs to interact with the
•• Inefficient utilization of memory operating system memory manager to perform com-
•• Faster program execution pile-time address binding. This type of address bind-
•• Use of static linking if static loading is used ing is done before loading the program into memory.
2. Load-time Address Binding: This type of
2. Dynamic Loading: It is the technique in which
address binding is performed after loading the pro-
the program is loaded into memory on demand.
gram into memory. The load-time address binding
Some points of dynamic loading are as follows:
•• Efficient utilization of memory
is performed by a loader.
•• Slower program execution
3. Run-Time (dynamic) Address Binding: In the
•• Use of dynamic linking if dynamic loading is used
dynamic address binding, address binding is post-
poned till the time of execution. Dynamic binding is
The majority of OS is used dynamic loading and dynamic performed by the processor. The majority of OS uses
linking. dynamic binding.
7.8.3.2 Linking (or Address Binding) 7.8.4 Paging
Association of program instruction and data to the actual Paging is one of the memory-management schemes
physical memory location is called as address binding. where a computer can store and retrieve data from sec-
ondary storage for use in the main memory (Fig. 7.19).
Instruction Address This technique maps processor-generated logical address
I1 20 to physical address. As we know,
I2 30 Logical address > Physical address
I3 40 There are few terms used in paging:
I4 50 1. Logical Address Space (LAS): Defines the size
of logical memory.
Address binding track to increment program counter. 2. Physical Address Space (PAS): Defines the
There are three types of address binding: size of physical memory.
13 bit 12 bit
13LA
bit
CPU LA 2 12 bit10
3 10
CPU 2 10
3P 10d g d
P d g d
00
0 00
01 11 (F)
01
011 12 11 (F)
10 (F) 01
011 23
10 P3
34 10 (F)
01 (F) 10 P3 10
45
10
56 01 (F) 11
67 00 (F)
11
7 00 (F)
Page table PAS
Page Page table PAS
numbers
Page
numbers
P= Page number bits
fP = Framenumber
Page numberbits
bits
d =
f = Frame number bitssize
Page size or frame
Figure 7.19 | A memory paging scheme.
d= Page size or frame size
Chapter 7.indd 350 4/9/2015 10:01:18 AM

3. Logical address (LA): Number of bits to repre-

Problem 7.5: Consider the following system:
sent logical address space.
4. Physical address (PA): Number of bits to rep- Number of pages = 2K
resent physical address space. Page size = 4K words
5. Page: Logical memory is divided into equal parts
Physical address = 18 bits
called pages.
6. Frame: Physical memory is divided into equal Calculate logical address space and number of frames.
parts called frames.
Solution:
Assume that processor-generated logical address (LA) is
13-bit long and physical address (PA) is 12-bit long, then 1. We know that
Logical address space = 213 words = 23 × 210 words LAS
Number of pages =
= 8K words Page size
Physical address space = 212 words = 22 × 210 words So, LAS = Number of pages × Page size
LAS = (2K × 4K) words = 211 × 212 words
= 4K words
The logical address is divided into equal size pages. In = 223 words = 8M words
our case, page size is given as 1K words. So,
PAS 218
2. Number of frames = = 12
LAS Frame size 2
Number of pages =
Page size Because frame size = page size, the number of
frames is 26 frames = 64 frames.
8K
Number of pages = =8
1K
Problem 7.6: Consider a system with LA = 32 bits,
To represent eight pages, we required 3 bits. So, logical PAS = 64 MB and page size is 4 KB. The memory
address is divided into two parts P (page number or is byte addressable. Page table entry size is 2 bytes.
number of bits required to represent the pages of LAS) What is the approximate page table size?
and d (page size or number of bits required to represent
the page size of LAS). This means LA is divided into Solution:
P = 3 and d =10 bits. LA = 32 bits. So, LAS = 232 bytes
The physical address is divided into equal-size frames. LAS 232
Generally, frame size = page size, given as 1K words. So, Number of pages = = 12 = 220
Page size 2
Number of frames =
PAS Page table size = Number of entries × Page table
Frame size entry size ⇒ 220 × 2 bytes = 2MB (because number
4K of page table entries = number of pages)
Number of pages = =4
1K
To represent four frames, we required 2 bits. So, physi 7.8.4.1 Paging Performance with TLB
cal address is divided into two parts f (frame number or
number of bits required to represent the frames of PAS) Translation lookaside buffer is a cache memory used
and d (frame size or number of bits required to represent by memory management unit to improve the transla-
the frame size of PAS). This means PA is divided into tion speed of virtual address. The use of virtual memory
f = 2 and d = 10 bits. becomes time consuming without TLB. This buffer keeps
the record of all the recently accessed pages.TLB works
as shown in Fig. 7.20.
Figure Description:
CPU first searches the page in TLB, if the page is
1. Whenever we apply paging, we have to maintain found in TLB then physical address is calculated. But if
a page table. there is TLB miss, then CPU goes to the page table. It
2. Number of entries in page table is same as the checks the page in page table, if found then it calculates
number of pages in LAS. the effective address and make the entry in TLB too.
3. The page table entries contain frame number But if page is not found in page table, it generates the
(binary format). page fault.
4. P is mapped to page table and corresponding The following points show how the performance is
entries go to PAS. improved by using TLB:
Chapter 7.indd 351 4/9/2015 10:01:20 AM

CPU p d f d
P
LA TLB 0 PA
miss 1
2
P7
3
4 P3
5 P2
TLB 6 P1
Page 7
numbers Physical
TLB Page table address space(PAS)
HIT
LA Logical address P = Page number

PA Physical address f = Frame number
TLB Trantition look-aside buffer d = Page size or
frame size
Figure 7.20 | TLB.
1. If page table is kept in the main memory and the 7.8.5 Multi-Level Paging
main memory access time is M, then formula for
effective memory access time for page table is Paging technique faces problem of large memory require-
EMAT = 2M ment due to the large logical address space. Multi-level
paging technique provides the solution to the problem
2. Translation look-aside buffer (TLB) is added to that occurred in paging technique (Fig. 7.21).
improve the performance of page table. TLB con-
tains the frequently referenced page number and Example:
corresponding frame number. If the TLB access Assume that processor-generated logical address (LA) is
time is C and TLB hit ratio is X, then effective 35-bit long and physical address (PA) is 26-bit long, then
memory access time for page table is Logical address space = 235 words
= 25 × 230 words
EMAT = X(C + M ) + (1 − X )(C + 2M )
= 32G words
Physical address space = 226 words
= 26 × 220 words
Problem 7.7: Consider a system with TLB access = 64M words
time 20 ns and main memory access time 100 ns.
Calculate the effective memory access time if TLB The logical address is divided into equal size pages. In
hit ratio is 95%. our case, page size is given as 4K words. So,
LAS 32G words
Solution: Number of pages = = = 8M pages
Page size 4K words
Given that C = 20 ns, M = 90 ns, X = 95%. Without
TLB effective memory access time LAS 32G words
Number of pages = = = 8M pages
Page size 4K words
EMAT = 2M = 2 × 100 = 200 ns
To represent 8M pages, we require 23 bits. So, the logi-
With TLB, effective memory access time cal address is divided into two parts P (page number or
number of bits required to represent the pages of LAS)
EEMAT =X
MAT = M))++((11−−X
X((CC++M X)()(CC++22M
M))
and d (page size or number of bits required to represent
==00.95
.95((20 100))++((11−−00.95
20++100 20++22××100
.95)()(20 100)) the page size of LAS). This means LA is divided into
==11
1144++11 11==125
125 nsns P = 23 and d = 12 bits. And P is further divided into
two parts P1 and P2.
Chapter 7.indd 352 4/9/2015 10:01:22 AM

CPU LA
P p d
1st level page table 2nd level page table

P1 P2 or outer page table or inner page table
0 0
1 (f) 1 f d
2 2 PA
3 3
(x) (f)
Physical address
space
n n
{ P1
P2
Number of bits required to represent pages of page table
Number of bits required to represent size of page of page table
{
Figure 7.21 | Multi-level paging scheme.
1. P1 is the page number or number of bits required table for every page, inverted paging is used (Fig. 7.22).
to represent the pages of page table. In inverted paging, only one table is maintained for all
2. P2 is the page size or number of bits required to
represent the page size of page table.
The physical address is divided into equal-size frames. Problem 7.8: Consider a system with logical address
Generally, frame size = page size, given as 4K words. So, = 34 bits, physical address = 29 bits and page size
= 16 KB. The memory is byte addressable and the
PAS 64M words page table entry size is 8 bytes.
Number of frames = = = 214 = 16K frames
Frame size 4K words (a) What is the conventional page table size?
PAS 64M words (b) What is the inverted page table size?
es = = = 214 = 16K frames
Frame size 4K words Solution:
To represent 16K frames, we required 14 bits. So, physi- (a) Conventional page table size
cal address is divided into two parts f (frame number or
number of bits required to represent the frames of PAS) LAS 234
and d (frame size or number of bits required to represent Number of pages = = 14 = 220
Page size 2
the frame size of PAS). Means PA is divided into f = 14
and d = 12 bits. Page table size = Number of pages × Page table
Problems associated with multilevel paging: entry size = 220 × 8 bytes × 8 B = 8 MB
Memory requirements are low for multilevel paging but
(b) Inverted page table size
address translation time is more due to more number of
levels. This can be solved by using Translation lookaside
PAS 229
buffer. Number of frames = = 14 = 215
Frame size 2
7.8.5.1 Inverted Paging Page table size = Number of frames × Page table
entry size
In paging technique, every process has its own page table = 215 × 8 bytes × 8 B
and because of this there is a problem of huge page table
= 256 KB
size. To overcome overhead of maintaining the page
Chapter 7.indd 353 4/9/2015 10:01:23 AM

CPU P d F d
LA
Page number Process - id

0 P0 Pri
1 P1 Pri PAS
2 P2 Pri
3 P0 Prj
4 P4 Prj
5 P2 Prk
6 P3 Prk
7 P4 Prk
Frame Inverted page table

number
Figure 7.22 | An inverted paging scheme.
the processes. The number of entries in the inverted page S is the segment number; it is equal to the number of
table is the same as number of frames in the PAS. bits required to represent the segment of LAS.
7.8.6 Segmentation D is the number of bits required to represent the seg-

ment size. It is also known as segment offset. So,
Paging technique does not follow user’s view of memory
allocation. To achieve user’s view of memory allocation, Number of segments in segment table
segmentation technique is used (Fig. 7.23). In this tech- = Number of segments in LAS
nique, logical address space is divided into several seg-
ments. Segments may vary in size. Segmentation also suffers from external fragmentation.
LA
CPU S D
D D
Base address Size 1500 S4

≤ +
Yes 2500
0 3600 300 3600 S0
3900 S1
1 3900 400 No 4300
2 7000 1000 5800 S5
Trap to OS 6300 S3
3 5800 200 6500 S2
7500
4 1500 1200
PAS
5 9000 600
Segment
6 12000 700
number
Segmentation table
Figure 7.23 | A memory segmentation scheme.
Chapter 7.indd 354 4/9/2015 10:01:24 AM

7.8.7 Virtual Memory 7.8.7.2 Effective Memory Access Time
Virtual memory is a memory management technique The time taken to service a page fault is called as page
that is implemented using both hardware and software. fault service time. If the main memory access time is M,
It maps memory addresses used by a program, called page fault service time is S, which is much larger than M
virtual addresses, into physical addresses in computer (S >> M) and page fault rate is P, then
memory. Virtual memory gives an illusion to the pro- Effective memory access time,
grammer that programs which are larger in size than
actual memory can be executed. Virtual memory can be EMAT = P × S + (1 − P ) × M
implemented with demand paging.
7.8.7.1 Demand Paging Problem 7.9: Consider a system with page fault ser-
vice time (S) = 100 ns, main memory access time
Demand paging is a method of virtual memory manage- (M) = 20 ns, and page fault rate (P) = 65%. Calculate
ment (Fig. 7.24). In a system that uses demand paging, effective memory access time.
the operating system copies a disk page into physical Solution:
memory only if an attempt is made to access it and
that page is not already in memory (i.e., if a page fault EMAT = P × S + (1 − P ) × M
occurs).
Figure Description:
= 0.65 × 100 + (1 − 0.65) × 20
Step 1: Processor is looking for page number 4. But = 0.65 × 100 + 0.35 × 20
page 4 is not available in the memory, that is, page fault. = 65 + 7 = 72 ns
Without page 4 currently executing, program will stop.
Step 2: The signal is send to OS that processor is look- Table 7.4 summarizes the different memory management
ing for page 4, which is currently not available in the techniques.
main memory.
Step 3: OS will search for required page 4 in LAS.
Step 4: OS will replace the page from LAS to PAS by 7.8.8 Page Replacement Algorithms
using the page replacement algorithm.
Whenever there is a demand of any page which is not in
Step 5: OS updates the page table accordingly. memory then that must be loaded. But if there is no free
Step 6: The signal is sent to the processor, which says frame in memory, one page is selected for replacement.
the required page is brought into the memory and then Page replacement algorithm decides which page is to be
the processor continues the program execution. replaced at the time of page fault.
3 Search for page 4

LA
CPU p d OS
0
1 2 Request
0 PA 1
1 F d 0 P7 2
6 2 P5
1 3
3 5 2 P Replace
3 P4 4
4 4
Update 3 P1 5
5
6 6
PAS
N — Nth step in the digram 7 Frame 7
number
Page table LAS
Figure 7.24 | A demand paging scheme.
Chapter 7.indd 355 4/9/2015 10:01:25 AM

Table 7.4 | Comparison of different memory management techniques

Technique Pros Cons
Fixed Partitioning Simple, No overhead on OS Internal fragmentation
Dynamic Partitioning No internal fragmentation, External fragmentation, Inefficient use of
Efficient use of RAM CPU
Paging No external fragmentation Very little internal fragmentation
Segmentation No internal fragmentation, External fragmentation
Less overhead as compared to dynamic
partitioning
Virtual Memory Paging No external fragmentation, Overhead on OS
High degree of multiprogramming
Virtual Memory No internal fragmentation, Extra overhead on OS
Segmentation Higher degree of multiprogramming,
Support for protection and sharing
7.8.8.1 First In, First Out 34

In this technique, the new page will replace the page 210
which is entered first in the memory frame. This means 02
that page is selected which has been in the main memory Number of hits: 2, 1, 1 = 3
for the longest time. Queue data structure is used.
Number of page faults = 1,2,0,3,1,2,4,0 = 8
Example 7.8 Number of hits 3

Hit ratio = =
Total page references 11
We have a reference string: 1, 2, 2, 1, 0, 3, 1, 2, 4, 1, 0
Reference String Problems with FIFO technique: FIFO suffers from
1 2 2 1 0 3 1 2 4 1 0 Belady’s anomaly. According to Belady’s anomaly,
sometimes with the increase in number of frames, page
1 1 1 11 3 3 3 4 4 4 fault rate also increases. The performance of the OS
2 2 22 2 1 1 1 1 0 drops instead going up (Fig. 7.25).
0 0 0 2 2 2 2
Miss Miss Hit Hit Miss Miss Miss Miss Miss Hit Miss
16
Number of page faults
Step 1: Suppose we have three empty frames. We insert 14

values in frames 1, 2. Then there is a hit for 2 and 1. 12
Because they are already present in frames, so the next 10
0 will be inserted in frame.
8
1 6
2 4
0
2
Step 2: When page 3 is referenced, there is no free
frame in memory; 3 will be replaced with the page which 1 2 3 4 5 6 7
is in memory for the longest time. That page is 1. Number of frames
1 3 Figure 7.25 | Belady’s anomaly.

2
0 7.8.8.2 Optimal Replacement Algorithm
Step 3: Next page referred is 1, which will be replaced In the case of optimal page replacement algorithm, page
with 2 and so on. will be replaced that will not be referenced in future.
Chapter 7.indd 356 4/9/2015 10:01:26 AM

Reference String Reference String

1 2 2 1 0 3 1 2 4 1 0 1 2 2 1 0 3 1 2 4 1 0
1 1 1 11 1 1 1 1 1 0 1 1 1 1 1 1 1 1 1 1 1
2 2 22 2 2 2 4 4 4 2 2 2 2 3 3 3 4 4 4
0 3 3 3 3 3 3 0 0 0 2 2 2 0
Miss Miss Hit Hit Miss Miss Hit Hit Miss Hit Miss Miss Miss Hit Hit Miss Miss Hit Miss Miss Hit Miss
Step 1: We have three empty frames. We insert values

in frames 1, 2. Then there is a hit for 2 and 1. Because Step 1: We have three empty frames. We insert values
they are already present in frames, so the next 0 will be in frames 1, 2. Then there is a hit for 2 and 1. They are
inserted in the frame. already present in frames. So, next 0 will be inserted in
frame.
1
1
2
2
0
0
Step 2: When page 3 is referenced, there is no free
frame in memory; 3 will be replaced with the page which Step 2: When page 3 is referenced, there is no free
will not be used for long. That page is 0. frame in memory. So, 3 will be replaced with the page
1 which is not recently used. That page is 2.
2 1
0 3 2 3
Step 3: Next page referred is 1, 2 will be hit. Now page 0
4 is demanded. No free frame is there in memory. So it
Step 3: Next page referred is 1, which will be hit. Now
will be replaced by either 2 or 3. Both are not required
page 2 is demanded. No free frame is there in memory.
in future. To decide among the two, FIFO is used. 2 will
So it will be replaced by 2, due to LRU page.
be replaced with 4.
1 1
2 4 3
3 0 2
Step 4: Next in reference string is 1. This is already Step 4: Next in reference string is 4. This will be
present, so it is a hit. Next is 0, 0 will also use FIFO to replaced with 1.
break the tie among 1, 4, 3. And 0 will be replaced with 1.
1
1 0 3 4
4 2
3
Step 5: Next in reference string is 1. This will be hit.
• When there is a tie for replacement in optimal page Next, page 0 will be replaced with 2.
replacement (OPR), FIFO is used to break the tie.
1
Number of page faults = 6
4
Number of hits = 5 2 0
• OPR is the most efficient algorithm. Lowest num- Number of hits = 4
ber of page faults occurs for this algorithm. Number of page faults = 7
• Realization of this algorithm is difficult, because
OS needs to know that how long it will take to • Counter is maintained after each page reference
refer that page again. with the last access time.
• Expensive.
7.8.8.3 Least Recently Used
Frame allocation: Each process needs minimum
Least recently used (LRU) page replacement algorithm number of pages. The allocation can be done in the
replaces the page which has not been used recently. following ways.
Chapter 7.indd 357 4/9/2015 10:01:26 AM

Fixed allocation: 7.8.8.5 Thrashing

• Equal allocation: Each process gets equal num-
ber of frames. For example, there are 80 frames and A program that causes page faults more frequently is
4 processes. So, each process will get 20 frames. said to be thrashing, which means that the system more
• Proportional allocation: This is a dynamic way busy in dealing with page faults instead of doing some
of frame allocation. Allocation is done according to constructive work. This leads to the decrease in through-
the size of process. put of the system.
s Further, it has been observed that there is a degree
Allocation for Pi = i × m
S of multiprogramming, which is optimal for any sys-
Here Pi: ith process tem’s performance, because if we increase the degree
of multiprogramming it might result in thrashing
si: size of ith process
(Fig. 7.26).
S: total size
m: total number of frames
Processor
Priority allocation: This method uses priority instead utilization
of size to allocate pages. Thrashing
7.8.8.4 Global versus Local Replacement
As discussed earlier, a page replacement algorithm

decides which page is to be replaced at the time of page
fault. Now, the questions that arise are:
1. Which page (that we have seen through different Degree of
page replacement algorithms, above), and multiprogramming
2. From where (to be discussed in the following Figure 7.26 | Thrashing in multiprogramming
section).
Global Replacement: The process can replace any
There are different remedies to overcome thrashing, one
page available in the main memory when page fault
of them is discussed below:
occurs. All the pages in the system are viewed as
common pool. Further, fixed allocation method cannot Working set model: According to Peter Denning,
be implemented using this technique. One of the major working set is a collection of pages that are being used
problems with this scheme is that each process will by some process during a finite time interval. This gives
spend more time in recovering lost pages, which is an idea about approximate set of pages the process will
called `thrashing’. require in future.
Local Replacement: When the page fault occurs,
local replacement algorithm can replace the pages from ∆= Working setwindow
the same process. Other processes will not suffer by this
scheme.
Example:
1. Local page replacement does not remove thrashing,
but it reduces it to some extent. Window size = 10
2. Working set model is solution of thrashing. As shown in the figure below, active pages are in set.
Working set:
{1,2,3,4}
2 6 1 5 7 7 7 7 5 1 6 2 3 4 1 2 3 4 4 4 3 4 3 4 4 4 1 3 2 3
D=10
Chapter 7.indd 358 4/9/2015 10:01:27 AM

7.9 FILE SYSTEM 359
7.9 FILE SYSTEM Such a type of file system is more popular in present day
OS. A path name is the path from the root, through all
the subdirectories, to a specified file. There are two types
7.9.1 Directory Structures of path names:
1. Absolute path name: It begins at the root and
The directory structures are used to manage and orga- follows a path down to the specified file, giving
nize data. This can be done in the following ways: directory names on the path. For example, root/
1. The disks can be split into one or more partitions. spell/mail/sys/units.
These partitions provide separate areas within one 2. Relative path name: It defines a path from the
disk and are treated as a separate storage devices. current directory. Considering the same example
2. Each partition contains information about files. as above, let us suppose that we are working in a
The device directory contains information such as current directory named root/spell/mail, then the
name, location, size and type. relative path can be sys/units, instead of specifying
the full path name.
7.9.1.1 Single-Level Directory
7.9.1.4 Acyclic Graph Directories
It is the simplest directory structure (Fig. 7.27), in which
there is one directory which contains all the files. Major An acyclic graph is a graph with no cycles and is a natural
limitation lies in the number of files (memory size) a generalization of the tree-structured directory scheme (Fig.
directory can hold and there is more than one user that 7.29). A tree structure prohibits the sharing of files or direc-
needs to access the files. tories, whereas an acyclic graph allows directories to have
shared subdirectories and file, which means the same file
7.9.1.2 Two-Level Directory or subdirectory may be in two different directories. With a
shared file, only one actual file exists, so any changes made
To overcome the above problems, a separate directory can
by one person are immediately visible to the other.
be created for each user, known as — User File Directory
(UFD). The major limitation is the sharing of files among
the users. This scheme is represented in Fig. 7.28. 7.9.2 Allocation Methods
7.9.1.3 Tree-Structured Directories The direct access nature of disks allows us flexibility in
the implementation of files and usually many files are
Directory structure is extended to a tree of arbitrary stored on the same disk. The main problem is how to
height. In this, users can create their own subdirectories allocate space to these files so that disk space is uti-
and manage their own files. There is a root directory too. lized effectively and files can be accessed quickly. Three
Directory cat bo a test data mail cont hex records
Files
Figure 7.27 | A single-level directory.
MFD User 1 User 2 User 3 User 4
UFD cat zo b test b data b test z data f
Figure 7.28 | Two-level directory structures.

Note: User 2 and User 4 have a file named data, thus name-collision problem is solved.
Chapter 7.indd 359 4/9/2015 10:01:28 AM

mail spell
list a b c c zzz dd
list e3 w4
Figure 7.29 | Acyclic graph directory structures (the sub directory list is shared by zzz and b).
widely used methods of allocating disk space are dis- and external fragmentation. It supports sequential and
cussed in the following text. random access.
7.9.2.2 Linked Allocation
7.9.2.1 Contiguous Allocation
In linked allocation, each data block contains the address
In contiguous allocation method, each file occupies a
of the next block in the file (Fig. 7.31). In contrast to con-
set of contiguous blocks on the disk. All the blocks
tiguous allocation, the disk blocks are scattered on disk.
are continuously kept together. There is an index that
The directory entry contains an index, which points to the
specifies the starting address and the length (size)
starting block as well as the ending block. Each block in
indicating the blocks (Fig. 7.30). Different allocation
turn contains an address of the next block, similarly, all the
strategies are first-fit, best-fit or worst-fit (already dis-
blocks are linked together. The last block does not have a
cussed earlier). This technique suffers from internal
0
File Start Length
5 Cnt 0 3
10 List 9 4
.
.
.
.
.
.
.
Figure 7.30 | Contiguous allocation of disk space.
Chapter 7.indd 360 4/9/2015 10:01:29 AM

7.9 FILE SYSTEM 361
Directory
0 File Start End
5 Cnt 7 31
10
30
35
Figure 7.31 | Linked allocation of disk space.
next block address (just as linked list). The major advan- dynamic access without any external fragmentation and
tage being, no external fragmentation with a limitation of there is very less loss of information in case of breaking
slow access (because we have to traverse all the blocks in of links. Let us take an example of a loss of pointer, in
the list before the required block). Another major drawback this case only information about that block will be lost
is the breaking of the link (similar to dangling references). whereas in linked allocation information stored in the
rest of blocks will also be lost.
7.9.2.3 Indexed Allocation It is hotly debated how large the index block should
be. Every file must have an index block, so we want the
There is an index block that contains the address of index block to be as small as possible. If the index block
every block that contains the information associated with is too small, however, it will not be able to hold enough
that file (Fig. 7.32). This approach is better as provides pointers for a large file, and a mechanism will have to be
Directory
File index block
1 Cut 15
7 9
16
1
10 10
15 13
-1
13 15 -1
-1
16
Figure 7.32 | Indexed allocation of disk space (-1 represents not used).
Chapter 7.indd 361 4/9/2015 10:01:30 AM

available to deal with this issue. The following are the • Number of disk block possible in one disk block =
possible solutions:  DBA size 

1. Linked scheme: An index block is normally of  DBA 
one disk space. Thus, it can be read and written
directly by itself. •• If asked for maximum file size, then calculate for
2. Multilevel index: The first index block contains highest power only such as Problem 7.10.
the set of pointers to secondary index blocks, which
in turn contain pointers to the actual data blocks. Problem 7.10: Consider the UNIX i-node which uses
3. Combined scheme: It is used in the UNIX File 12 direct DBAs, 1 single indirect, 1 double indirect, 1
System (UFS) i-nodes (see Fig. 7.33, which rep- triple indirect. The disk block address requires 32 bits
resents the UFS i-node scheme), where first few and the disk block size is 1 KB. What is the maxi-
data block pointers of the index block are stored mum file size?
directly in the i-node and then single, double and (a) 8 GB (b) 16 GB
triple indirect pointers provide access to more data (c) 32 GB (d) 64 GB
blocks as needed.
  DBA size 
Solution:  DBA size 2  DBA size 3 
Total size of file system = Number of direct DBA +       ×DB size

 DBA   DBA   DBA  DBA 3 
+ + +
   

  size  
  DBA size   DBA sizeMaximum
2  DBA file 3 =  DBA  × DB size
size
Total size of file system = Number of direct DBA +  +  +
size
+ ×DB size
  DBA   DBA   DBA  
  DBA size   1 KB 3 3  210 3
    1 KB =  2  × 210 bytes
3
  ×
 ×size    × DB size
DBA size
 2 
=size
  DBA file
Maximum file size =Maximum DB =
 DBA size   DBA size 2  DBA size 3  4 bytes
DBA 
f direct DBA +        ×DB
 DBA   DBA   DBA 
+ + + size
  1 KB 3 = 1610GB 3 3
   21 KB  210 3
=   × 1 KB = =  2  × 210×bytes   × 210 bytes
 
    1 KB =  2 
where DBA is disk block address. 4 bytes 2
4 bytes  2 
= 16 GB = 16 GB
Mode
Owners (2)
Timestamps (3)
data
Size block
data
Count data
Direct blocks
data
. data
Single indirect . data .
.
.
. data
data .
Double indirect
.
.
Triple indirect . data
.
.
data
Figure 7.33 | The UNIX i-node combined scheme.
Chapter 7.indd 362 4/9/2015 10:01:32 AM

7.10 I/O SYSTEMS 363
7.10 I/O SYSTEMS Problem 7.11: Consider a disk which has 16 plotters,
each plotter is divided into two surfaces. Each surface
has 1K tracks and each track has 512 sectors. Each
There are different devices that are connected with a
sector can store 2 KB data.
computer. These devices are broadly classified as:
(a) Find the capacity of the disk.
1. Machine readable: Most of the electronic equip- (b) How many bits are required to identify a sector?
ments such as hard disk, floppy disk, CD, USB, bar
Solution:
code reader, actuators, etc.
2. Human readable: Devices such as printer, key- (a) Capacity of disk = (Plotters × Surface × Tracks
board and Video Display Unit (VDU), usually × Sector × Sector Size) bytes
called monitor screen (but is different from moni- = (16 × 2 × 1024 × 512 × 2048) bytes
= (32 × 210 × 29 × 211) bytes
tor, as used in OS). Such devices transfer the data
at a slower rate as compared to the data transferred
between the memory and CPU. To overcome such = 32 × 23 bytes
differences in the rate of transfer, the concept of = 32 GB
SPOOLing (Simultaneous Peripheral Operations
(b) Number of bits required to identify a sector
OnLine) is used.
3. Communication: Network Interface Cards (NIC) = (Bits for plotter + bits for surface + bits for
are used for network communication or modems, tracks + bits for sectors)
etc. = (4 + 1 + 10 + 9) bits
There are different techniques used for I/O, such as = 24 bits
Programmed I/O, Interrupt driven I/O and DMA I/O.
These are covered in detail in Chapter 2.
7.10.1 Disk Structure 7.10.1.1 Time Related to Disk

Due to advent of technology, the gap in the access speeds 1. Seek time: The amount of time required to move
of a processor (as well as primary memory, i.e., RAM) the read/write head from its current position to
and disks (the secondary memory, i.e., hard disks) have desired track.
increased. In order to provide a faster access to user, it 2. Rotational latency: The amount of time to
is equally important for the OS to manage the disks effi- rotate the track when the read/write head comes
ciently by making use of suitable algorithms. A simple to desired sector position. In simple disk, rotational
disk structure is shown in Fig. 7.34. latency is the time to rotate 1/2 disk to the access.
Note: Number of cylinders per track = Number of 3. Transfer time: The amount of time taken to
tracks per surface transfer the required data is called as transfer time.
Spindle
Track
Read/write header
Upper surface
Lower surface
Plotter
Cylinders
Sector
Surface
Figure 7.34 | The disk structure.
Chapter 7.indd 363 4/9/2015 10:01:32 AM

Transfer time depends on the total size of the track

and the rotational rate of the disk. 8 1 8 1 6 1
4. Transfer Rate: The number of bytes transferred 7 2 4 5 3 4
in one unit time is called as the transfer rate of the
6 3 7 2 8 7
disk.
5 4 3 6 5 2
7.10.1.2 Disk Interleaving Max Single Double

interleaving interleaving interleaving
Figure 7.35 | Disk interleaving.
When the disk rotation speed is higher, then read/write
head skips sector 2 after reading sector 1 because it does
not get enough time to read the next sector. This problem
is solved by the disk interleaving technique (Fig. 7.35): Note:
1. Single interleaving: There is a gap between two 1. In single interleaving technique, to read a track,
consecutive sectors. two rotations are required.
2. Double interleaving: There are two gaps 2. In double interleaving technique, to read a track,
between two consecutive sectors. 2.75 rotations are required.
Problem 7.12: Consider a disk which has average seek time of 32 ns and rotational rate of 360 rpm (round per
minute). Each track of the disk has 512 sectors, each of size 512 bytes.
(a) What is the time taken to read four continuous sectors?
(b) What is the data transfer rate?
Solution:
(a) Time to read four continuous sectors = (Seek time + rotational latency + transfer time)
I. Seek time = 32 ns = 32 × 10−9 s
II. Rotational latency
⇒360 rotations took 60 s
1 rotation → (60/360) s
1/2 rotation → (60/360) × (1/2) ⇒ 0.083 s
III. Transfer time
⇒1 Track can be read in one rotation, and 1 track has 512 sectors each having 512 bytes.
So, (512 × 512) bytes data can be read in one rotation time (1/6) s.
256 KB data read time = (1/6) s
 1 
1 KB data read time = 
 6 × 256 
s
 1 
2 KB data read time = 2 × 
 6 × 256 
s
(1 sector size= 512 byte, so four sector = 2KB) = 0.0013 s

Time to read four continuous sectors = (32 × 10−9 + 0.083 + 0.0013) s = 0.0843 s
(b) Transfer rate = Number of bytes transferred in one unit time (1 s)
1/6 s → 256 KB
1 s → (256 × 6) KB = 1536 Kbps
Chapter 7.indd 364 4/9/2015 10:01:33 AM

Problem 7.13: Consider an interleaving disk where the track is divided into eight sectors, each of size 2K. Seek time
is 30 ms and rotation rate of disk is 3600 rpm. If half rotation is required to place the read/write head at the start
of the sector, then
(a) How much time is required reading all eight sectors of the track using double-interleaving disk?
(b) What is the time difference to read all eight sectors if it is a non-interleaving disk?
(c) What is the data transfer rate using double-interleaving disk?
Solution:
(a) Time to read eight continuous sectors = (Seek time + rotational latency + transfer time)
I. Seek time = 30 ms
II. Rotational latency = (1/2) rotation
3600 rotations took 60 s
1 rotation → (60/3600) = 16.66 ms
1/2 rotation → (16.66) × (1/2) = 0.0083 = 8.33 ms
III. Transfer time
We have to read all of the eight sectors mean one whole track. 1 Track can be read in 2.75 rotations (double
interleaving).
So, transfer time of eight sectors = Transfer time of 1 track = 2.75 Rotation time
Time to read eight continuous sectors = (Seek time + rotational latency + transfer time)
= (Seek time + (1/2) rotation time + 2.75 rotation time)
= (Seek time + 3.25 rotation time)
={(30 + 3.25 × (16.66)} = 84.14 ms
(b) Time to read eight continuous sectors = (Seek time + rotational latency + transfer time)
I. Seek time = 30 ms
II. Rotational latency ⇒ 8.33 ms
III. Transfer time
We have to read all of the eight sectors means one whole track.
1 Track can be read in 1 rotation (non-interleaving).
So, transfer time of eight sectors = Transfer time of 1 track = 1 Rotation time
Time to read eight continuous sectors = (Seek time + rotational latency + transfer time)
= (Seek time + (1/2) rotation time + 1 rotation time)
= (Seek time + 1.5 rotation time)
= {(30 + 1.5 × (16.66)} = 55 ms
Time difference between double interleaving and non-interleaving = (84.14 − 55)
= 29.14 ms
(c) Transfer rate = Number of bytes transferred in one unit time (1 s)
2.75 Rotation time → 16 KB
60 s = 3600 rotations
1 s = 3600/60 = 60 rotations
2.75 rotation = 16 KB data
1 rotation = (16/2.75) KB
60 rotations = {(16/2.75) × 60} KB ≈ 349 KB
So, transfer rate ≈ 349 Kbps
Chapter 7.indd 365 4/9/2015 10:01:33 AM

7.10.2 Disk Scheduling Algorithms 7.10.2.1 FCFS Scheduling
One of the responsibilities of the operating system is to The simplest form of the disk scheduling is, of course,
use the hardware efficiently, which means that we should the first-come, first-served (FCFS) algorithm. This algo-
have fast disk access time and a broader disk bandwidth. rithm is intrinsically fair, but it generally does not pro-
Whenever a process needs I/O to or from the disk, it vide the fastest service.
issues a system call to the operating system. The request
specifies several pieces of information, such as:
7.10.2.2 SSTF Scheduling
1. Whether this operation is input or output?
2. What the disk address for the transfer is? A variation of shortest job first (SJF), applied in
3. What the memory address for the transfer is? CPU scheduling. It serves the request that has a
4. What the number of bytes to be transferred is? minimum seek time from the current head position.
If the desired disk drive and controller are available, To calculate the minimum seek time, we look for the
the request can be serviced immediately. If the driver or minimum number of cylinders to be traversed. The
controller is busy, any new requests for service will be major limitation is – starvation, as encountered in
placed on the queue of pending requests for that device. SJF algorithm.
We require disk-scheduling algorithms to service these
pending requests.
Problem 7.15: Consider, the same problem, a disk

Problem 7.14: Consider, for example, a disk queue queue with requests for I/O to blocks on cylinders
with requests for I/O to blocks on cylinders
88, 83, 7, 13, 90, 79, 11, 63
88, 83, 7, 13, 90, 79, 11, 63
The disk head is initially at cylinder 53.
The disk head is initially at cylinder 53.
Solution: Solution:
The head will first move from 53 to 88 (being the first The head is positioned at cylinder 53, the minimum
request), then to 83, 7, 13, 90, 79, 11, and finally to distance it has to move is 10 when it moves to cyl-
63, for a total head movement of 339 cylinders. This inder 63, which is minimum, then from 63 the next
schedule is shown in Fig. 7.36. access will be 79, then 83, to 88, then to 90, to 13,
11 and finally 7. This schedule is shown in Fig. 7.37.
0 25 50 75 99
0 25 50 75 99
53 88
7 83 53
63 79
13 83
90
88
79 90
13
11
63
11
7
Figure 7.36 | FCFS disk-scheduling algorithm. Figure 7.37 | SSTF disk-scheduling algorithm.
Total head movement = (53~88) + (88~83) +
(83~7) + (13~7) + (99~13) + (79~90) + (11~79) Total head movement = (53~63) + (63~79) +
+ (11~63) (79~83) + (83~88) + (88~90) + (90~13) + (13~11)
= 35 + 5 + 76 + 6 + 86 + 11 + 68 + 52 + (11~7)
= 339 cylinders = 10 + 4 + 4 + 5 + 2 + 77 + 2 + 4 = 108 cylinders
Chapter 7.indd 366 4/9/2015 10:01:34 AM

7.10.2.3 SCAN Scheduling is that the requests are served in any one direction of
the motion of the head, not in both the directions as in
Also, called the elevator algorithm, here the disk arm SCAN. This can be compared by looking at the following
begins serving the requests at one end and moves to numerical problem.
the other end (it does not come back, as in earlier algo-
rithms) and then returns back serving the requests, simi-
lar to an elevator. This technique eventually results in Problem 7.17: Consider, the same problem, a disk
reduced variance as compared to SSTF. queue with requests for I/O to blocks on cylinders
88, 83, 7, 13, 90, 79, 11, 63
Problem 7.16: Consider, the same previous problem,
a disk queue with requests for I/O to block on The disk head is initially at cylinder 53 and the disk
cylinders arm is moving towards 0.
88, 83, 7, 13, 90, 79, 11, 63 Solution:

The disk head is initially at cylinder 53 and the disk The head is positioned at cylinder 53 and the disk
arm is moving towards 0. arm is moving towards 0. So, the cylinders accessed
will be 13,11 and 7, the disk arm will touch 0 and
Solution: then will return to the end (in this case) and start
The head is positioned at cylinder 53 and the disk reading 90, 88, 83, 79, and finally read 63. This sched-
arm is moving towards 0. So, the cylinders accessed ule is shown in Fig. 7.39.
will be 12, 11 and 7, the disk arm will touch 0 and
then will read 63, then 79, 83, 88 and finally 90. This 0 25 50 75 99
schedule is shown in Fig. 7.38.
0 25 50 75 99 53
53 13
11
7
13
11
7 90
88
83
63
63 79
70
83
88 Figure 7.39 | C-SCAN disk-scheduling algorithm.
90
Figure 7.38 | SCANdisk-scheduling algorithm. (11~7) + (7~0) + (0~99) + (99~90) + (90~88) +
(88~83) + (83~79) + (79~63)
(11~7) + (7~0) + (0~63) + (63~79) + (79~83) + = 40 + 2 + 4 + 7 + 99 + 9 + 2 + 5 + 4 + 16
(83~88) + (88~90) = 188 cylinders
= 40 + 2 + 4 + 7 + 63 + 16 + 4 + 5 + 2
= 143 cylinders
7.10.2.5 LOOK and C-LOOK Scheduling
It should take into account the head movement from
cylinder 7 to 0 and 0 to 63 also. LOOK and C-LOOK are the variants of SCAN and
C-SCAN algorithm, respectively. Here the head does not
7.10.2.4 C-SCAN Scheduling move till the end of the cylinder, it only moves to the
farthest cylinder which requires to be served. After serv-
This algorithm treats the cylinders as a circular list, ing the farthest request, it changes the direction (in case
hence provides a uniform wait time as compared to the of LOOK) or jumps to another end (in case of C-LOOK).
SCAN algorithm discussed above. The major difference The same is illustrated below.
Chapter 7.indd 367 4/9/2015 10:01:34 AM

Problem 7.18: Consider, the same problem, a disk 0 25 50 75 99

queue with requests for I/O to blocks on cylinders
88, 83, 7, 13, 90, 79, 11, 63 53
13
The disk head is initially at cylinder 53 and the disk
11
arm is moving towards 0.
Solution: 7
The head is positioned at cylinder 53 and the disk
arm is moving towards 0. So, the cylinders accessed 90
will be 13, 11 and 7, now the disk arm will not touch 88
0 (as in SCAN algorithm) but will read 63, then
79, 83, 88 and finally 90. This schedule is shown in 83
Fig. 7.40. 79
63
0 25 50 75 99 Figure 7.41 | C-LOOK disk-scheduling algorithm.
53 Total head movement = (53~13) + (13~11) +

(11~7) + (7~90) + (90~88) + (88~83) + (83~79)
13 + (79~63)
11 = 40 + 2 + 4 + 83 + 2 + 5 + 4 + 16 = 156 cylinder
7
63
79 7.10.2.6 Selection of a Disk-Scheduling
83 Algorithm
88
90 Given so many disk-scheduling algorithms, how do we
Figure 7.40 | LOOK disk-scheduling algorithm.

choose the best one? SSTF is common and has a natu-
ral appeal because it increases performance over FCFS.
SCAN and C-SCAN perform better for system that place
a heavy load on the disk, because they are less likely to
(11~7) + (7~63) + (63~79) + (79~83) + (83~88)
have starvation problem. Therefore, it is dependent on
+ (88~90)
the type of application we want to use.
= 40 + 2 + 4 + 56 + 16 + 4 + 5 + 2 = 129 cylinder
7.11 PROTECTION AND SECURITY

Problem 7.19: Consider, the same example, a disk
queue with requests for I/O to blocks on cylinders As huge information is stored in computer systems, the
need to protect it is equally important. Usually, security
88, 83, 7, 13, 90, 79, 11, 63
and protection are used interchangeably.
The disk head is initially at cylinder 53 and the disk
arm is moving towards 0. Simulate the C-LOOK disk- 7.11.1 Security
scheduling algorithm.
There are three important aspects, known as CIA triad,
Solution:
which are basic to providing security. These are:
The head is positioned at cylinder 53 and the disk
1. Confidentiality: It involves data confidentiality
arm is moving towards 0. So, the cylinders accessed
as well as user privacy.
will be 13,11 and 7, the disk arm will not touch 0 but
2. Integrity: It involves two terms–data integrity
will return to 90, then read 88, 83, 79 and finally read
(which means that data is handled in an autho-
63. This schedule is shown in Fig. 7.41.
rized manner) and system integrity (which means
Chapter 7.indd 368 4/9/2015 10:01:35 AM

7.11 PROTECTION AND SECURITY 369
that the system is free from unauthorised access or àdversaries’, and are capable of deploying active attacks
manipulation). as well as passive attacks.
3. Availability: It means the timely access to user
(for which he has the rights). 7.11.2.3 Accidental Data Loss
7.11.2 Security Environment Malicious intruders cause threats to the system. Hence,
valuable data can be lost by accident. The common
Security has many facets. Three of the most important causes of accidental data loss are as follows:
ones are the nature of the threats, the nature of intrud- 1. Hardware or software errors: CPU malfunc-
ers and accidental data loss. tions, unreadable disks or tapes, program bugs, etc.
There are two different kinds of attack. A `Passive 2. Human errors: Incorrect data entry, wrong tape
Attack’ does not affect the system resources but observes or disk mounted, wrong program run, lost disk or
the various acts of the system (such as — trying to know tape, etc.
the password by looking over the shoulders of a user), 3. Natural disaster: Fires, floods, earthquakes, wars,
whereas in an Àctive Attack’ the system resources are riots, or rats gnawing tapes or floppy disks, etc.
affected (such as — an unauthorised person deleting a file).
7.11.3 Cryptography
7.11.2.1 Threats
First we will discuss about cryptography, in brief, to
From a security perspective, computer systems have understand more about security.
three general goals, with corresponding threats to them
(see Table 7.5): The aim of cryptography is to encrypt plaintext (mes-
Table 7.5 | Goal and threats.

sage) into cipher text in such a way that only the autho-
rized person or user can decrypt it back to plaintext.
Goal Threat In symmetric cryptography (or private key cryptogra-
phy), one makes use of a single key. This key is used for
Data confidentiality Exposure of data encryption at the sender side and the same key is used
Data integrity Tempering with data at the receiver side to decrypt the message.
System availability Denial of Service (DOS) Whereas, in asymmetric cryptography (or public key
cryptography), two keys are used in which one is private
7.11.2.2 Intruders and the other one is public. The public key is used to
encrypt the message, whereas the private key, which is
In the securityperspective, intruders are defined as unau- only known to the receiver, is used for decryption.
thorized people, who are entering into the system they The encryption process is illustrated in the Fig. 7.42
do not have business with. Intruders are also known as below:
Decryption key Encryption key

KE KD
C = E(P, KE) P =D(C, KD)

P E D P
Plaintext in Cipher-text Plaintext out
Encryption Decryption
Algorithm Algorithm
Encryption Decryption
Figure 7.42 | Mechanism of cryptography.
Chapter 7.indd 369 4/9/2015 10:01:35 AM

7.11.4 Attacks from within the System Viruses can be divided into different classes based
upon the following characteristics:
When an adversary enters into a system, he or she can 1. Environment
start doing damage. The damage is of the following 2. Operating system (OS)
types: 3. Destructive capabilities
1. Trojan horses: Trojan horse is a program, which
contains a code hidden within a block of code, used to (Not to forget: there also exist other `harmful’ programs
modify/delete/encrypt the user’s data (functioning is or so called `malware’, such as Trojan horses).
similar to the Trojan horse, we read in history). The 1. Environment: Viruses can be described as:
Trojan horse cannot be differentiated from a normal •• File viruses: Most common virus that infects
code of program by reading the source code of the the executable (*.exe) files.
program housing the Trojan horse. •• Boot viruses: The virus infects the boot sector
2. Login spoofing: Related to Trojan horse, it is a program, the contents are replaced by infected
technique to obtain a user’s password. The user is code. Nowadays, modern OS are capable of
provided with the same lookalike page, as his login thwarting the attack by such kind of virus.
and he is unable to differentiate between the actual •• Macro viruses: Macros are defined as a group
page and this page; so he enters his login password of lines, technically speaking a subroutine,
and the vital information does not remain private. which is capable of infecting documents like —
3. Logic bombs: Similar to a bomb, this code explodes MSOffice, that make use of OLE (Object Link-
(causes destruction) when certain condition(s) are ing and Embedding) formats.
met like — a particular day of time or a particular •• Network viruses: These viruses make use of
date or when the user is running a particular file. protocols and commands of computer network
Such a code is embedded in a legitimate program, as or email to spread themselves in the system.
in a Trojan horse. These codes can modify or delete a
There is a large number of combinations possible,
file or can even cause a system to halt or do any other
for example, file-boot viruses infecting both files
damage.
and boot sectors on the disks. Another example of
4. Trap doors: These are also small programs embed-
the combo — network macrovirus, not only infect
ded in the programs to quickly gain access at a later
the documents which are being edited, but also
time. These are also known as `backdoors’. It is very
send copies of itself by email.
difficult to block these backdoors by an OS.
5. Buffer overflow: It is a program written to block 2. Mode of Operation: Viruses can be described
memory or buffer. It allows adversary to modify por- as follows:
tions of the target process address space. •• TSR: A terminate-and-stay-resident (TSR)
virus infects a computer and remains in RAM
and then performs an action for which it was
7.11.5 Attacks from Outside the System
planted. The virus becomes a part of the OS
In the previous sections, we have discussed about the and remains active always. In contrast a non-
threats largely caused by users already (logged) in the resident virus is active for a limited time only.
system; however, for machines connected to the network, •• Stealth algorithm: These viruses do not
there are external threats too. reveal their identity.
•• Self-encrypting and polymorphic: Such
viruses are not detected by antivirus as they
7.11.5.1 Virus Damage Scenarios do not have a fixed signature. They are able
to change themselves with every infection, and
A virus is a program that can do anything a program can
therefore, are very hard to detect.
do. It replicates itself, infects user’s data or files, stops
the operation of the system or functioning of the system. 3. Destructive capabilities: Viruses can be
It gets its name because of the similarities in its func- described as follows:
tionality to a biological virus. The common acronym for •• Harmless: Do not affect the computing
virus is Vital Information and Resources Under Siege. resources.
Nowadays, there are certain programs known as `Kit’, •• Not dangerous: Have a limited effect on com-
capable of generating new viruses automatically. Some puting resources. For example, a certain sound
of the well-known anti-virus softwares are: Symantec, can be produced after the computer is ON for
McAfee, AVG, Kaspersky, etc. say 2 hours.
Chapter 7.indd 370 4/9/2015 10:01:35 AM

•• Dangerous: They disrupt the normal work- 2. Identify: Identify the virus (may be your antivi-
ing of computing resources, such as you might rus scanner might tell you).
not be able to read from the USB port of the 3. Remove: All the traces of the virus are removed
computer. (disinfected by the antivirus scanner).
•• Very dangerous: Certain viruses result in
loss of data, such as erasing system files which The antivirus scanners use different approaches to
required when a system is to be booted, so as a protect the system, some of these approaches are:
result, the system does not work (boot). 1. Signature scanning: Each virus has a certain
signature (a code), which is located and deleted.
2. Heuristic scanning: Makes use of heuristics
7.11.6 Anti Virus Approaches
(rules of thumb or a guess) based upon certain
analysis.
The best approach is the protection against the virus. If
3. Generic decryption: The virus decrypts itself
at all the system has been infected by virus, the follow-
after the execution of code.
ing steps are commonly used:
4. Behaviour blocking: Its function is similar to
1. Detect: Locate it (by checking the file size of vital a firewall, it blocks the execution of a suspicious
files). piece of code.
IMPORTANT FORMULAS
1. Turnaround Time: TAT = CT − AT [(CT: com- 12. EMAT = X(C + M ) + (1 − X )(C + 2M )

= P × S + (1 − P ) × M
pletion time, AT: arrival time)]
13. Effective memory access time (E (EMAT
MAT)) =
2. Waiting Time: WT = TAT − BT [(BT: (EMAT ) = P × S + (1 − P ) × M
burst
time)]
  DBA size 
3. Waiting Time for Round Robin: (n − 1)q + nc 14. Total size of file system = Number of direct DBA + 
  DBA 


4. Complexity of Safe Sequence using Banker’s   DBA size   DBA size 2  DBA s
Total size of file system = Number of direct DBA +    +
Algorithm: m × n2(n: processes, m: resources)  DBA   DBA   DBA
+


5. To Avoid Deadlock:
• Total need < m + n   DBA size   DBA size 2  DBA size 3 
Total size of file system = Number of direct DBA +       ×DB size
should be between 1 DBA   DBA   DBA 
+ + +

• Max need of each process 
 
and m resources
w+s 15. Number of disk block possible in one disk block =
6. Response Ratio = (w = waiting time, s =
s  DBA size 
 
 DBA 
service time or burst time)
7. Thread maintains different registers and stack but
shares code, data and files. 16. Number of cylinder per track = Number of tracks
per surface
8. Need[I ] = Max[I ] − Allocation[I ]
17. Capacity in disk = (Plotters × Surface × Tracks ×
9. If program contains n fork() call, then the number
Sector × Sector size) bytes
of created child process are (2 − 1).
n
18. Number of bits required to identify a sector = (Bits

LAS
10. Number of pages = for plotter + bits for surface + bits for tracks +
Page size bits for sectors)
PAS 19. Time to read four continuous sectors = (Seek time
11. Number of frames =
Frame size + rotational latency + transfer time)
Chapter 7.indd 371 4/9/2015 10:01:37 AM

SOLVED EXAMPLES
1. The term thrashing is used to define (c) RR scheduling.

(d) highest penalty ratio next.
(a) A reduce page I/O
(b) A decreased degree of multiprogramming Solution: It uses fair pre-emptive scheduling.
(c) An excessive page I/O Ans. (b)
(d) Improvement(s) in the system performance
6. In modern operating systems such as Windows
Solution: Thrashing means that the system more 2000, all the processor-dependent code is isolated
busy in dealing with page faults instead of doing in a dynamic link library called
some constructive work. Thus, it leads to excessive
page I/O. (a) NTFS file system
Ans. (c) (b) Hardware abstraction layer
(c) Microkernel
2. Page fault occurs when (d) CORBA Process Manager
(a) The page is not in cache memory. Solution: The hardware abstraction layer (HAL)
(b) The page is in main memory. provides logical link between NT-based operating
(c) The page is not in main memory. systems and physical hardware of the computer.
(d) The page has an address, which cannot be Ans. (b)
loaded.
7. Translation look-aside buffer (TLB) is
Solution: If page is not found in page table, it
generates the page fault. (a) A cache-memory in which item to be searched
Ans. (c) is compared one by one with the keys.
(b) A cache-memory in which item to be searched
3. If m = number of resources, n = number of processes
is compared with all the keys simultaneously.
in the system, then Banker’s algorithm although
(c) A main memory in which item to be searched
a general algorithm may require
is compared one by one with the keys.
operations.
(d) An associative memory in which item to
(a) n × n × m be searched is compared with all the keys
(b) (m × m) × n simultaneously.
(c) (m × n) × n
Solution: TLB is a cache memory used by
(d) (n × m) × m
memory management unit to improve the transla-
Solution: Complexity of Safe Sequence using tion speed of virtual address.
Banker’s Algorithm is m × n2 Ans. (d)
where n is the number of processes, and m is the
number of resources. 8. Windows 2000 operating system include the follow-
Ans. (a, c) ing process states:
4. Mutual exclusion problem occurs (a) Ready, running and waiting
(b) Ready, standby, running, waiting, transition
(a) Between two disjoint processes that do not and terminated
interact. (c) Ready, running, waiting, transition, sleep and
(b) Among processes that share resources. terminated
(c) Among processes that do not use the same (d) Standby, running, transition, sleep and
resource. terminated
(d) Between two processes that uses different
resources of different machines. Solution: It includes ready, standby, running,
waiting, transition and terminated states.
Solution: In this problem, if process Pi is execut-
ing in its critical section, then no other process can Ans. (b)
execute their critical sections. 9. A memory management algorithm, realizing vir-
Ans. (b) tual memory, partially swaps out a process. This is
5. Linux operating system uses similar to which kind of CPU scheduling.
(a) i-Node scheduling. (a) Short-term scheduling
(b) fair pre-emptive scheduling. (b) Long-term scheduling
Chapter 7.indd 372 4/9/2015 10:01:37 AM

SOLVED EXAMPLES 373
(c) Medium-term scheduling 14. The P and V operations on counting semaphores,

(d) Mutual exclusion where s is a counting semaphore are defined as
Solution: Medium-term scheduling is one in which follows:
decision is taken to add the number of processes P(s):s = s − 1
that are present either partially or fully in the main if s < 0 then wait;
memory. This is similar to CPU scheduling. In short- V(s):s = s + 1
term scheduling the available process is executed by if s = 0 then wake up a process waiting on s.
the processor. In long-term scheduling, the process Assume that Pb and Vb are the wait and signal
is added to the pool of processes to be executed. operation on binary semaphore. Two binary sema-
Ans. (c) phores Xb and Yb are used to implement the sema-
phore operation P(s) and V(s) as follows:
10. The working set model is used in memory manage-
ment to implement the concept of P(s): Pb(Xb);
(a) Segmentation (b) Principle of locality s=s−1
(c) Paging (d) Thrashing if(s<0){
Vb(Xb);
Solution: Fact Pb(Yb);
Ans. (b) }
11. Which scheduling algorithm gives a minimum else Vb(Xb);
average waiting time? V(s): Pb(Yb);
s= s +1
(a) RR (b) SJF
if(s<=0)
(c) FCFS (d) Priority
Vb(Yb);
Solution: Shortest job first algorithm.
Ans. (b) The initial value of Xb and Yb are, respectively,
12. If the disk is rotating at 3600 rpm, determine the (a) 0 and 0 (b) 0 and 1
effective data transfer rate which is defined as the (c) 1 and 0 (d) 1 and 1
number of bytes transferred per second between Solution: From the give code, we get P(s) and
disk and memory. (Given size of track = 512 bytes) V(s) decrements and increment the value of sema-
(a) 20Kbps (b) 30 Kbps phore, respectively. So, from the given conditions,
(c) 40Kbps (d) 50Kbps we conclude that to avoid mutual exclusion the
value of Xb should be 1 and that of Yb should
Solution: 3600 rotation → 60 s
be 0.
1 rotation → 60/3600 = 1/60 s Ans. (c)
Transfer rate = (60 × track size) per second 15. Consider the following table of arrival time and
= 60 × 512 bytes/s burst time for three processes P0, P1 and P2.
= 30 Kbps
Ans. (b)
Process Arrival Time Burst Time
13. Which of the following features will characterize an
OS as multiprogrammed OS? P0 0 9
1. More than one program may be loaded into P1 1 4
main memory at the same time.
2. If a program waits for certain event, another P2 2 9
program is immediately scheduled.
3. If the execution of a program terminates, The pre-emptive shortest job first scheduling
another program is immediately scheduled. algorithm is used. Scheduling is carried out only
(a) Only at arrival or completion of process. The average
(b) (1) and (2) only waiting time for the three processes is ———.
(c) (1) and (3) only Solution: Waiting time for P0 = 5 − 1 = 4 s
(d) (1), (2) and (3) only Waiting time for P1 = 1 − 1 = 0 s
Solution: Multiprogrammed OS performs all the Waiting time for P2 = 13 − 2 = 4 s
given three tasks. Average waiting time = 4 + 0 + 11/3 = 5 s
Ans. (d) Ans. (5)
Chapter 7.indd 373 4/9/2015 10:01:37 AM

16. A CPU-scheduling algorithm determines an order (c) k2

for the execution of its scheduled processes. Given
k processes to be scheduled on one processor, how (d) k !
many different schedules are possible? Give a for-
mula in terms of k. Solution: k! (k factorial = k × (k − 1) × (k − 2) × × 2 × 1
k factorial = k × (k − 1) × (k − 2) × × 2 × 1 )
(a) k
Ans. (d)
1
(b)
k
1. Using a larger block size in a fixed block size file following scheduling strategies will result in the
system leads to least CPU utilization (over a long period of time)
for this system?
(a) better disk throughput but poorer disk space
utilization. (a) First come first served scheduling
(b) better disk throughput and better disk space (b) Shortest remaining time first scheduling
utilization. (c) Static priority scheduling with different priori-
(c) poorer disk throughput but better disk space ties for the two processes
utilization. (d) Round robin scheduling with a time quantum
(d) poorer disk throughput and poorer disk space of 5 ms
utilization. (GATE 2003: 2 Marks)
(GATE 2003: 1 Mark)
Solution: Round robin will result in least cpu uti-
Solution: Larger block size in a fixed block size lization for time slice = 5 ms
file system reduces the number of disk access, but Consider process A and B, both process have 10 ms
at the same time increases chances of internal cpu utilization and then 90 ms I/O. Sequence will
fragmentation. be ABAB. After 15 ms process, A can do I/O and
Ans. (a) I/O can be done in parallel; therefore, A finishes
2. In a system with 32-bit virtual addresses and 1-KB I/O at 105th ms and B finishes I/O at 110th ms so
page size, use of one-level page tables for virtual cpu remains idle from 20th to 105th ms leading to
to physical address translation is not practical a total of 85 ms.
because of Ans. (d)
(a) the large amount of internal fragmentation. 4. Which of the following will always leads to an
(b) the large amount of external fragmentation. output starting with `001100110011’?
(c) the large memory overhead in maintaining page (a) P (S) at W, V(S) at X, P(T) at Y, V(T) at Z, S
tables. and T initially 1
(d) the large computation overhead in the transla- (b) P(S) at W, V(T) at X, P(T) at Y, V(S) at Z, S
tion process. initially 1, and T initially 0
(GATE 2003: 1 Mark) (c) P(S) at W, V(T) at X, P(T) at Y, V(S) at Z, S
Solution: Page size = 210 bytes and T initially 1
Total number of pages required = 232/210 = 222, (d) P(S) at W, V(S) at X, P(T) at Y, V(T) at Z, S
which is very large. initially 1, and T initially 0
There will be overhead in maintaining page tables. (GATE 2003: 2 Marks)
Ans. (c) Solution: To print `001100110011’. Process P
3. A uni-processor computer system only has two pro- must start first and process Q must wait until P
cesses, both of which alternate 10 ms CPU bursts prints 00 then process P must wait until process Q
with 90 ms I/O bursts. Both the processes were prints 11 and so on.
created at nearly the same time. The I/O of both In all of the four options P(T) at Y is given,
processes can proceed in parallel. Which of the so process Q have to waits for T and T must be
Chapter 7.indd 374 4/9/2015 10:01:38 AM

initialize to 0. Process P must signal T (V(T)) at 7. Which of the following will always lead to an
X. similarly in all four option P(S) is given at W so output staring with `001100110011’?
S must be initially 1 and Process Q signal S at Z to
(a) P (S) at W, V(S) at X, P(T) at Y, V(T) at Z, S
print desired output.
and T initially 1
Ans. (b)
(b) P(S) at W, V(T) at X, P(T) at Y, V(S) at Z, S
Common Data Questions 5 and 6: A processor initially 1, and T initially 0
uses two-level page tables for virtual to physical (c) P(S) at W, V(T) at X, P(T) at Y, V(S) at Z, S
address translation. Page tables for both levels are and T initially 1
stored in the main memory. Virtual and physical (d) P(S) at W, V(S) at X, P(T) at Y, V(T) at Z, S
addresses are both 32 bits wide. The memory is initially 1, and T initially 0
byte addressable. For virtual to physical address
Solution: To get output 001100110011, P and Q
translation, the 10 most significant bits of the vir-
should be executed alternatingly. P(S) at W, V(T)
tual address are used as index into the first-level
at X, P(T) at Y, V(S) at Z, S initially 1, and T
page table while the next 10 bits are used as index
initially 0 will give the correct output.
into the second-level page table. The 12 least sig-
nificant bits of the virtual address are used as offset Ans. (b)
within the page. Assume that the page table entries
8. Consider the following statements with respect to
in both levels of page tables are 4 bytes wide.
user-level threads and kernel-supported threads
Further, the processor has a translation look-
aside buffer (TLB), with a hit rate of 96%. The (i) Context switch is faster with kernel-supported
TLB caches recently used virtual page numbers threads.
and the corresponding physical page numbers. The (ii) For user-level threads, a system call can block
processor also has a physically addressed cache the entire process.
with a hit rate of 90%. Main memory access time (iii) Kernel-supported threads can be scheduled
is 10 ns, cache access time is 1 ns, and TLB access independently.
time is also 1 ns. (iv) User-level threads are transparent to the kernel.
5. Assuming that no page faults occur, the average Which of the above statements are true?
time taken to access a virtual address is approxi- (a) (ii), (iii) and (iv) only
mately (to the nearest 0.5 ns) (b) (ii) and (iii) only
(a) 1.5 ns (b) 2 ns (c) (i) and (iii) only
(c) 3 ns (d) 4 ns (d) (i) and (ii) only

Solution: Average time taken to access a virtual Solution: For option (i): User level thread may
address load only those registers which are required,
= [0.96(1 + 0.9 × 1 + 0.1 × 1 +10)] + [(0.04)(21 whereas kernel-level thread loads each registers, so
+ 0.9 × 1 + 0.1 × 1 + 10)] context switching is slower in kernel-level thread.
= 3.87 ≈ 4 ns For option (ii): Kernel do not have information
Ans. (d) about all the user-level thread, so for blocking a
6. Suppose a process has only the following pages user-level thread a system call is used.
in its virtual address space: two contiguous code Options (iii) and (iv) are correlated: Kernel-level
pages starting at virtual address 0 × 00000000, two threads are well known to kernel, so kernel sup-
contiguous data pages starting at virtual address ported thread can be scheduled independently and
0 × 00400000, and a stack page starting at virtual user level thread are transparent to kernel.
address 0 × FFFFF000. The amount of memory Ans. (a)
required for storing the page tables of this process is
9. Consider an operating system capable of loading
(a) 8 KB (b) 12 KB and executing a single sequential user process at a
(c) 16 KB (d) 20 KB time. The disk head scheduling algorithm used is
(GATE 2003: 2 Marks) first come first served (FCFS). If FCFS is replaced
by shortest seek time first (SSTF), claimed by the
Solution: Total memory required = 4 × 212 bytes vendor to give 50% better benchmark results, what
= 16 KB is the expected improvement in the I/O perfor-
Ans. (c) mance of user programs?
Chapter 7.indd 375 4/9/2015 10:01:38 AM

(a) 50% (b) 40% 12. Consider a system with a two-level paging scheme
(c) 25% (d) 0% in which a regular memory access takes 150 ns,
and servicing a page fault takes 8 ms. An aver-
(GATE 2004: 1 Mark)
age instruction takes 100 ns of CPU time, and two
Solution: I/O performance is determined by memory accesses. The TLB hit ratio is 90%, and
input and output devices. Disk scheduling will not the page fault rate is one in every 10000 instruc-
improve the performance of I/O for users. tions. What is the effective average instruction
Ans. (d) execution time?
10. The minimum number of page frames that must be (a) 645 ns (b) 1050 ns
allocated to a running process in a virtual memory (c) 1215 ns (d) 1230 ns
environment is determined by
(a) the instruction set architecture.
(b) page size. Solution: Total memory access time = 0.90 ×
(c) physical memory size. 150 + 0.10 × 300 = 165 ns
(d) number of processes in memory. Time taken to execute instructions = 100 + 2(165
+ 5000 × 8 × 106) = 1230 ns
(GATE 2004: 1 Mark) Ans. (d)
Solution: The minimum number of page frames
13. Consider two processes P1 and P2 accessing the
required by a process is based on instruction set
shared variables X and Y protected by two binary
architecture.
semaphores Sx and Sy, respectively, both initialized
Ans.(a)
to 1. P and V denote the usual semaphore opera-
11. Consider the following set of processes, with the tors, where P decrements the semaphore value, and
arrival times and the CPU-burst times given in V increments the semaphore value. The pseudo-
milliseconds. codes of P1 and P2 are as follows:
Process Arrival Time Burst Time P1: P2:
P1 0 5 While true do { While true do {

P2 1 3 L1;................ L3...............
P3 2 3
L2;................ L4;................
P4 4 1
X = X + 1; Y = Y + 1;
What is the average turnaround time for these Y = Y - 1; X = Y - 1;
processes with the pre-emptive shortest remaining
V(Sx); V(Sy);
processing time first (SRPT) algorithm?
(a) 5.50 (b) 5.75 V(Sy); V(Sx);
(c) 6.00 (d) 6.25
In order to avoid deadlock, the correct operators at
(GATE 2004: 2 Marks) L1, L2, L3 and L4 are, respectively,
Solution: (a) P(Sy), P(Sx); P(Sx), P(Sy)
(b)P(Sx), P(Sy); P(Sy), P(Sx)
P1 P2 P4 P3 P1 (c) P(Sx), P(Sx); P(Sy), P(Sy)
0 1 4 5 8 12 (d)P(Sy), P(Sy); P(Sx), P(Sy)
Turnaround time for (GATE 2004: 2 Marks)
P1 = 12 − 0 = 12 Solution: To avoid deadlock, values of L1, L2, L3
P2 = 4−1=3 and L4 should be considered as P(Sx), P(Sy), P(Sx)
P3 = 8−2=6 and P(Sy).
P4 = 5−4=1 Ans. (d)
Average turnaround time = (12 + 3 + 6 + 1)/4 14. A Unix-style i-node has 10 direct pointers and one
= 5.50 single, one double and one triple indirect point-
Ans. (a) ers. Disk block size is 1 KB, disk block address is
Chapter 7.indd 376 4/9/2015 10:01:38 AM

32 bits, and 48-bit integers are used. What is the operated in cycle stealing mode whereby whenever
maximum possible file size? one 4 byte word is ready it is sent to memory;
similarly, for writing, the disk interface reads a 4
(a) 224 bytes (b) 232 bytes
byte word from the memory in each DMA cycle.
(c) 234 bytes (d) 248 bytes
Memory cycle time is 40 nsec. The maximum per-
(GATE 2004: 2 Marks) centage of time that the CPU gets blocked during
DMA operation is:
Solution: Number of disk block pointers in one
block = 210 byte/32 bit = 256
256 × 256 × 256 × 1KB
Maximum size of file = (a) 10 (b) 25 (c) 40 (d) 50
= 234 bytes Solution: Data transfer in one rotation = 512 ×
Ans. (c) 1024 Bytes
15. What is the swap space in the disk used for? 1 rotation takes = 60/3000 s
(a) Saving temporary html pages 512KB is transferred in = 60/3000 s
(b) Saving process data
60
(c) Storing the super-block 1 byte will be transferred =
(d) Storing device drivers 3000 × 512 × 1024
(GATE 2005: 1 Mark) 4 bytes will be transferred
Solution: Swap space is used to store the process 60 × 4

= = 152.58 s
data. 3000 × 512 × 1024
Ans. (b)
16. Increasing the RAM of a computer typically 40
Block % = = 26%
improves performance because 152.58
(a) virtual memory increases. Ans. (b)
(b) larger RAMs are faster.
19. Consider the following code fragment:
(c) fewer page faults occur.
(d) fewer segmentation faults occur.
if (fork () ==0)
(GATE 2005: 1 Mark) {
Solution: With the increase in memory, page a = a + 5; printf(“%d,%dn”, a, &a);
faults will become less. }
Ans. (c) else
{
17. Suppose n processes P1, …, Pn share m identical
a = a – 5; printf(“%d, %dn”, a, &a);
resource units, which can be reserved and released
}
one at a time. The maximum resource requirement
of process Pi is si, where s > 0. Which one of the Let u and v be the values printed by the parent
following is a sufficient condition for ensuring that process, and x, y be the values printed by the child
deadlock does not occur? process. Which one of the following is TRUE?
(a) i, si < m (b) i, si < n (a) u = x + 10 and v = y
(b) u = x + 10 and v! =y
n n
∑ si < (m + n) ∑ si < (m × n)
(c) u + 10 = x and v = y
(c) (d)
(d) u + 10 = x and v! = y
i =1 i =1
Solution: To remain in safe state, the maximum Solution: Fork function returns 0 for child pro-
resource requirement must be less than m + n. cess and it gives Process id of child to parent pro-
Ans. (c) cess. Parent process executes else part of code. So
u = a − 5 for parent and for child x = a + 5.
18. Consider a disk drive with the following specifica-
Therefore, u + 10 = x. As there are two copies of a
tions:
is created, so they have different memory locations.
16 surfaces, 512 tracks/surface, 512 sectors/track, Hence, v ! = y.
1 KB/sector, rotation speed 3000 rpm. The disk is Ans. (d)
Chapter 7.indd 377 4/9/2015 10:01:39 AM

Linked Answer Questions 20 and 21: We are given any intervening access to the memory location x.
9 tasks T1, T2, … T9. The execution of each task Consider the following implementation of P and V
requires one unit of time. We can execute one task functions on a binary semaphore S.
at a time. Each task Ti has a profit Pi and a dead-
line di. Profit Pi is earned if the task is completed void P (binary_semaphore *S)
before the end of the (di)th unit of time. {
unsigned y;
Tank T1 T2 T3 T4 T5 T6 T7 T8 T9 unsigned *x = &(S->value);
Profit 15 20 30 18 18 10 23 16 25 do
Deadline 7 2 5 3 4 5 2 7 3 {
fetch-and-set x, y;
} while (y);
20. Are all tasks completed in the schedule that gives
}
maximum profit?
void V (binary_semaphore *S)
(a) All tasks are completed {
(b) T1 and T6 are left out S->value = 0;
(c) T1 and T8 are left out }
(d) T4 and T6 are left out
Which one of the following is true?
(a) The implementation may not work if context
Solution: switching is disabled in P.
T2 T7 T9 T5 T3 T8 T1 (b) Instead of using fetch-and-set, a pair of normal
load/store can be used.
T4 and T6 are left out. (c) The implementation of V is wrong.
Ans. (d) (d) The code does not implement a binary
semaphore.
21. What is the maximum profit earned?
(a) 147 (b) 165 (c) 167 (d) 175
Solution: If context switching is disabled, this
(GATE 2005: 2 Marks) implementation may not work. Then processes will
Solution: Maximum profit = 20 + 23 + 25 + 18 not be synchronized.
+ 30 + 15 + 16 = 147 Ans. (a)
Ans. (a)
24. A CPU generates 32-bit virtual addresses. The
22. Consider three CPU-intensive processes, which page size is 4 KB. The processor has a translation
require 10, 20 and 30 time units and arrive at look-aside buffer (TLB) which can hold a total of
times 0, 2 and 6, respectively. How many context 128 page table entries and is 4-way set associative.
switches are needed if the operating system imple- The minimum size of the TLB tag is
ments a shortest remaining time first scheduling
algorithm? Do not count the context switches at (a) 11 bits (b) 13 bits
time zero and at the end. (c) 15 bits (d) 20 bits
(a) 1 (b) 2 (GATE 2006: 2 Marks)

(c) 3 (d) 4
Solution: Virtual address = 32 bits
(GATE 2006: 1 Mark)
Page size = 4KB = 12 bits
Solution: Number of pages = 128 = 27 = 7 bits are required
In 4-way set-associative memory = 128/4 = 32 =
P1 P2 P3 5 bits required
0 10 30 60 Tag bits = 32 − 12 − 5 = 15
Two times context switching will be done. Ans. (c)
Ans. (b)
25. A computer system supports 32-bit virtual
23. The atomic fetch-and-set x, y instruction uncon- addresses as well as 32-bit physical addresses. Since
ditionally sets the memory location x to 1 and the virtual address space is of the same size as
fetches the old value of x n y without allowing the physical address space, the operating system
Chapter 7.indd 378 4/9/2015 10:01:40 AM

designers decide to get rid of the virtual memory as possible. For what percentage of time does the
entirely. Which one of the following is true? CPU remain idle?
(a) Efficient implementation of multi-user support (a) 0% (b) 10.6%
is no longer possible. (c) 30.0% (d) 89.4%
(b) The processor cache organization can be made (GATE 2006: 2 Marks)
more efficient now.
(c) Hardware support for memory management is Solution: Let three processes be p0, p1 and p2.
no longer needed. Their execution time is 10, 20 and 30, respectively.
(d) CPU scheduling can be made more efficient now. p0 spends first 2 time units in I/O, 7 units of CPU
(GATE 2006: 2 Marks) time and finally 1 unit in I/O. p1 spends first
4 units in I/O, 14 units of CPU time and finally
Solution: Special hardware support is required
2 units in I/O. p2 spends first 6 units in I/O, 21
for address translation in case of virtual addressing.
units of CPU time and finally 3 units in I/O.
Ans. (c)
26. Consider three processes (process id 0, 1, 2, respec- Ideal P0 P1 P2 ideal
tively) with compute time bursts 2, 4 and 8 time 0 2 9 23 44 47
units. All processes arrive at time zero. Consider
Total time spent = 47
the longest remaining time first (LRTF) schedul-
Idle time = 2 + 3 = 5
Percentage of idle time = (5/47) × 100 = 10.6%
ing algorithm. In LRTF, ties are broken by giving
priority to the process with the lowest process id.
The average turnaround time is Ans. (b)
(a) 13 units (b) 14 units 28. Consider the following snapshot of a system run-
(c) 15 units (d) 16 units ning n processes. Process i is holding xi instances of
a resource R, 1 n. Currently, all instances of R are
(GATE 2006: 2 Marks) occupied. Further, for all i, process i has placed a
Solution: request for an additional yi instances while holding
the xi instances it already has. There are exactly
Process Burst Remaining Turnaround two processes p and q such that yp = yq = 0. Which
Time Time Time one of the following can serve as a necessary condi-
P0 2 2−1=1 12 tion to guarantee that the system is not approach-
ing a deadlock?
4 − 1 = 3 − 1 = 13
(a) min (xp, xq) < maxk ≠ p,qyk
P1 4
2−1=1
(b) xp+ xq ³≥ mink ≠ p,qyk
P3 8 8 − 4 = 4 − 1 = 14 (c) max (xp, xq) > 1
3 − 1 = 2 − 1=1 (d) min (xp, xq) > 1
P2 P1 P2 P1 P2 P0 P1 P2 P0 P1 P2 Solution: Processes p and q have sufficient
resources so they do not require any additional
0 4 5 6 7 8 9 10 11 12 13 14
resources and both of them can complete and
Average turnaround time = 12 + 13 + 14 = 39/3 release all resources acquired by them which are (xp
= 13 + xq). If the resources released by process p and q
Ans. (a) are sufficient for another process which is waiting
for yk resources, then system will not be in deadlock.
27. Consider three processes, all arriving at time zero,
Ans. (b)
with total execution time of 10, 20 and 30 units,
respectively. Each process spends the first 20% of Linked Answer Questions 29 and 30: Barrier is a
execution time doing I/O, the next 70% of time synchronization construct where a set of processes
doing computation, and the last 10% of time doing synchronizes globally, that is, each process in the
I/O again. The operating system uses a shortest set arrives at the barrier and waits for all others
remaining compute time first scheduling algorithm to arrive and then all processes leave the barrier.
and schedules a new process either when the run- Let the number of processes in the set be three and
ning process gets blocked on I/O or when the run- S be a binary semaphore with the usual P and V
ning process finishes its compute burst. Assume functions. Consider the following C implementa-
that all I/O operations can be overlapped as much tion of a barrier with line numbers shown on left.
Chapter 7.indd 379 4/9/2015 10:01:40 AM

void barrier (void) { Group I Group II

1: P(S);
2: process_arrived++; (P) Gang Scheduling (1) G
uaranteed
3. V(S); Scheduling
4: while (process_arrived !=3); (Q) Rate Monotonic (2) R
eal-time
5: P(S); Scheduling Scheduling
6: process_left++; (R) F
air Share (3) T
hread
7: if (process_left==3) { Scheduling Scheduling
8: process_arrived = 0;
9: process_left = 0; (a) P − 3; Q − 2; (b) P − 1; Q − 2;
10: } R − 1 R−3
11: V(S); (c) P − 2; Q − 3; (d) P − 1; Q − 3;
} R−1 R−2
The variables process_arrived and process_left are (GATE 2007: 1 Mark)

shared among all processes and are initialized to
zero. In a concurrent program, all the three pro- Solution: Gang scheduling: Used in parallel sys-
cesses call the barrier function when they need to tems like for thread scheduling
synchronize globally. Rate monotonic scheduling: Used in real-time
29. The above implementation of barrier is incorrect. systems
Which one of the following is true? Fair share scheduling: Used for guaranteed
scheduling
(a) The barrier implementation is wrong due to Ans. (a)
the use of binary semaphore S.
(b) The barrier implementation may lead to a 32. Consider a disk pack with 16 surfaces, 128 tracks
deadlock if two barriers in invocations are used per surface and 256 sectors per track. 512 bytes of
in immediate succession. data are stored in a bit serial manner in a sector.
(c) Lines 6 to 10 need not be inside a critical The capacity of the disk pack and the number of
section. bits required to specify a particular sector in the
(d) The barrier implementation is correct if there disk are, respectively,
are only two processes instead of three. (a) 256 Mbyte, 19 bits
(GATE 2006: 2 Marks) (b) 256 Mbyte, 28 bits
(c) 512 Mbyte, 20 bits
Solution: Fact
(d) 64 Gbyte, 28 bits
Ans. (b)
(GATE 2007: 1 Mark)
30. Which one of the following rectifies the problem in
the implementation? Solution: Disk capacity = 16 surfaces × 128
(a) Lines 6 to 10 are simply replaced by tracks × 256 sectors × 512 bytes = 256 MB
process_arrived. Total number of sectors = 16 × 128 × 256 = 219
(b) At the beginning of the barrier the first process Ans. (a)
to enter the barrier waits until process_arrived
becomes zero before proceeding to execute 33. Consider the following statements about user-level
P(S). threads and kernel-level threads. Which one of the
(c) Context switch is disabled at the beginning of following statements is FALSE?
the barrier and re-enabled at the end. (a) Context switch time is longer for kernel-level
(d) The variable process_left is made private threads than for user-level threads.
instead of shared. (b) User-level threads do not need any hardware
(GATE 2006: 2 Marks) support.
(c) Related kernel-level threads can be scheduled
Solution: Fact on different processors in a multiprocessor
Ans. (b) system.
31. Group 1 contains some CPU scheduling algorithms (d) Blocking one kernel-level thread blocks all
and Group 2 contains some applications. Match related threads.
entries in Group 1 to entries in Group 2. (GATE 2007: 1 Mark)
Chapter 7.indd 380 4/9/2015 10:01:40 AM

Solution: Kernel-level threads are managed by kernel. (a) Both P and Q are true, and Q is the reason for P.
Blocking one thread may not block other threads. (b) Both P and Q are true, but Q is not the reason
Ans. (d) for P.
34. An operating system uses shortest remaining time (c) P is false, but Q is true.
first (SRT) process scheduling algorithm. Consider (d) Both P and Q are false.
the arrival times and execution times for the fol- (GATE 2007: 2 Marks)
lowing processes:
Solution: P: True, it is Belady’s anomaly.
Process Execution Time Arrival Time Q: True, but not reason for Belady’s anomaly.
Ans. (b)
P1 20 0
36. A single processor system has three resource types
P2 25 15 X, Y and Z, which are shared by three processes.
P3 10 30 There are 5 units of each resource type. Consider
P4 15 45 the following scenario, where the column alloc
denotes the number of units of each resource type
allocated to each process, and the column request
What is the total waiting time for process P2?
denotes the number of units of each resource type
(a) 5 (b) 15 requested by a process in order to complete execu-
(c) 40 (d) 55 tion. Which of these processes will finish LAST?
(GATE 2007: 2 Marks) alloc request
Solution:
X Y Z X Y Z
P1 P2 P3 P2 P4
P0 1 2 1 1 0 3
0 20 30 40 55 70 P1 2 0 1 0 1 2
Waiting time of P2 = C ompletion time − (Arrival P2 2 2 1 1 2 0
time + Execution time)
= 55 − (15 + 25) (a) P0
= 15 (b) P1
Ans. (b) (c) P2
35. A virtual memory system uses first in first out (d) None of the above, since the system is in a
(FIFO) page replacement policy and allocates a deadlock
fixed number of frames to a process. Consider the (GATE 2007: 2 Marks)
following statements:
Solution: Available resources: 0 1 2
P: Increasing the number of page frames allocated P1’s request can be fulfilled, it will start executing.
to a process sometimes increases the page fault rate. Available after P1 execution: 2 1 3
Q: Some programs do not exhibit locality of P0’s request can be fulfilled and then P2 will execute.
reference. P2 will finish last.
Which one of the following is TRUE? Ans. (c)
37. Two processes, P1 and P2, need to access a critical section of code. Consider the following synchronization con-
struct used by the processes:
In process P1 and P2, variables wants1 and wants2 both become true, but process P1 and P2 are stuck for infinite time
in their while loops waiting for each other to finish. So, they do not prevent deadlock. But mutual exclusion is ensured.
P1 P2
while (true) {while (true) { while (true) {while (true) {

wants1 = true;wants1 = true; wants2 = true;wants2 = true;
while (wants2==true);
while (wants2==true); while (wants1==true);
while (wants1==true);
/* Critical /* Critical /* Critical /* Critical
Section */ Section */ Section */ Section */
Wants1=false; Wants1=false; Wants2=false; Wants2=false;
} } } }
/* Remainder section
/* Remainder
*/ section */ /* Remainder section
/* Remainder
*/ section */
Chapter 7.indd 381 4/9/2015 10:01:41 AM

Which one of the following statements is TRUE 40. The data blocks of a very large file in the Unix file
about the above construct? system are allocated using
(a) It does not ensure mutual exclusion. (a) contiguous allocation.
(b) It does not ensure bounded waiting. (b) linked allocation.
(c) It requires that processes enter the critical sec- (c) indexed allocation.
tion in strict alternation. (d) an extension of indexed allocation
(d) It does not prevent deadlocks, but ensures
(GATE 2008: 1 Mark)
mutual exclusion.
(GATE 2007: 2 Marks) Solution: i-node is used in Unix file system.
Ans.(d)
Solution: When wants1 and wants2 both become
true, both processes will be in wait state. But 41. For a magnetic disk with concentric circular tracks,
mutual exclusion is ensured. the seek latency is not linearly proportional to the
Ans. (d) seek distance due to
Linked Answer Questions 38 and 39: A process has (a) non-uniform distribution of requests.
been allocated 3-page frames. Assume that none (b) arm starting and stopping inertia.
of the pages of the process are available in the (c) higher capacity of tracks on the periphery of
memory initially. The process makes the following the platter.
sequence of page references (reference string): 1, 2, (d) use of unfair arm scheduling policies.
1, 3, 7, 4, 5, 6, 3, 1.
(GATE 2008: 1 Mark)
38. If optimal page replacement policy is used, how
many page faults occur for the above reference Solution: For a magnetic disk with concentric
string? circular tracks, the seek latency is not linearly pro-
portional to the seek distance due to arm starting
(a) 7 (b) 8
and stopping inertia.
(c) 9 (d) 10
Ans. (b)
Solution: 42. In an instruction execution pipeline, the earliest
Using optimal page replacement: page faults = 7 that the data TLB (Translation Lookaside Buffer)
can be accessed is
1
(a) before effective address calculation has started.
2 → 7 → 4 → 5 →6 (b) during effective address calculation.
3 (c) after effective address calculation has
Ans. (a) completed.
39. Least recently used (LRU) page replacement (d) after data cache lookup has completed.
policy is a practical approximation to optimal page (GATE 2008: 2 Marks)
replacement. For the above reference string, how
many more page faults occur with LRU than with Solution: For effective address calculation, TLB
the optimal page replacement policy? can be accessed earliest.
Ans. (b)
(a) 0 (b) 1
(c) 2 (d) 3 43. The P and V operations on counting semaphores,
(GATE 2007: 2 Marks) where s is a counting semaphore, are defined as
follows:
Solution:
Using LRU: page faults = 9 P(s) : s = s - 1;
if s < 0 then wait;
1→4→3
V(s) : s = s + 1;
2→7→6 if s <= 0 then wakeup a
process waiting on s;
3→5→1
9 − 7 = 2 more page faults when LRU is used. Assume that Pb and Vb the wait and signal opera-
Ans. (c) tions on binary semaphores are provided. Two
Chapter 7.indd 382 4/9/2015 10:01:41 AM

binary semaphores xb and yb are used to imple- 45. Which of the following is NOT true of deadlock
ment the semaphore operations P(s) and V(s) as prevention and deadlock avoidance schemes?
follows:
(a) In deadlock prevention, the request for resources
P(s) : Pb (xb); is always granted if the resulting state is safe.
s = s -1; (b) In deadlock avoidance, the request for resources
if (s < 0) { is always granted if the result state is safe.
Vb (xb); (c) Deadlock avoidance is less restrictive than
Pb (yb); deadlock prevention.
} (d) Deadlock avoidance requires knowledge of
else Vb (xb); resource requirements a priori.
(GATE 2008: 1 Mark)
V(s) : Pb (xb); Solution: Deadlock avoidance scheme is less
s = s + 1; restrictive than deadlock prevention because in
if (s <=0) Vb (yb); the latter, the request for a resource may not be
Vb (xb); granted even if the resulting state is safe.
Ans. (c)
The initial values of xb and yb are, respectively,
46. A process executes the following code
(a) 0 and 0 (b) 0 and 1
(c) 1 and 0 (d) 1 and 1 for (i = 0; i < n; i + +) for ( );
(GATE 2008: 1 Mark) The total number of child processes created is

(a) n (b) 2n − 1
Solution: xb and yb are binary semaphores. P(S)
decrements the value of semaphore and V(S) incre- (c) 2n (d) 2n+1 − 1
ments it. To ensure the property of mutual exclu- (GATE 2008: 1 Mark)
sion, value passed to xb should be 1 and value for
yb should be 0. Solution:
Ans.(c) 1 fork system call creates 1 child processes.
2 fork calls create 3 child processes, and so on
44. Which of the following statements about synchro- With n fork calls, there will be 2n − 1 child pro-
nous and asynchronous I/O is NOT true? cesses created.
Ans. (b)
(a) An ISR is invoked on completion of I/O in syn-
chronous I/O but not in asynchronous I/O. 47. A processor uses 36-bit physical addresses and
(b) In both synchronous and asynchronous I/O, 32-bit virtual addresses, with a page frame size of
an ISR (Interrupt Service Routine) is invoked 4 KB. Each page table entry is of size 4 bytes. A
after completion of the I/O. three-level page table is used for virtual to physical
(c) A process making a synchronous I/O call waits address translation, where the virtual address is
until I/O is complete, but a process making an used as follows
asynchronous I/O call does not wait for com-
•B its 30—31 are used to index into the first-level
pletion of the I/O.
page table.
(d) In the case of synchronous I/O, the process
• Bits 21—29 are used to index into the second-level
waiting for the completion of I/O is woken up
page table.
by the ISR that is invoked after the completion
• Bits 12—20 are used to index into the third-level
of I/O.
page table.
(GATE 2008: 1 Mark) • Bits 0—11 are used as offset within the page.
Solution: Interrupt is generated on completion of The number of bits required for addressing the
both synchronous and asynchronous I/O. In syn- next level page table (or page frame) in the page
chronous, the interrupt wakes up another process table entry of the first, second and third level page
that is waiting for I/O, whereas in asynchronous, tables are, respectively,
interrupt informs the other processes that the
(a) 20, 20 and 20 (b) 24, 24 and 24
I/O is complete and they can process their I/O
(c) 24, 24 and 20 (d) 25, 25 and 24
operation.
Chapter 7.indd 383 4/9/2015 10:01:41 AM

Solution: Process P1: Process P2: Process P3:

Given that
Physical address size = 36 bits t = 0: requests t = 0: requests t = 0: requests
Virtual address size = 32 bits 2 units of R2 2 units of R3 1 unit of R4
Physical memory size = 236 bytes t = 1: requests t = 2: requests t = 2: requests
Page frame size = 4K bytes = 212 bytes 1 unit of R3 1 unit of R4 2 units of R1
Number of bits required to access physical memory t = 3: requests t = 4: requests t = 5: releases 2
frame = 36 − 12 = 24, so in third level of page 2 units of R1 1 unit of R1 units of R1
table to access an entry 24 bits are required. t=5: releases 1 t = 6: releases t = 7: requests
To access second-level page table entry, we require unit of R2 and 1 unit of R3 1 unit of R2
9-bit of virtual address and size of pages in second 1 unit of R1 t = 8: Finishes t = 8: requests
level of page table is 4 bytes. So, the size of second 1 unit of R3
level page table is (29) × 4 = 211 bytes.
t = 7: releases 1 t = 9: Finishes
There are (236)/(211) possible locations to store
unit of R3
this page table. Therefore, the second page table
requires 25 bits to address it. t = 8: requests
Similarly, we can calculate that for first level page 2 units of R4
table needs 25 bits to address it. t = 10: Finishes
Ans. (b)
48. In which one of the following page replacement Which one of the following statements is TRUE
policies, Belady’s anomaly may occur? if all three processes run concurrently starting at
time t = 0?
(a) FIFO (b) Optimal
(c) LRU (d) MRU (a) All processes will finish without any deadlock.
(b) Only P1 and P2 will be in deadlock.
(GATE 2009: 1 Mark) (c) Only P1 and P3 will be in a deadlock.
(d) All three processes will be in deadlock.
Solution: Belady’s anomaly refers to the phenom-
enon where increasing the number of page frames (GATE 2009: 2 Marks)
results in increase in the number of page faults. Solution: Apply deadlock detection algorithmand
This phenomenon is experienced when FIFO page see that there is no process waiting indefinitely for
replacement algorithm is used. a resource.
Ans. (a) Ans. (a)
49. The essential content(s) in each entry of a page 51. Consider a disk system with 100 cylinders. The
table is/are requests to access the cylinders occur in the follow-
(a) virtual page number ing sequence:
(b) page frame number 4, 34, 10, 7, 19, 73, 2, 15, 6, 20
(c) both virtual page number and page frame Assuming that the head is currently at cylinder 50,
number what is the time taken to satisfy all requests if it
(d) access right information takes 1ms to move from one cylinder to adjacent
(GATE 2009: 1 Mark) one and shortest seek time first policy is used?
Solution: (a) 95 ms (b) 119 ms
Page frame number is must in page table. (c) 233 ms (d) 276 ms
50. Consider a system with four types of resources R1 Solution: 4, 34, 10, 7, 19, 73, 2, 15, 6, 20
(3 units), R2 (2 units), R3 (3 units), R4 (2 units). A Since shortest seek time first policy is used, head
non-preemptive resource allocation policy is used. will first move to 34. This move will cause 16 × 1
At any given instance, a request is not entertained ms. After 34, head will move to 20 which will cause
if it cannot be completely satisfied. Three processes 14 × 1 ms and so on. So, cylinders are accessed in
P1, P2, P3 request the sources as follows if executed the following order 34, 20, 19, 15, 10, 7, 6, 4, 2, 73
independently. and total time will be
Chapter 7.indd 384 4/9/2015 10:01:41 AM

(16 + 14 + 1 + 4 + 5 + 3 + 1 + 2 + 2 + 71) × I. The above solution to CS problem is

1 = 119 ms deadlock-free.
Ans. (b) II. The solution is starvation-free.
III. The processes enter CS in FIFO order.
52. In the following process state transition diagram IV. More than one process can enter CS at the
for a uniprocessor system, assume that there are same time.
always some processes in the ready state: Which of the above statements is TRUE?
B (a) I only (b) I and II
(c) II and III (d) IV only
Start A Ready Running D Terminated
C Solution:
E F The code segment checks that deadlock should
Blocked not occur, as the functions test-and-set and
leave_CS initialize memory location x to 0.
Ans. (a)
Now, consider the following statements: 54. A multilevel page table is preferred in comparison
I. If a process makes a transition D, it would to a single-level page table for translating virtual
result in another process making transition A address to physical address because
immediately.
(a) it reduces the memory access time to read or
II. A process P2 in blocked state can make transi-
write a memory location.
tion E, while another process P1 is in running
(b) it helps to reduce the size of page table needed to
state.
implement the virtual address space of a process.
III. The OS uses pre-emptive scheduling.
(c) it is required by the translation look-aside
IV. The OS uses non-preemptive scheduling.
buffer.
Which of the above statements are TRUE?
(d) it helps to reduce the number of page faults in
(a) I and II (b) I and III page replacement algorithms.
(c) II and III (d) II and IV
Solution: To reduce the page table size, multi-
Solution: Arrow C shows pre-emptive scheduling. level paging is used.
II is also true, a process can be in ready queue after Ans. (b)
completing I/O operation.
Ans. (c) 55. Consider the methods used by processes P1 and
P2 for accessing their critical sections whenever
53. The enter_CS() and leave_CS() functions to needed, as given below. The initial values of shared
implement critical section of a process are realized Boolean variables S1 and S2 are randomly assigned.
using test-and-set instruction as follows:
Method used by P1 Method used by P2
void enter_CS(X) while (S1 = = S2); while (S1 != S2);
{
( ) Critica1 section Critica1 section
while test-and-set(X) ; S1 = S2; S2 = not (S1);
}
void leave_CS(X) Which one of the following statements describes
{ the properties achieved?
X=0;
(a) Mutual exclusion but not progress
}
(b) Progress but not mutual exclusion
(c) Neither mutual exclusion nor progress
In the above solution, X is a memory location asso-
(d) Both mutual exclusion and progress
ciated with the CS and is initialized to 0. Now con-
sider the following statements: (GATE 2010: 1 Mark)
Chapter 7.indd 385 4/9/2015 10:01:42 AM

Solution: Both P1 and P2 are checking different

Process P0 Process P1 Process P2
conditions, so ensuring mutual exclusion. Progress
while they
is not shown by P1 and P2, because (true) while (true)
are making wait(true)
while (S1); wait (S1); waitwait
(S1);(S2); wait (S2); wai
{
their own condition true again. { { release (S0);release (S0);
release (S0);
release (S0);
release (S0);
rel
wait (S0); Ans.wait(a)(S0); wait (S0);
56. Two alternative packages A print '0'
and B are print
available for '0' print '0'
processing a database havingrelease (S1);
101 records. release (S1);
Package release (S1);
A requires 0.0001 n2 time release
units and(S2); release
package B (S2);release (S2);
requires 10nlog10n time units} to process n records.
} }
What is the smallest value of k for which package
B will be preferred over A? How many times will process P0 print `0’?
(a) 12 (b) 10 (c) 6 (d) 5 (a) At least twice (b) Exactly twice
(c) Exactly thrice (d) Exactly once
Solution:
0.0001 n2 < 10 n log n (GATE 2010: 2 Marks)
For k = 6, B will be preferred.
Solution: S0 = 1, only P0 will execute, it will
Ans. (c)
print 0. P0 will increment S1 to 1, process P1 will
57. A system uses FIFO policy for page replacement. start executing which will again make S0 to 1.
It has 4-page frames with no pages loaded to begin Process P0 will again print 0. Similarly, P2 will
with. The system first accesses 100 distinct pages perform the action. Hence, at least twice 0 will be
in some order and then accesses the same 100 printed.
pages, but now in the reverse order. How many Ans. (a)
page faults will occur?
60. A system has n resources R0, …, Rn−1, and k
(a) 196 (b) 192 (c) 197 (d) 195 processes P0, …, Pk−1. The implementation of
(GATE 2010: 1 Mark) the resource request logic of each process Pi is as
follows:
Solution: Page faults that occur = 100 + 100 − 4
= 196 if (i%2==0)
4 is subtracted because the last four will be present {
in frames. if (i<n) request Ri;
Ans. (a) if (i+2<n)request Ri+2;
58. Which of the following statements are true? }
else
I. Shortest remaining time first scheduling may
{
cause starvation.
if (i<n) request Rn-i;
II. Pre-emptive scheduling may cause starvation.
if (i+2<n)request Rn-i-2;
III. Round robin is better than FCFS in terms of
}
response time.
(a) I only (b) I and III only In which one of the following situations is a dead-
(c) II and III only (d) I, II and III lock possible?
(GATE 2010: 1 Mark) (a) n = 40, k = 26 (b) n = 21, k = 12
(c) n = 20, k = 10 (d) n = 41, k = 19
Solution:
I. High remaining time jobs will starve. (GATE 2010: 2 Marks)
II. In pre-emptive scheduling, low priority jobs
will suffer. Solution: Using option (b), we have
III. Response time will be better round robin if k = 12, then p12−1 = p11 ⇒ i = 10 then else block
because each process will get chance to execute is executed and Rn−i, Rn−i−2 = R21−11, R21−11−2 =
on periodic basis. R10, R8. But R10 is also requested when k = 11, so
Ans. (d) this results in dead lock.
59. The following program consists of three concurrent Ans. (b)
processes and three binary semaphores. The sema- 61. Let the time taken to switch between user and
phores are initialized as S0 = 1, S1 = 0, S2 = 0. kernel modes of execution be t1 while the time
Chapter 7.indd 386 4/9/2015 10:01:43 AM

taken to switch between two processes be t2. Which The pre-emptive shortest job first scheduling
of the following is TRUE? algorithm is used. Scheduling is carried out only
(a) t1 > t2 at arrival or completion of processes. What is the
(b) t1 = t2 average waiting time for the three processes?
(c) t1 < t2 (a) 5.0 ms (b) 4.33 ms
(d) Nothing can be said about the relation between (c) 6.33 ms (d) 7.33 ms
t1and t2
(GATE 2011: 1 Mark)
Solution:
Solution: Process switching requires more time. P0 P1 P0 P2
Ans. (c)
0 1 5 13 22
62. A thread is usually defined as a `light weight pro-
cess’ because an operating system (OS) maintains Waiting time
smaller data structures for a thread than for a pro- P0 = 5 − 1 = 4 ms
cess. In relation to this, which of the following is P1 = 1 − 1 = 0 ms
TRUE? P2 = 13 − 2 = 11 ms
(a) On per-thread basis, the OS maintains only Average waiting time = 4 + 11/3 = 5 ms
CPU register state. Ans. (a)
(b) The OS does not maintain a separate stack for 65. An application loads 100 libraries at startup.
each thread. Loading each library requires exactly one disk
(c) On per-thread basis, the OS does not maintain access. The seek time of the disk to a random loca-
virtual memory state. tion is given as 10 ms. Rotational speed of disk
(d) On per-thread basis, the OS maintains only is 6000 rpm. If all 100 libraries are loaded from
scheduling and accounting information. random locations on the disk, how long does it take
to load all libraries? (The time to transfer data
(GATE 2011: 1 Mark)
from the disk block once the head has been posi-
Solution: Operating System maintains separate tioned at the start of the block may be neglected.)
stack, register and thread-specific data for differ-
ent threads. OS does not maintain virtual memory (a) 0.50 s (b) 1.50 s
state. So option (c) is correct. (c) 1.25 s (d) 1.00 s
63. Let the page fault service time be 10 ms in a com- Solution: Since transfer time can be neglected,
puter with average memory access time being 20 the average access time is sum of average seek
ns. If one page fault is generated for every 106 time and average rotational latency. Average seek
memory accesses, what is the effective access time time for a random location time is given as 10 ms.
for the memory? The average rotational latency is half of the time
needed for complete rotation. It is given that 6000
(a) 21 ns (b) 30 ns rotations need 1 min. So one rotation will take
(c) 23 ns (d) 35 ns 60/6000 s, which is 10 ms. Therefore, average rota-
(GATE 2011: 1 Mark) tional latency is half of 10 ms, which is 5 ms.
Solution: p = page fault rate = 1/106 Average disk access time = S eek time + rotational
latency
Effective access time = p × page fault service time = 10 ms + 5 ms = 15 ms
+ (1 − p) × memory access time For 100 libraries, the average disk access time will
= 1/106 × 10 + [(1 − (1/106)] × 20 ≈ 30 ns be 15 × 100 ms.
Ans. (b) Ans. (b)
64. Consider the following table of arrival time and 66. A process executes the code
burst time for three processes P0, P1 and P2. fork();
fork();
Process Arrival Time Burst Time fork();
P0 0 ms 9 ms The total number of child processes created is
P1 1 ms 4 ms (a) 3 (b) 4
P2 2 ms 9 ms (c) 7 (d) 8
(GATE 2012: 1 Mark)
Chapter 7.indd 387 4/9/2015 10:01:43 AM

AcquireLock(L){
388 Chapter 7: Operating System while (Fetch_And_Add(L,1))
L = 1;
Solution: fork() system call is used to create child }
processes. ReleaseLock(L){
Total number of processes: 2n L = 0;
Child processes: 2n − 1, here n represents the }
number of fork calls.
This implementation
Ans. (c)
(a) fails as L can overflow.
67. Consider the three processes, P1, P2 and P3 shown (b) fails as L can take on a non-zero value when
in the table. the lock is actually available.
(c) works correctly but may starve some processes.
Process Arrival Time Time Units (d) works correctly without starvation.
Required
P1 0 5
P2 1 7 Solution: Let a process has just released the lock
and made L = 0 and there is one more process wait-
P3 3 4 ing for the lock, that process will then execute the
The completion order of the three processes under AcquireLock(). Just after the L was made 0, let the
the policies FCFS and RR2 (round-robin schedul- waiting processes executed the line L = 1. And then
ing with CPU quantum of 2 time units) are all the waiting process will execute AcquireLock()
and they cannot come out of the while loop.
(a) FCFS: P1, P2, P3 (b) FCFS: P1, P3, P2
Ans. (b)
RR2: P1, P2, P3 RR2: P1, P3, P2
(c) FCFS: P1, P2, P3 (d) FCFS: P1, P3, P2 69. A file system with 300 GB disk uses a file descrip-
RR2: P1, P3, P2 RR2: P1, P2, P3 tor with 8 direct block addresses, 1 indirect block
(GATE 2012: 2 Marks) address and 1 doubly indirect block address. The
size of each disk block is 128 bytes and the size of
Solution: FCFS Scheduling: each disk block address is 8 bytes. The maximum
P1 P2 P3 possible file size in this file system is
0 5 12 16 (a) 3 KB
Time of completion for P1: 5 − 0 = 5 (b) 35 KB
P2: 12 − 1 = 11 (c) 280 bytes
P3: 16 − 3 = 13 (d) Dependent on the size of the disk
Order of completion is P1, P2, P3. (GATE 2012: 2 Marks)
RR2 Scheduling:
P1 P2 P1 P3 P2 P2 P3 P2 P2 Solution:
Disk block size = 128 bytes
0 2 4 6 8 10 11 13 15 16 Disk block addresses = 8 bytes
Time of completion for P1: 11 − 0 = 11 Number of entries = 128/8 = 16
P2: 16 − 1 = 15 8 direct block addresses = 8 × 128 = 1024
P3: 13 − 3 = 10 1 indirect block addresses = 16 × 128 = 2048
Order of completion is P1, P3, P2. 1 doubly indexes = 16 × 16 × 128 = 32768
Ans. (c) 35 KB
68. Fetch_And_Add(X, i) is an atomic read-modify- Ans. (b)
write instruction that reads the value of memory
location X, increments it by the value i, and returns 70. Consider the virtual page reference string
the old value of X. It is used in the pseudocode 1, 2, 3, 2, 4, 1, 3, 2, 4, 1
shown below to implement a busy-wait lock. L is On a demand paged virtual memory system run-
an unsigned integer shared variable initialized to 0. ning on a computer system with main memory size
The value of 0 corresponds to lock being available, of 3 pages frames which are initially empty. Let
while any non-zero value corresponds to the lock LRU, FIFO and OPTIMAL denote the number of
being not available. page faults under the corresponding page replace-
AcquireLock(L){ ments policy. Then
while (Fetch_And_Add(L,1)) (a) OPTIMAL < LRU < FIFO
L = 1; (b) OPTIMAL < FIFO < LRU
}
ReleaseLock(L){
L = 0;
}
Chapter 7.indd 388 4/9/2015 10:01:44 AM

(c) OPTIMAL = LRU 72. A scheduling algorithm assigns priority propor-

(d) OPTIMAL = FIFO tional to the waiting time of a process. Every pro-
(GATE 2012: 2 Marks) cess starts with priority zero (the lowest priority).
The scheduler re-evaluates the process priorities
Solution: every T time units and decides the next process to
LRU: 1 1 1 4 4 4 2 2 2 9 page faults schedule. Which one of the following is TRUE if
22223331 the processes have no I/O operations and all arrive
3311144 at time zero?
FIFO: 1 1 1 4 4 4 6 page faults
(a) This algorithm is equivalent to the first-come-
22211
first-serve algorithm.
3332
(b) This algorithm is equivalent to the Round-
OPTIMAL: 1 1 1 1 1 5 page faults
Robin algorithm.
2244
(c) This algorithm is equivalent to the shortest-
332
job-first algorithm.
OPTIMAL < FIFO < LRU
(d) This algorithm is equivalent to the shortest-
Ans.(b)
remaining-time-first algorithm.
71. Consider the following transactions with data
(GATE 2013: 1 Mark)
items P and Q initialized to zero:
Solution: Because time quantum T is given here
T1: read (P);
to reschedule priority. So, it is equivalent to Round
read (Q);
Robin algorithm.
if P = 0 then Q := Q + 1;
Ans. (b)
write (Q).
T2: read (Q); 73. Consider a hard disk with 16 recording surfaces
read (P); (0-15) having 16384 cylinders (0-16383) and each
if Q = 0 then P := P + 1; cylinder contains 64 sectors (0-63). Data storage
write (P). capacity in each sector is 512 bytes. Data are orga-
Any non-serial interleaving of T1 and T2 for con- nized cylinder-wise and the addressing format is
current execution leads to <cylinder no., surface no., sector no.>. A file of
size 42797 KB is stored in the disk and the starting
(a) a serializable schedule disk location of the file is <1200, 9, 40>. What is
(b) a schedule that is not conflict serializable the cylinder number of the last sector of the file, if
(c) a conflict serializable schedule it is stored in a contiguous manner?
(d) a schedule for which a precedence graph cannot
be drawn (a) 1281 (b) 1282 (c) 1283 (d) 1284
Solution: Two or more actions are said to be in Solution: The number of sectors required to store
conflict if:
42797 × 1024
(i) The actions belong to different transactions. the file = = 85594 sectors
512
(ii) At least one of the actions is a write operation.
Number of sectors in a cylinder = 16 × 64 = 1024
(iii) The actions access the same object (read or
write). Total number of cylinders required = 85594/1024
= 84
Take the following schedule:
Last sector will be stored on 1284th cylinder.
S: r1(P); r1(Q); r2(Q); r2(P); w1(Q); w2(P)
Ans. (d)
74. A certain computation generates two arrays a and
b such that a[i]=f(i) for 0 ≤ i < n and b[i]=g(a[i]) for
T1 T2 0 ≤ i < n. Suppose this computation is decomposed
into two concurrent processes X and Y such that
X computes the array a and Y computes the array
b. The processes employ two binary semaphores
Cycle exists between two transactions, so T1 and R and S, both initialized to zero. The array a is
T2 are not conflict serializable. shared by the two processes. The structures of the
Ans. (b) processes are shown below.
Chapter 7.indd 389 4/9/2015 10:01:44 AM

Process Y: Process Y:
private i; private i; private i; private i;

for (i=0; i for
< n;(i=0;
i++) i{ < n; i++) { for (i=0; i for
< n;(i=0;
i++) i{ < n; i++) {
a[i] = f(i);a[i] = f(i); EntryY(R, S);
EntryY(R, S);
ExitX(R, S);ExitX(R, S); b[i]=g(a[i]);
b[i]=g(a[i]);
} } } }
Which one of the following represents the Solution: S = 2 (only two processes can run at
CORRECT implementations of ExitX and EntryY? the same time)
(A) (B) x=0
ExitX(R, S) { ExitX(R, S) {
W X Y Z
P(R); V(R);
V(S); V(S); P(S) P(S) P(S) P(S)
} } R(x) R(x) R(x) R(x)
EntryY (R, S) { EntryY(R, S) {
P(S); P(R);
x++ x++ x=x-2 x=x-2
V(R); P(S); W(x) W(x) W(x) W(x)
} }
V(S) V(S) V(S) V(S)
(C) (D)
ExitX(R, S) { ExitX(R, S) { Let X and Y are running:
P(S); V(R); X: R(x) → 0 Y: R(x) → 0
V(R); P(S); x++ = 1 x = x − 2 = −2
} } w(x) =1 w(x) = −2
EntryY(R, S) { EntryY(R, S) { Let X wrote after Y, so x will contain 1.
V(S); V(S); Now when W and Z will execute, W will increment
P(R); P(R); it to 2 and Z will decrement to −1, so maximum
} } possible values is 2.
Ans. (d)
(a) A (b) B
(c) C (d) D Common Data Questions 76 and 77: A computer
(GATE 2013: 2 Marks) uses 46-bit virtual address, 32-bit physical address,
Solution: P(S) is for wait, V(S) is for signal. In and a three-level paged page table organization.
option (c), mutual exclusion is ensured. Deadlock The page table base register stores the base address
also does not occur. Other options are leading either of the first-level table (T1), which occupies exactly
to deadlock or do not ensure mutual exclusion. one page. Each entry of T1 stores the base address
Ans.(c) of a page of the second-level table (T2). Each entry
of T2 stores the base address of a page of the third-
75. A shared variable x, initialized to zero, is operated
level table (T3). Each entry of T3 stores a page
on by four concurrent processes W, X, Y, Z as fol-
table entry (PTE). The PTE is 32 bits in size. The
lows. Each of the processes W and X reads x from
processor used in the computer has a 1 MB 16-way
memory, increments by one, stores it to memory,
set-associative virtually indexed physically tagged
and then terminates. Each of the processes Y and
cache. The cache block size is 64 bytes.
Z reads x from memory, decrements by two, stores
it to memory, and then terminates. Each process 76. What is the size of a page in KB in this computer?
before reading x invokes the P operation (i.e., wait)
(a) 2 (b) 4
on a counting semaphore S and invokes the V oper-
(c) 8 (d) 16
ation (i.e., signal) on the semaphore S after storing
x to memory. Semaphore S is initialized to two. (GATE 2013, 2 Marks)
What is the maximum possible value of x after all Solution: All the page table resides in the physi-
processes complete execution? cal memory (RAM) since the physical address is
(a) −2 (b) −1 (c) 1 (d) 2 32 bit addressable. Each level table holds a 32-bit
(GATE 2013: 2 Marks) (4-byte) address of the RAM.
Chapter 7.indd 390 4/9/2015 10:01:45 AM

Assume the size of a page is x bytes (which we to different sets in the processor cache of this
need to find). As the size of level-1 table is also x computer?
(as given in the question that level-1 table accom- (a) 2 (b) 4
modates exactly a page), it can hold x/4 base (c) 8 (d) 16
addresses of x/4 level-2 tables (which also accom-
modates exactly a page) and each level-2 table (GATE 2013: 2 Marks)
holds base addresses of x/4 level-3 tables. Solution:
The level-3 table has entities equal to 246/x Let page sixe is x bytes.
That is, (x/4)(x/4)(x/4) = 246/x Size of level 1 table is also x (given in the question).
On solving, x = 8192, which is in bytes so 8192/1024 It can hold x/4 base addresses of x/4 level 2 tables
= 8 KB and each level 2 table can hold base addresses of
x/4 level 3 tables.
Ans. (c) The level 3 table has entities equal to 246/x, that is,
77. Consider the same data as in the above question. (x/4)(x/4)(x/4) = 246/x
What is the minimum number of page colours On solving x = 8192 bytes = 8 K bytes
needed to guarantee that no two synonyms map Ans. (c)
PRACTICE EXERCISES
Set 1 6. Paging suffers from the following limitation(s)
1. Context switching is part of (a) Internal fragmentation

(b) External fragmentation
(a) Buffering (c) Middle fragmentation
(b) Process execution state (d) Both internal and external fragmentation
(c) Interrupt handling
(d) Ready state 7. Segmentation is solution to overcome
2. Interrupt is (a) Internal fragmentation
(a) External to the execution of the current
(c) Middle fragmentation
instruction
(d) Both internal and external fragmentation
(b) Associated with the execution of the current
instruction 8. Segmentation suffers from the following
(c) Explicit request limitation(s)
(d) Internal to the execution of the current program
(a) Both internal and external fragmentation
3. Process no longer exist, but it leaves a record for (b) External fragmentation
its parent process to collect, the state of the pro- (c) Middle fragmentation
cess is (d) Internal fragmentation
(a) spawned. (b) slipend.
(c) zombie. (d) running. 9. Physical base address of each page in physical
memory is associated with?
4. In round-robin scheduling algorithm, throughput
will (a) Main memory
(b) Virtual memory
(a) Remain as it is (c) Page table
(b) Will be decreased (d) Cache memory
(c) Will be increased
(d) It depends, may either increase or decrease 10. External fragmentation can be overcome by
5. Paging is a solution to overcome 1. Compaction
2. Paging
(a) Internal fragmentation 3. Segmentation
(c) Middle fragmentation (a) 1 and 2 (b) 2 and 3
(d) Both internal and external fragmentation (c) 1 and 3 (d) 1 and 2 and 3
Chapter 7.indd 391 4/9/2015 10:01:45 AM

11. In semaphores, the term P and V stands for (c) M

ay change the state of the interrupt process
to block and schedule another process
(a) Wait, Signal
(d) Informs the process control block (PCB)
(b) Signal, Wait
(c) Both a and b 20. Disk I/O is measured in terms of
(d) None
(a) Blocks (b) Bits
12. The time taken to move the desired sector under (c) Sectors (d) Tracks
the R/W head is called 21. Simplest method for deadlock recovery is
(a) Seek time delay (a) Roll back the processes one by one
(b) Rotational time delay (b) Pre-empt the resource
(c) Transfer time delay (c) Lock one of the resources
(d) Queuing time delay (d) Kill one of the processes
13. Banker’s algorithm is applied for 22. Let the time taken to switch between user mode
and kernel mode of execution be T1 while time
(a) Deadlock detection
taken to switch between the user mode1 and user
(b) Deadlock avoidance
mode2 be T2. Which of the following is true?
(c) Deadlock recovery
(d) Both a and c (a) T1 < T2
(b) T1 > T2
14. Solution to the thrashing is (c) T1 = T2
(a) Working set model (d) T1 and T2 cannot be compared
(b) Increasing the degree of multiprogramming 23. A thread is usually defined as a light-weight pro-
(c) Increasing the size of page table cess because an operating system (OS) maintains
(d) Decreasing the size of the page smaller data structure for a thread than for a pro-
15. Consider a memory paging system where TLB cess. In relation to this, which of the following
access time is 20 ns and main memory access time statement is correct?
is 200 ns. If the hit ratio in TLB is 90%, find the (a) O S maintains only scheduling and virtual
effective access time in ns is ––. memory information for each thread.
(b) OS maintains only CPU registers for each
16. In the following resource allocation graph, every thread.
resource has single instance, then cycle means (c) OS does not maintain a separate stack for each
thread.
(a) Deadlock exists (d) OS does not maintain virtual memory state for
(b) Deadlock may exists each thread.
(c) Deadlock can be avoided
(d) Deadlock can be recovered 24. The Windows operating system such as Windows
XP uses the following file system:
17. In a resource allocation graph, every resource has
multiple instances, then cycle means (a) FAT-16, NTFS
(b) FAT-32, NTFS
(a) Deadlock exists (c) NTFS (NT File System) only
(b) Deadlock may exists (d) FAT-16, FAT-32 and NTFS
(c) Cycle cannot exists
(d) Deadlock can be avoided 25. Which of the following statements is not true about
disk-arm scheduling algorithms?
18. Semaphores provide a solution for
(a) S STF (shortest seek time first) algorithm
(a) Mutual exclusion increases performance of FCFS.
(b) Process synchronization (b) The number of requests for disk service is not
(c) Circular wait influenced by file allocation method.
(d) None of the above (c) Caching the directories and index blocks in
main memory cannot help in reducing disk arm
19. When an interrupt occurs, an operating system movements.
(a) Ignores the interrupt (d) SCAN and C-SCAN algorithms are less likely
(b) Stops the execution of the current instruction to have a starvation problem.
Chapter 7.indd 392 4/9/2015 10:01:45 AM

26. maintains the list of free disk blocks (a) 4 (b) 2

in a Unix operating system. (c) 3 (d) 1
(a) i-Node 33. Fork system call creates the child process on suc-
(b) Boot node pointer cessful completion of fork, which of the following
(c) Super block value is returned in the child process?
(d) File allocation table (a) p_id of parent process
27. A part of Windows 2000 operating system that is (b) p_id of child process
not transportable is (c) p_id of both the child process and parent process
(d) zero
(a) Device Management Scheme
(b) Virtual Memory Management Scheme 34. Suppose that a process is in blocked state waiting
(c) Process Management Scheme for some I/O devices, now when the I/O device is
(d) Graphical User Interface Scheme done with the serving to a process, it was earlier
doing, the process changes its state to the
28. Windows 32 API supports
(a) Running state
(a) 16 Bit (b) Ready state
(b) 32-Bit and 64-bit Windows (c) Suspended state
(c) Only 32-bit Windows (d) Terminated state
(d) 16-Bit, 32-bit and 64-bit Windows 35. What is shell?
29. How an operating system does make sure that a par- (a) It is hardware component.
ticular program has had a successful completion? (b) It is a command interpreter.
(a) The return value is 0. (c) It is a compiler component.
(b) The return value is 1. (d) It is a tool used for CPU scheduling.
(c) The return value is −1. 36. When an OS creates a process at the explicit
(d) The program does not return a value. request of another process, this action is known as
30. Consider the following program (a) Process cloning
(b) Process slipping
main() (c) Process spawning
{ (d) Process borne
fork();
fork(); 37. In threads, which mapping is called pure user level
fork(); threads?
} (a) 1:1 Mapping
How many new processes will be created? (b) N:1 Mapping
(c) N:M Mapping
(a) 9 (b) 6
(d) Null Mapping
(c) 7 (d) 5
38. Which type of scheduler controls degree of
31. A fence register is basically used for multiprogramming?
(a) Memory protection (a) Long-term scheduler
(b) Block protection (b) Short-term scheduler
(c) File protection (c) Medium-term scheduler
(d) CPU protection (d) Hybrid-term scheduler
32. Consider the following program 39. Consider the following program
main() P1 P2
{ P(mutex1) P(mutex2)
fork(); P(mutex2) P(mutex1)
exec(“c:/notepad.exe”); . .
fork(); . .
} . .
V(mutex1) V(mutex2)
How many new processes will be created? V(mutex2) V(mutex1)
Chapter 7.indd 393 4/9/2015 10:01:46 AM

Using this program (a) Safe sequence exists

(a) Deadlock cannot occur. (b) Safe sequence does not exists
(b) Deadlock can occur. (c) Insufficient data
(c) Data is insufficient. (d) Deadlock recovery does not exist
(d) Deadlock recovery cannot be done. 47. Consider a process of 256 bytes and a system
40. If thread library is implemented in a kernel space, having main memory of size 1024 bytes with frame
then each function call will be considered as a size of 8 bytes.
(a) System call 47a. The number of pages will be:

(b) Local function call (a) 16 (b) 32
(c) Both a and b (c) 64 (d) 128
(d) User call
47b. How many frames are there in main memory?
41. An OS has to schedule three user processes each
requiring 2 units of a particular resource R. The (a) 128 (b) 256
minimum number of units of R such that no dead- (c) 512 (d) 64
locks will ever arise is 47c. How many entries are there in page table?
(a) 3 (b) 2 (a) 16 (b) 32
(c) 4 (d) 6 (c) 64 (d) 8
42. Which of the following need not be saved on con- 47d. What could be the size of page table?
text switch between the various processes?
(a) 124 bits (b) 224 bits
(a) General purpose register (c) 145 bits (d) 128 bits
(b) Translation look-aside buffer
(c) Program counter 48. Consider the following snapshot of an operating
(d) Stack register system required to schedule the execution of the
processes:
43. A computer hardware has N number of CPUs,
then the maximum number of processes that can Processes Arrival CPU Time Priority
be in running state on this hardware will be Time Needed
(in ms) (in ms)
(a) N/2 (b) N!
(c) N (d) log2N P1 0 10 5
44. Page swapping and page replacement is also called P2 0 5 2

P3 2 3 1
(a) Page cannibalizing
(b) Page copying P4 5 20 4
(c) Page snooping P5 10 2 3
(d) Page sharing
48a. If CPU scheduling policy is FCFS, then average
45. The commonly used term PCB means
waiting time will be
(a) Program control block
(a) 12.8 ms (b) 8 ms
(b) Process control block
(c) 16 ms (d) 15 ms
(c) Process communication block
(d) Program communication block 48b. If scheduling policy is SJF without pre-emption,
then average waiting time will be
46. Consider the following scenario, does it represent a:
(a) 12.8 ms (b) 6.8 ms
Process Allocated Need Available (c) 17 ms (d) 15 ms
ABC ABC ABC 48c. If CPU scheduling policy is SJF with pre-emption,
P1 010 135 020 then average waiting time will be
P2 203 542 (a) 8 ms (b) 14 ms
(c) 5.6 ms (d) 15 ms
P3 323 241
P4 212 030 48d. If CPU scheduling policy is priority scheduling with-
out pre-emption, then average waiting time will be
Chapter 7.indd 394 4/9/2015 10:01:46 AM

(a) 12.8 ms (b) 11.8 ms What will be the safety sequence of the above pro-
(c) 10.8 ms (d) 14 ms cesses to avoid the deadlock?
48e. If CPU scheduling policy is priority scheduling with (a) P0, P3, P1, P2 (b) P0, P1, P2, P3
pre-emption, then average waiting time will be (c) P0, P1, P3, P2 (d) P0, P2, P1, P3
(a) 19 ms (b) 7.6 ms 5. Assume the processes are arrived into the ready
(c) 8 ms (d) 14.2 ms queue in the correct order in the above question
at time to. What will be the average waiting time
using round-robin scheduling with a time quantum
Set 2 5 ms? (The CPU burst requirements of P0, P1, P2,
P3 are 5 ms, 10 ms, 7 ms and 8 ms, respectively.)
1. Which scheduling policy is most suitable for time-
shared operating systems? (a) 14.5 ms (b) 13 ms
(c) 12.5 ms (d) 15 ms
(a) Shortest job first
(b) Round robin 6. When the result of a computation depends on the
(c) First come first serve speed of the processes involved, there is said to be
(d) Elevator (a) Cycle stealing (b) Race condition
2. Four jobs to be executed on a single processor (c) A time lock (d) A deadlock
system arrive at time 0+ in the order A, B, C, D. 7. Each process Pi, I = 1 … 9 is coded as follows:
Their burst CPU time requirements are 4, 1, 8, 1
time units, respectively. The completion time of A Repeat
under round-robin scheduling with time slice of one P (mutex)
time unit is . {critical section}
V (mutex)
3. The sequence is an optimal non-preemptive sched-
uling sequence for the following jobs which leaves Forever
the CPU idle for unit(s) of time. The code uses P10 is identical except that it uses
V (mutex) in place of P (mutex). What is the larg-
Job Arrival Time Burst Time est no of processes that can be inside the critical
section at any moment?
1 0.0 9
(a) 1 (b) 2
2 0.6 5 (c) 3 (d) None of these
3 1.0 1 8. Consider Peterson’s algorithm for mutual exclusion
between two concurrent processes i and j. The pro-
(a) {3,2,1},1 gram executed by process is shown below:
(b) {2,1,3},0
Repeat
(c) {3,2,1},0
(d) {1,2,3},5 Flag[i] = true;
Turn=j;
4. The Banker’s algorithm for avoiding the deadlock
While (P) do no -op;
has got the following data structure, with four pro-
cesses P0,P1,P2,P3 and resourcetypes A, B, C and D. Enter critical section, perform actions,then
Then
Exit critical section
Process Alloca- Maxi- Available Flag[i]=false;
tion mum Perform other non-criticalsection actions.
ABCD ABCD ABCD Until false;
P0 2120 4132 3012 For the program to guarantee mutual exclusion,

the predicate P in the while loop should be
P1 1002 4103
(a) Flag[j]=true and Turn=i
P2 2021 5441 (b) Flag[j]=true and Turn=j
P3 2341 3365 (c) Flag[i]=true and Turn=j
(d) Flag[i]=true and turn=i
Chapter 7.indd 395 4/9/2015 10:01:46 AM

9. Consider an operating system containing `n’ pro- (a) R elease all resources before requesting a new
cesses each requiring four resources. The maximum resource.
number of processes that can be executed without (b) Number the resources uniquely and never
any deadlock if there are 18 resources is request a lower numbered resource than the
last one requested.
10. The average page fault service time in a system is (c) Never request a resource after releasing any
25 ms and memory access time is 220 ns. If page resource.
fault rate is 0.0004, what will be the effective access (d) Request and all required resources be allocated
time? before execution.
(a) 10 (b) 7.36 15. Consider the disk drive with the following
(c) 10.21 (d) 11.36 specification:
Eight surfaces, 1024 tracks/surface, 512 sectors/
11. Consider n processes sharing the CPU in a round- track, 4KB/sector, rotation speed is 3000 rpm and
robin fashion. Assuming that each process switch the disk is operated in burst mode. The processor
takes s seconds, what must be the quantum size runs at 600 MHz and takes 300 and 900 clock cycle
q such that the overhead resulting from process to initiate and complete DMA transfer, respec-
switching is minimized but at the same time each tively, if the size of transferred data is 20KB.
process is guaranteed to get its turn at the CPU at
least every t seconds? What is the size of the hard disk?
(a) 16 GB (b) 8 GB
t − ns t − ns
(a) q ≤ (b) q ≥ (c) 16 MB (d) 8 MB
n −1 n −1
16. What is the percentage of processor time consumed
t − ns t − ns
(c) q ≤ (d) q ≥ for the transfer operation in the above case?
n +1 n +1 (a) 0.69% (b) 27.77%
(c) 97% (d) 20%
12. A solution to the Dining Philosophers Problem
17. Consider a system having m resources of the same
which avoids deadlock is
type. These resources are shared by three processes
(a) Ensure that all philosophers pick up the left A, B and C which have peak demands of 3, 5 and
fork before the right fork 7, respectively. For what value of m deadlock will
(b) Ensure that all philosophers pick up the right not occur?
fork before the left fork (a) 13 (b) 14
(c) Ensure that one particular philosopher picks up (c) 15 (d) 16
the left fork before the right fork, and that all
18. A ROM is used to store the table for multiplication
other philosophers pick up the right fork before
of two 8-bit unsigned integers. The size of ROM
the left fork
required is
(d) None of these
(a) 256 × 16 (b) 64K × 8
13. A computer has six tape drives, with n processes (c) 4K × 16 (d) 64K × 16
competing for them. Each process may need two 19. Dirty bit for a page in a page table
drives. What is the maximum value of n for the
system to be deadlock free? (a) helps avoid unnecessary writes on paging device
(b) helps maintain LRU information
(a) 6 (b) 5 (c) allows only read on a page
(c) 4 (d) 4 (d) none of the above
14. Which of the following is NOT a valid deadlock 20. Consider the following heap (figure) in which blank
prevention scheme? regions are not in use and hatched regions are in use.
50 150 300 350 600
Increasing addresses
Chapter 7.indd 396 4/9/2015 10:01:48 AM

The sequence of requests for blocks of size 300, 25, (a) 7 (b) 8
125, 50 can be satisfied if we use (c) 9 (d) 10
(a) Either first-fit or best-fit policy(any one) 27. If number of frames is 3, then how many more
(b) First-fit but not best-fit policies page faults will occur using the same page reference
(c) Best-fit but not first-fit policy string?
(a) 0 (b) 1
21. The capacity of memory unit is defined by the (c) 2 (d) 3
number of words multiplied by the number of bits/
word. How many separate address and data lines 28. Which of the following disk-scheduling strategies is
are needed for a memory of 4K × 16? likely to give the best throughput?
(a) 10 address, 16 data lines (a) Farthest cylinder next
(b) 11 address, 8 data lines (b) Nearest cylinder next
(c) 12 address, 16 data lines (c) First come first serve
(d) 12 address, 12 data lines (d) Elevator algorithm
22. If an instruction takes i μs and a page fault takes an 29. In the index allocation scheme of blocks to a file,
additional j μs, the effective instruction time if on the maximum possible size of the file depends on
the average a page fault occurs every k instruction is
j (a) T he size of blocks, and the size of the address
(a) i + (b) i + j ∗ k of the blocks.
k
(b) The number of blocks used for the index, and
i+j
(d) (i + j) ∗ k
the size of the blocks.
(c)
k (c) The size of blocks, the number of blocks used
23. Suppose the time to service a page fault is on the for the index and the size of the address of the
average 10 ms, while a memory access takes 1 μs. blocks.
Then a 99.99% hit ratio results in average memory (d) None of the above
access time of 30. Suppose two jobs, each of which needs 10 min of
(a) 1.9999 ms CPU time, start simultaneously. Assume 50% I/O
(b) 1 ms wait time.
(c) 9.999 μs How long will it take for both to complete if they
(d) 1.9999 μs run sequentially?
24. Consider a process that has been allocated 3 page (a) 10 (b) 20
frames. Assume that system uses pure demand (c) 30 (d) 40
paging. While undergoing execution, the process
31. How long will it take if they run in parallel in the
makes the following sequence of page references
above case?
1,4,7,1,6,3,1,3,2,4,9,1.
(a) 10 (b) 20
If optimal page replacement policy is used, then
(c) 30 (d) 40
the number of page faults for the above reference
string is 32. A multilevel page table is preferred in comparison
to a single-level page table for translating virtual
(a) 7 (b) 6
address to physical address because
(c) 8 (d) None of these
(a) It reduces the memory access time for read and
25. If LRU is used for the same reference string in the
write.
above question then which of the following is true?
(b) It helps to reduce the size of page table needed
(a) Optimal has 22% lesser page faults than LRU to implement the virtual address space of a
(b) Optimal has 33% lesser page faults than LRU process.
(c) Optimal has 44% lesser page faults than LRU (c) It is required by translation look-aside buffer.
(d) None of these (d) It helps to reduce the number of page faults in
page replacement algorithms.
26. What is the number of page faults by least recently
used page replacements for a memory with 4 frames 33. A process executes the following code for(i = 0; i <
for the page reference string 2, 0, 1, 2, 4, 0, 5, 1, 4, n; i++) for ();the total number of child processes
6, 4, 2, 1, 3, 0? created is
Chapter 7.indd 397 4/9/2015 10:01:49 AM

(a) n (b) 2n−1 35. Match the following:

(c) 2n (d) 2(n+1)−1
List I List II
34. A UNIX style i-node has 10 direct pointers and one a. Critical region 1. Hoarse monitor
single, one double and one triple indirect pointers. b. Wait/signal 2. Mutual exclusion
Disk block size is 1 KB, disk block address is 32 bit
c. Working set 3. Principal of locality
and 48 bit integers are used. What is the maximum
possible file size? d. Dead lock 4. Circular wait
(a) 224 bytes (b) 232 bytes (a) 2 1 3 4 (b) 1 2 4 3
(c) 234 bytes (d) 248 bytes (c) 2 3 1 4 (d) 1 3 2 4
Set 1
1. (c) 12. (b) 23. (b) 34. (b) 45. (c)

2. (a) 13. (b) 24. (d) 35. (b) 46. (b)
3. (c) 14. (a) 25. (b) 36. (c) 47a. (b)
4. (b) 15. (240) 26. (c) 37. (a) 47b. (a)
5. (b) 16. (a) 27. (b) 38. (a) 47c. (b)
6. (a) 17. (b) 28. (d) 39. (b) 47d. (b)
7. (a) 18. (b) 29. (a) 40. (a) 48a. (a)
8. (b) 19. (c) 30. (c) 41. (c) 48b. (b)
9. (c) 20. (a) 31. (a) 42. (b) 48c. (c)
10. (a) 21. (d) 32. (b) 43. (c) 48d. (c)
11. (a) 22. (a) 33. (d) 44. (a) 48e. (b)
Set 2 4. (c) Refer text

5. (a) Refer text
1. (b) To schedule processes fairly, a round-robin
scheduler generally employs time-sharing, giving 6. (b) The condition is said to be race condition.
each job a time slot or quantum (its allowance of
7. (d) Every time when process P10 enters into criti-
CPU time), and interrupting the job if it is not
cal section, it performs the ÙP’ operation, so after
completed by then. It is designed especially for
that one more processor can enter into it. This
time-sharing systems.
process continues till end and so 10 processes can
2. (9) Draw Gantt chart according to the round-robin enter into critical section at a time.
algorithm with time quantum 1.
8. (b) To ensure the mutual exclusion into this pro-
0 1 2 3 4 5 6 7 8 9 10 gram, we have to take care that `j’ should not be
A B C D A C A C A C allowed to value `turn’.
So, completion time of A is 9. 9. (7) 5n ≤ n + 28 ⇒ n = 7 satisfy the above equation.
3. (a) In the first sequence (3, 2, 1), CPU is ideal for 10. (c) Effective access time = [(1 − p) × Memory
1 unit time because 3rd process arrives at time 1. access time + p × page fault service time]
Process with large burst time will not have to wait ⇒ [(1 − 0.0004) × 0.22 + 0.0004 × 2500] μs
more so CPU will not be ideal more time. ⇒10.219
Chapter 7.indd 398 4/9/2015 10:01:49 AM

11. (a) Next execution of the process should occur 24. (d)
within t seconds. Each process runs for q period, 1 7 7 1 6 3 1 3 2 4 9 1
and if there are n processes p1 p2 p3 … pn p1 p2 … 1 1 1 1 1 1 1 1 1 1 1 1
then p1 turn comes again when it has completed 4 4 4 4 4 4 4 4 4 9 9
time quanta for remaining process p2 to pn, that is, 7 7 6 3 3 3 2 2 2 2
it would take at most (n − 1)q time. F F F H F F H H F H F H
So each process in round-robin gets its turn after
≤ (n − 1)q time when we do not consider over- 25. (a)
heads, but if we consider overhead(s) then it would
1 4 7 1 6 3 1 3 2 4 9 1
be ns + (n − 1)q. 1 1 1 1 1 1 1 1 1 4 4 4
So, we have ns + (n − 1)q ≤ t overhead will be 4 4 4 6 6 6 6 2 2 2 1
reduced when time quantum is maximum allow- 7 7 7 3 3 3 3 3 9 9
able, that is, q = (t − ns)/(n − 1) F F F H F F H H F F F H
12. (a) It is a condition to implement correctly Dining
Philosopher Problem. We have to start checking 26. (d)
from left fork to the right fork.
String 2 0 1 2 4 0 5 1 4 6 4 2 1 3 0
13. (a) Every process needs two resources to complete
its task. If we allocate five resources to five pro- Frame 1 2 2 2 0 0 1 2 4 0 5 5 1 6 4 2
cesses as one process each, then no one process can Frame 2 0 0 1 1 2 4 0 5 1 1 6 4 2 1
finish its task. But if we give one more resource to
Frame 3 1 2 2 4 0 5 1 4 6 4 2 1 3
any process then it will complete its task and free
its resources to be used by the other. By which all Frame 4 4 0 5 1 4 6 4 2 1 3 0
process can finish their tasks. So answer is 6.
Bolds are representing misses. So, total 10 misses.
14. (c) This is not a valid deadlock prevention scheme.
27. (d)
15. (a) Disk size = 8 × 1024 × 512 × 4 K = 16 × 230
bytes = 16 GB String 2 0 1 2 4 0 5 1 4 6 4 2 1 4 0
16. (b) Refer text Frame 1 2 2 2 0 1 2 4 0 5 1 1 6 4 2 1
Frame 2 0 0 1 2 4 0 5 1 4 6 4 2 1 4
17. (d) The peak demand for each processes A, B and
C is given as 3, 5 and 7, respectively. Suppose the Frame 3 1 2 4 0 5 1 3 6 4 2 1 4 0
peak demand for all the three processes happens
simultaneously, then we will require 3 + 5 + 7 = Bolds are representing misses. So 13 misses. So dif-
15 resources to satisfy the demand. ference is 13 − 10 = 3.
18. (d) 28 × 28 = 64K × 16 28. (b) Nearest cylinder next approach gives the better
results.
19. (d) The pages which have dirty bits are the pages
written back in memory, which avoids unnecessary 29. (c) Number of disk block possible in one disk block
writing.
 DBA size 
= 
20. (b) Refer text  DBA 
21. (c) 4K = 212 → 12 addresses
Each process takes total of 10 min for I/O +
22. (c) Calculate by the given formula (EMAT = p × s 10 min for CPU, that is, 20 min
+ (1 − p) × m) Total time for complete = 20 min + 20 min
where p = page fault rate, s = page fault service = 40 mins
time, m = memory access time. 31. (b) Probability p for which time is spent in I/O
23. (a) Calculate by the given formula: [(% of Page miss × Page
p =fault
1/2 sorervice
50% time)
I/O
+ (% of Page hit × Memory access time)]
CPU utilization = 1 − p2 = 0.75
Average memory access time = T × 0.75 = 20
100
[(% of Page miss × Page fault service time) 32. (b) The use of multilevel page table reduces the
+ (% of Page hit × Memory access time)] size of the page table needed to implement the vir-
memory access time =
100 tual address space of a process.
Chapter 7.indd 399 4/9/2015 10:01:50 AM

33. (b) Total number of process is 2n, but we subtract 35. (a)
the main process, therefore total number of chil- (a) Critical region: Non-sharable region of code.
dren = (2n−1) (b) In concurrent programming, mutual exclusion
34. (c) Number of disk block pointer that will fit in one is achieved by semaphore.
block = 210/4 = 256 (c) Hoare’s monitor: Monitors used for synchroni-
zation were invented by C. A. R. Hoare.
Maximum file size = 256 × 256 × 256 × 1 KB (d) condition fordeadlockto occur: Hold and wait,
= 234 bytes circular wait, mutual exclusion, no pre-emption
Chapter 7.indd 400 4/9/2015 10:01:50 AM

UNIT VIII: DATABASES
8
7
Number of questions
6
5 Marks 1
4 Marks 2
2
1
0
2015 2014 2013 2012 2011 2010
Year Concepts
2015 SQL, ER Diagram, Disk, Functional Dependency, Transactions, B+ Trees,
Transaction Properties, Transaction Recovery
2014 Relational algebra, SQL query, Serializability, Attributes of relational schema
2013 Indexing, Relational schemas, Normalizations, Functional dependencies
2012 ER and relational models, SQL query, Normalizations, Relational algebra
2011 SQL query, Indexing
2010 B+ tress, Relational schema, SQL query, Concurrency control protocols,
Functional dependencies
Chapter 8.indd 401 4/9/2015 10:03:29 AM

Chapter 8.indd 402 4/9/2015 10:03:29 AM
CHAPTER 8
DATABASES
Syllabus: ER model, relational model (relational algebra, tuple calculus), database design (integrity constraints,
normal forms), query languages (SQL), file structures (sequential files, indexing, B and B+ trees), transactions and
concurrency control.
8.1 INTRODUCTION 8.1.1 Traditional File Processing Approach
File processing approach is generally more accurate

Database is a collection of organized information so and faster than the manual database system. Each
that it can easily be searched, retrieved, managed and user is responsible for the defining and implementation
updated. A database management system (DBMS) is of the files required for the specific application. The
a set of programs designed to manage a database. It implementation of required files sometimes creates
enables users to store, retrieve and modify information redundancy of data, for example, one user keeps records
in a database with utmost efficiency along with security for the savings account of the customer and another user
features. DBMS is applicable to various day-to-day may create the loan account of the same customer. This
fields such as transactions in banking; airline/railway/ causes the duplicity of the records of the same customer.
hotel reservations; maintenance of student information So, this practice is not feasible for real time applications.
in schools/universities; online retails; marketing and There is a need of centralized management of data. The
sales etc. It also allows its users to create their own data is created once and then accessed by different users.
databases. Different types of DBMS are available such The data should be shared for different transactions. It
as hierarchical, network, relational and object oriented. should be self-describing in nature, which means the
Chapter 8.indd 403 4/9/2015 10:03:29 AM

404 Chapter 8: Databases
database system contains not only data but it describes ••Operating system software like Microsoft Win-
the description of the database structure. dows, Linux OS, Mac OS.
••DBMS software such as Oracle 8I, MySQL,
8.1.2 Database Management System Access (Jet, MSDE), SQL Server etc.
••Network softwares are used for sharing share the
A database is a collection of related data. Data is a col- data of database among multiple users.
lection of raw facts or figures, processed to form infor- ••Application programs are developed like C++,
mation. Database management system is a collection of
VB, dotnet etc. are used to access database in
programs for the creation and maintenance of database.
dbms. These are used to access and manipulate
It is an efficient and reliable approach to retrieve data
the data in the database.
for many users. It provides various functions such as:
1. Redundancy control: It provides redundancy 2. Hardware: It consists of all system’s physical
by removing duplicity of data by following rules of devices such as computers, storage devices, I/O
normalization. channels, electromechanical devices etc. It also
2. Data independence: It provides independence includes peripherals, such as, keyboard, mouse,
to application programs from details of data repre- modems, printers, etc.
sentation and storage. It also provides an abstract 3. Data: It is the collection of facts. The database
view of the data to insulate application code from contains the data and the metadata.
such details. 4. Procedures: There are the instructions and rules
3. Data integrity: It promotes and enforces some to design and use the database system. These
integrity rules for reducing data redundancy and includes the following:
increasing data consistency. ••Steps for the installation of DBMS
4. Concurrency control: It supports sharing of ••Steps to use the DBMS or application program
data, so, it has to provide an approach for managing
••Steps for the backup of DBMS
concurrent access of the database. Hence, preserv-
ing the inconsistent state and integrity of the data. ••Steps to change the structure of DBMS
5. Transaction management: It provides an ••Steps for the generation of reports.
approach to ensure that either all the updates for a
5. Data access language: The users can use it to
given transaction will execute or that none of them
access the data to and from the database. The func-
would execute.
tion of data access language is the entry of new data,
6. Backup and recovery: It provides mechanisms
manipulation of the existing data and the retrieval of
for backing up data periodically and recovering
the existing data in the database. The most popular
from different types of failures, thus, preventing
database access language is SQL (Structured Query
loss of data.
Language). Users can perform these functions with
7. Non-Procedural query language: It provides
the help of commands. The role of administrator is
with query language for retrieval and manipulation
to access, to create and to maintain the database.
of data.
6. People: Persons involved to access, to create and
8. Security: It protects unauthorized access in the
to maintain the database are called users. These
database. It ensures the access to authorized users.
are of various types according to the role performed
by them (Fig. 8.1). These are as follows:
8.2 COMPONENTS OF DATABASE ••System Administrator: The role of system
SYSTEMS a dministrator is to supervise the general opera-
tions of DBMS.
••Database Administrator: The role of database
DBMS consists of several components, namely software,
hardware, data, procedures and data access language. administrator (DBA) is to manage the DBMS.
These components are responsible for the definition, col- ••Database Designer: The role of database
lection, management and use of data within the envi- designer is to design the structure of the database.
ronment. Figure 8.1 shows the components of database ••Application Programmer: The role of
system. The description of each component is as follows: application programmer is to create the data
1. Software: It is the collection of programs used entry forms, reports and procedures.
by the computers within the database system. It is ••End User: The role of end user is to use the
used to handle, control and manage the database. application programs by entering new data and
It includes the following software: manipulating and accessing existing data.
Chapter 8.indd 404 4/9/2015 10:03:29 AM

8.4 DATA MODELS 405
1. Internal schema or physical level: It is also

End Database Application
administrator programmer called Database Tier. Database exists in this tier. It
user
also includes query processing languages, all relations
and their corresponding constraints. It describes the
physical storage structure of the database.
2. Conceptual schema or logical level: It is also
Query called Application Tier. Server and program exists
processing in this tier. It also describes the structure of the
database.
3. External schema or view level: It is also called
Database Presentation Tier. End user exists in this tier. An
manager
end user is capable of multiple views of database.
It also includes all views generated by applications.
File
manager 8.3.2 Data Independence
Data independence is defined, as the change at one level

does not affect the higher level. It is of two types:
Database Metadata
1. Logical data independence: It is defined, as
the change in conceptual schema does not affect
Figure 8.1 | Components of database systems. the external schema. For example, if the format of
the table changes, the data lies in the table should
not be changed.
2. Physical data independence: It is defined, as
8.3 DBMS ARCHITECTURE
the change in internal schema does not affect the
conceptual schema. Thus, does not affect the exter-
It is an approach to outlook the database by users. It is nal schema. For example, if the storage system has
means for the representation of data in an understand- been changed, then it does not affect the logical
able way to the users. DBMS architecture can be used structure of the database.
to divide the whole system to related and independent
modules. It can be of 1-tier, 2-tier 3-tier or n-tier. 8.4 DATA MODELS
8.3.1 3-Tier Architecture
Data model is a collection of tools, which describes data,
DBMS can be most widely used as 3-tier architecture. In relationships, constraints and semantics. It gives the logi-
this architecture, the database is divided into three tiers cal structure of the database. It describes the relationship
depending upon the kind of users (see Fig. 8.2). of data in the database. Data models are of various types:
1. Relational Model: It is a collection of relations
(tables). In relational model, each table is stored as
a separate file.
View level 2. Entity—Relationship data model: This model
is based on the notion of real-world entities and
relationship among them.
3. Object-Based data model: It defines the
database as objects, its properties and its operations.
Logical level In this model, objects with the similar structures
comprise a class. The classes are organized into
hierarchies. The operations on these classes are
performed through methods.
4. Semi-Structured data model: It is also known
Physical level as XML model. This model is used to exchange
data over the web. It uses hierarchical tree struc-
Figure 8.2 | The 3-tier architecture of DBMS.

tures. In this model, data can be represented as
elements by using tags.
Chapter 8.indd 405 4/9/2015 10:03:30 AM

5. Network model: In this model, data is repre- Table 8.1 | Notations of fundamental operations
sented as record types. The data in this model has of relational algebra
many-to-many relationship.
6. Hierarchical model: In this model, the data is Fundamental Operation Symbol
represented as a hierarchical tree structure. Select s
Project p
8.4.1 Relational Model
Union ∪
The database in relational model is represented as a col- Intersection ∩
lection of relations (tables). A relation is a kind of set. It Set difference −
is also a subset of a Cartesian product of an unordered
set of ordered tuples. Relational model was proposed by Cartesian product ×
E. F. Codd, which stores data in a tabular form. It con- Rename r
sists of a table where rows represent records and columns Natural join 
represent the attributes. It has various terminologies as
follows:
Fundamental operations of relational algebra are as
1. Tuple: It represents a single row of a table, which follows (see Table 8.1):
contains a single record for that relation.
1. Select: It is used to select rows from a relation. It
2. Relation instance: It represents a finite set of
is denoted by s.
tuples in the relational database system.
3. Relation schema: It represents the relation Syntax of select s p (r), r is relation and p is prepo-
name, that is, table name, attributes and their sitional logic.
names. p uses connectors and operators Ù, Ú, =, ≠,<,>, £, ³.
4. Relation key: It represents the unique key for For example, s empname = "John" (emp).
the relation or table. Each row has one or more
attributes, which can identify the row in the table 2. Project: It is used to project columns in a relation.
uniquely. It is denoted by p . The duplicate tuples are auto-
5. Attribute domain: It represents the predefined matically eliminated.
value scope of each attribute. Syntax of projects p A (r), r is relation and A is the
attribute name in a relation.
8.4.1.1 Constraints in Relational Model For example, p empname, sal (emp)
Constraints are the restrictions that one wishes to apply 3. Union: It returns a relation instance, which con-
on database. The following constraints are applied on tains all tuples occurring in the first relation or in
relational model. the second relation. It is denoted as R∪S , where R
and S are two relations. The duplicate tuples are
1. Key constraints: Each relation has at least one automatically eliminated.
minimal subset of attributes, which can identify a This operation is valid for the following:
tuple uniquely. ••Both relations must have the same number of
••No two tuples have identical value for key at-
attributes.
tributes. ••Attribute domains must be compatible.
••Key attribute does not have NULL value.
4. Intersection: It returns a relation instance,
2. Domain constraints: Attributes have specific which contains all tuples occurring in both
domain values in real world. For example, value of relations. It is denoted as R ∩ S , where R and S
age can only be positive. are two relations.
3. Referential integrity constraints: If a relation
5. Set difference: It returns a relation instance,
refers to a key attribute of a different relation then
which contains all tuples that occur in the first
that key element must exist.
relation but not in the second relation. It is denoted
as R − S , where R and S are two relations.
8.4.1.2 Relational Algebra 6. Cartesian product: It returns a relation
instance, which contains all the fields of the first
It is a procedural query language. It takes instances of relation followed by all the fields of the second
relations as input and returns instances of relations as relation. It is denoted as R×S , where R and S are
output. Operators are used to perform queries. two relations.
Chapter 8.indd 406 4/9/2015 10:03:35 AM

8.4 DATA MODELS 407
7. Rename: It returns a relation but without any { t | Condition}, wherte t is a tuple variable and Condition
name. It is used to rename the output relation. It is a conditional expression.
is denoted as r. Preliminaries of tuple calculus are as follows:
8. Joins: It returns combined information from two
or more relations. 1. Constants
2. Predicates
••Condition joins: It accepts a join condition c 3. Boolean and, or, not
and a pair of relation instances as arguments, 4. $ there exists
and returns a relation instance. It is denoted as 5. " for all
s c (R×S ) .
••uijoin: It is a special case of condition joins
8.4.2 ER Model
where the condition c contains equalities.
••Natural join: It is a Cartesian product of two
ER model represents the conceptual view of a database.
relations. It is denoted by . It describes the relation of data to each other (Table 8.2).
It was developed by Peter Chen in 1976. It views real-
8.4.1.3 Tuple Calculus world data as systems of entities and relationships.
ER model has three basic elements: entity, attribute
It is a non-procedural query language. In this, number and relationship. These are discussed in the following
of tuple variables is specified. It is represented as sections.
Table 8.2 | Entity-relationship objects

Element Symbol Example
Entity
Student Employee
Weak entity
Order Order item
Attribute
Name
Simple attribute
Order
Composite attribute Last_name

First_name
Name
Derived attribute
Date−of−birth Age
Single-valued attribute
Roll_number
Multi-value attribute
Phone_number
(Continued)
Chapter 8.indd 407 4/9/2015 10:03:40 AM


Element Symbol Example
Relationship
One-to-One 1 1
relationship Department has Manager
One-to-Many
1 N
relationship Manager manages Employee
Many-to-One N 1
relationship Employee works Department
Many-to-Many N N
relationship Employee handles Projects
8.4.2.1 Entity For example, Details of Employee’s spouse, order item, etc.
Entities represent the real-world things. These are data Order Order item
objects which maintain different relationships with each
other, for example, Employee, Department, etc. These
are represented by means of rectangles.
8.4.2.2 Attributes
Attribute is a property, trait or characteristic of an

entity, relationship or another attribute. All attributes
have values. A domain or range of values can be assigned
For example, to attributes. For example, name, class, age are attributes
of the entity student. These are represented by ovals.
Student Employee
Entity set is a collection of similar types of entities. For

example, all Employees set may contain all employees
of all departments. For example,
Weak entity depends on the existence of another
entity. A weak entity cannot be identified by its own
attributes. It uses a foreign key combined with its attri- Name
butes to form the primary key. It is represented by
means of double rectangles.
There are different types of attributes, listed as follows:
1. Simple attribute: Simple attributes contains

atomic values. Atomic values cannot be divided
Chapter 8.indd 408 4/9/2015 10:03:42 AM

8.4 DATA MODELS 409
into sub parts. Examples are mobile number, roll For example contact number, email ids.
number.
Order Phone−number
2. Composite attribute: Composite attributes are

composition of many simple attributes. For exam- 8.4.2.3 Relationships
ple address can be divided into house number,
street number, locality and city. It represents the association among entities in a specified
way. For example, employee entity has relation works at
with department entity. Relationships are represented
First−name Last−name by diamond-shaped boxes.
Name
3. Derived attribute: Derived attributes are those Some basic terminologies related to relationship are given
whose value is derived from some other attribute below.
in the database. For example, age of person can be As we have discussed, relations are the core compo-
calculated from date of birth. Dotted oval is used nents in RDBMS, these relations are defined by two
to represent derived attributes. major characteristics–relationship set and the degree of
relationship, defined in the following text.
1. Relationship set: Relationship of similar type
is called relationship set. It has attributes. These
attributes are called descriptive attributes.
2. Degree of relationship: It defines the number
For example, of participating entities in a relationship. They are
of the following types:
••Unary relationship (Degree 1): One entity
participates. For example,
Date−of−birth Age
Manages
4. Single-valued attribute: Single valued attri- Managed by

butes contain only one value for that attribute. For
example age for person, blood group. Employee
Roll−number ••Binary relationship (Degree 2): Two entities

participate. For example,
5. Multi-value attribute: In this, an attribute may Received

contain more than one value. Multiple values are Customer Invoice
represented by double ovals.
Sent to
••Ternary relationship (Degree 3): Three entities

participate. For example,
Chapter 8.indd 409 4/9/2015 10:03:45 AM

of generalization. For example, employees are

Student specialized as salaried and contract as shown in
Fig 8.3.
Specialization
Course Registration Faculty Employee
••n-ary relationship (Degree n): n entities participate.
3. Cardinality: It defines the number of instances of Contract

Salaried
an entity, which can be associated to the number of
Generalization
instances of other entity via relationship set.
••One to one: One instance of an entity is associ- Figure 8.3 | Specialization and Generalization.
ated with at most one instance of another entity 6. Aggregation: It allows a relationship set partici-
with the relationship. It is represented as `1-1’. pate in another relationship set (see Fig. 8.4).
For example,
1 1
Department has Manager Employee works Department
••One to many: One instance of an entity is

a ssociated with more than one instance of
another entity with the relationship. It is repre- handles Projects
sented as `1-N’. For example,
1 N Figure 8.4 | Aggregation.

Manager manages Employee
••Many to one: More than one instance of an 8.5 DATABASE DESIGN

e ntity is associated with another entity with
the relationship. It is represented as `N-1’. For
The design of a database consists of the following steps:
example,
1. Identifying entities
N 1 2. Identifying relationships
Employee works Department 3. Identifying attributes
4. Presenting entities and relationships: Entity rela-
••Many to many: More than one instance of an
tionship diagram (ERD)
5. Assigning keys
entity is associated with more than one instance
6. Defining the attribute’s data type
of another entity with the relationship. It is
7. Normalization
represented as `N-N’. For example,
8.5.1 Integrity Constraints

N N
Employee handles Projects
Integrity constrains are applied to maintain the consis-
tency in a database. This helps in providing the unique
4. Generalization: It is a collection of entity sets, answer to a given query on the database. For example,
having similar characteristics, brought together if the answer to particular query is `x’ then it should be
into one generalized entity. For example, salaried `x’ if such a query is carried out again (without adding/
and contract employees are generalized as deleting/modifying) on the same table.
employee. 1. Domain integrity: It defines a valid set of values
5. Specialization: It is the process of identifying for an attribute, for example, length or size, data
subsets of an entity set. It is a reverse process type, etc.
Chapter 8.indd 410 4/9/2015 10:03:46 AM

8.5 DATABASE DESIGN 411
2. Entity integrity constraint: It defines that Also, sometimes called, Project-Join Normal Form
the primary keys cannot be null. There must be a (PJNF).
proper value in the primary key field.
3. Referential integrity constraint: It is speci- 8.5.3 Attribute Closure
fied between two tables. It is used to maintain
the consistency among rows between the two Set of all attributes functionally determined by X is
tables. called closure of X. Closure set of X is denoted by X+.
4. Foreign key integrity constraint: There are
two types of foreign key integrity constraints:
Problem 8.1: Functional dependencies F = {A → B,
••Cascade update related fields: Whenever the B → C, C → D} is given, find closure of A, B, C
primary key of a row in the primary table is and D.
changed, the foreign key values are updated in
the matching rows in the related table. Solution:
••Cascade delete related rows: Whenever a row in
Closure of A = A+ = (A, B, C, D)
the primary table has been deleted, the match-
ing rows are automatically deleted in the related Closure of B = B+ = (B, C, D)
table. Closure of C = C+ = (C, D)
Closure of D = D+ = (D)
8.5.2 Normal Forms
Normalization is a technique of organizing the database 8.5.4 Key

tables, such that they have minimum redundancy.
Following are the normal forms that are to be achieved It is a minimum set of attributes used to differentiate all
for the process of normalization. The underlying concept the tuples of the table.
is that, if we say a table to be satisfying an `x’ level
normal form, then it is understood that it has also satis-
fied the `x-1’ level normal form. 8.5.4.1 Superkey
1. First normal form: It defines that all the attri- It is a set of attributes that uniquely identifies each
butes in a relation must have atomic domains. record in a table. It is a superset of candidate key. In
2. Second normal form: It defines that every non- other words, let R be the schema and X be the set of
prime attribute should be fully functionally depen- attributes over R. If closure of X (X+) determines all the
dent on prime key attribute. A prime attribute is attributes of R, then X is called superkey.
a part of prime key. The relation must be in the
First normal form.
3. Third normal form: It defines that no non- Problem 8.2: Consider a schema R(ABCDE) and
prime attribute is transitively dependent on prime FDs {AB → C, C → D, B → E}. Find superkey.
key attribute. The relation must be in the Second
normal form. Solution:
4. Boyce—Codd normal form: It defines that for
any non-trivial FD, X → A, then X must be a
Find closure set of (A, B, C, AB).
superkey. It is an extension of the Third normal Closure of A = A+ = {A}
form. Closure of B = B+ = {B, E}
5. Fourth normal form: It defines that for every
Closure of C = C+ = {C, D}
multivalued dependency X →→ Y that holds
over R, one of the following statements is true: Closure of AB = AB+ = {A, B, C, D, E}
(a) Y is the subset of X and (b) X is a superkey. So, AB is the superkey.
Also, sometimes called, Multi-valued Dependency
Normal Form (MDNF).
6. Fifth normal form: It denies that for every join
dependency  {R1, …, Rn} that holds over R, 8.5.4.2 Candidate Key
one of the following statements is true: (a) Ri = R
and (b) the join dependency is implied by the set It is an attribute or set of attribute that can act as a pri-
of those FDs over R in which the left side is a key. mary key of a table that uniquely identifies each record
Chapter 8.indd 411 4/9/2015 10:03:47 AM

in a table. Every candidate key is a superkey but not (a) Algorithm for finding decomposition is lossless:
vice versa.
tep 1: Union of all decomposed sub-relation
S
In other words, candidate key is the minimal super- should be equal to relation R.
key. If X is a superkey and none of the proper subset of
X is a superkey, then X is called the minimal superkey R1 ∪ R2 ∪ R3 ∪ … ∪ Rn = R
or candidate key.
tep 2: Any two sub-relations Ri and Rj can be
S
merged into Rij with R1 ∪ R2 only if
Problem 8.3: Consider a schema R(ABCDE) i. Ri ∩ Rj ≠ ∅ (null)

and FDs {AB → C, C → D, B → EA}. Find the
ii. Ri ∩ Rj = X then closure of X(X + ) ⊇ Ri
superkey.
or
Solution: Ri ∩ R j = X then closure of X(X + ) ⊇ Rj
Find closure set of (A, B, C, AB)
tep 3: Repeat step 2 until `N’ relations become
S
Closure of A = A+ = {A} one relation. If `N’ relations become single relation,
Closure of B = B+ = {B, E, A, C, D} then composition is called lossless, otherwise not.
Closure of C = C+ = {C, D}
Problem 8.4: Consider a schema R (ABCDEFGHIJ)
Closure of AB = AB+ = {A, B, C, D, E}
and functional dependencies {FDs = (AB → C, A →
So, AB and B are superkeys. But only B is the can- D, B → F, F → GH, D → IJ)} and decompositions
didate key. (a) {D = (ABCDE, BFGH, DIJ)}
(b) {D = (ABCD, DE, BF, FGH, DIJ)}
Check whether the decomposition is lossless or not.
8.5.4.3 Primary Key
Solution:
It is one or more data attributes that uniquely identify (a) Given
an entity. It does not allow null values. R1 = (ABCDE)
1. Alternate key: The candidate key, which is not R2 = (BFGH)
selected as a primary key.
2. Composite key: It consists of two or more R3 = (DIJ)
attributes. Apply algorithm for lossless decomposition:
3. Foreign key: It is an entity that is the reference
to the primary key of another entity. Step 1:
R1 ∪ R2 ∪ R3 = {(ABCDE ) ∪ (BFGH ) ∪ (DIJ )} = (ABCDEFGHIJ
8.5.5 Decomposition R1 ∪ R2 ∪ R3 = {(ABCDE ) ∪ (BFGH ) ∪ (DIJ )} = (ABCDEFGHIJ ) = R
It is required to eliminate redundancy from the schema. Step 1 satisfies the given condition, so it is true.
If a relational schema R has redundancy in the data, Step 2:
then decompose R into two R1 and R2 schema. There For R1 and R2:
are two properties, which should be maintained when
we perform decomposition, it should be a lossless join as R1 ∩ R2 = (ABCDE ) ∩ (BFGH ) = B
well as dependency preserving.
Find closure of B = B+ = {B, F, G, H} .
Condition (ii) is satisfied, so R1 and R2 can be
8.5.5.1 Lossless Join
merged together. After merging R1 and R2,
Let R be a relation schema and let F be a set of func- R12 = (ABCDEFGH)
tional dependency (FD) over R. R is decomposed into R1
and R2. R1 and R2 are called lossless-join decomposition Now merge R12 and R3.
if R = R1  R2 or if we can recover original relation
from the decomposed relation. R12 ∩ R3 = (ABCDEFGH ) ∩ (DIJ ) = D
Chapter 8.indd 412 4/9/2015 10:03:49 AM

8.5 DATABASE DESIGN 413
Find closure of D = D+ = {D, I, J }. Condition (ii) of step 2 is satisfied, so R134 and R5
can be merged together. After merging R134 and R5
Condition (ii) is satisfied, so R12 and R3 can be
merged together. After merging R12 and R3 R1345 = (ABCDFGHIJ )
R123 = (ABCDEFGHIJ ) For R1345 and R2
So, R123 = R, therefore, it is lossless decomposition. R1345 ∩ R2 = (ABCDFGHIJ ) ∩ (DE ) = D
(b) Given that Find closure of D = D+ = {D, I, J }.

R1 = (ABCD) Condition (ii) of step 2 does not satisfy, so R1345
R2 = (DE) and R2 cannot be merged together.
R3 = (BF) So, finally we left with two decompositions which

cannot be merged into a single relation. So condi-
R4 = (FGH) tion given in step 3 does not satisfy, therefore, it is
R5 = (DIJ) not lossless decomposition.
Apply algorithm for lossless decomposition.
Step 1:
8.5.5.2 Dependency Preserving
R1 ∪ R2 ∪ R3 ∪ R4 ∪ R5 = {(ABCD) ∪ (DE ) ∪ (BF ) ∪ (FGH ) ∪ (DIJ )} = (ABCDEFGHIJ ) = R
{(ABCD) ∪ (DE ) ∪ (BF ) ∪ (FGH ) ∪ (DIJ )} = (ABCDEFGHIJ ) = R All the dependency should be preserved after the decom-
position of schema R.
Step 1 satisfies the given condition, so it is true.
Let R be the relational schema with FD set F decom-
Step 2: posed into R1, R2, R3, …, Rn with FD sets F1, F2, F3, …,
For R1 and R2 Fn, respectively. The decomposition is said to be depen-
dency preserved if
R1 ∩ R2 = (ABCD) ∩ (DE ) = D
F1 ∪ F2 ∪ F3 ∪ … ∪ Fn = F
Find closure of D = D+ = {D, I, J }.
Condition (ii) of step 2 does not satisfy, so R1 and And composition is called non-dependency preserved if
F1 ∪ F2 ∪ F3 ∪ … ∪ Fn ⊂ F
R2 cannot be merged together.
For R1 and R3
(a) Algorithm to check dependency is preserved or not:
R1 ∩ R3 = (ABCD) ∩ (BF ) = B Step 1: Find all the FDs of sub-relations.
Find closure of B = B+ = {B, F, G, H}. tep 2: Check that all the FDs of relation F is
S
covered by FDs of sub-relation in any form directly
Condition (ii) of step 2 is satisfied, so R1 and R3 or indirectly.
can be merged together. After merging R1 and R3
R13 = (ABCDF)
For R13 and R4 Problem 8.5: Consider a schema R(ABCD) and FDs
set F = (A → B, B → C, C → D, D → A) and decom-
R13 ∩ R4 = (ABCDF ) ∩ (FGH ) = F positions D = (AB, BC, CD). Check decomposition is
dependency preserving or not.
Find closure of F = F + = {F, G, H }
Solution:
Condition (ii) of step 2 is satisfied, so R13 and R4
can be merged together. After merging R13 and R4 Step 1: Find all the FDs of sub-relations.
R134 = (ABCDFGH) R1 = (AB) R2 = (BC ) R3 = (CD)

For R134 and R5 Direct A→B B→C C→D
Dependency
R134 ∩ R5 = (ABCDFGH ) ∩ (DIJ ) = D
Indirect B→A C→B D→C
Find closure of D = D+ = {D, I, J } Dependency
Chapter 8.indd 413 4/9/2015 10:03:51 AM

To calculate indirect dependency, use closure property. Functional Covered by F

In relation R1, we have indirect dependency B → A. Dependency
We have to first find closure of all attributes in that
decomposition (like in R1 we have A and B) by using B→C Direct covered
FD set F and then check that we are getting attribute A→C Indirect covered (A+ = ABC ),
A in closure of B. So closure of B = B + = ABCD. So we so A can determine B and C.
can write that indirectly as we have three dependencies
{B → A, B → C and B → D} but only B → A is valid (b) G covers F
for R1 as we have only two attributes (A and B) in
relation R1. Functional Dependency Covered by G
tep 2: Check that all the FDs of relation F is cov-
S A→B Direct covered
ered by FDs of sub-relation in any form directly or
indirectly. B→C Direct covered
(a) A → B (Directly covered by sub-relation FD)

In our example, F covers G and vice versa. So F = G.
(b) B → C (Directly covered by sub-relation FD)
(c) C → D (Directly covered by sub-relation FD)
(d) D → A (Indirectly covered by sub-relation FD by
using {D → C then C → B and then B → A}). 8.6 QUERY LANGUAGES (SQL)
All the dependency is preserved, so it is dependency-
preserved decomposition. SQL stands for Structured Query Languages, which
is a standard computer language for relational data-
8.5.5.3 Relation between Two Functional base management and data manipulation. It is used to
Dependency (FD) Sets query, insert, update and modify data in the table. Some
common RDBMS that use SQL are Oracle, Microsoft
Let F and G be two FD sets: SQL Server, Access, Ingres, Sybase, etc. Raymond Boyce
and Donald Chamberlin developed it in the early 1970s at
1. Set F and G are equal if and only if closure of IBM, but commercially released by Relational Software
F(F +) = Closure of G(G +) Inc. (now, Oracle Corporation) in 1979. Looking into the
2. Set F and G are equal only if both of the below history it was the software named- VULCAN, that was
conditions satisfy: procured by Ashton Tate and then by FoxPro and later
••F covers G: All FD in G is logically implied by F. was purchased by Microsoft. Other popular softwares
••G covers F: All FD in F is logically implied by G.
were/are — Clipper and Gupta Technologies.
3. If all FD in G is logically implied by F but all FD 8.6.1 SQL Commands

in F is not logically implied by G then G ⊂ F.
The SQL commands are used to interact with relational
databases. These commands can be classified into the
Problem 8.6: Let there be two FD sets F and G, following groups (see Tables 8.3 and 8.4):
given as follows:
1. Data Definition Language (DDL): It con-
F = {A → B, B → C } tains metadata, that is, data about data. All the
G = {A → B, AB → C, B → C, A → C } integrity constraints and data base schemas are
defined through DDL. These commands are used
Find the relation between both FDs.
to create, modify and delete database objects.
CREATE, ALTER, TRUNCATE and DROP are
Solution:
part of DDL. TRUNCATE command is used to
(a) F covers G: delete complete data from an existing table, while
DROP command is used to remove a table defini-
Functional Covered by F tion and all data, indexes, triggers and constraints
Dependency for a table.
A→B Direct covered ••Syntax of create command
AB → C +
Indirect covered (AB = ABC ), CREATE TABLE table_name( column1
so AB can determine C. datatype, column2 datatype, column3
Chapter 8.indd 414 4/9/2015 10:03:51 AM

8.6 QUERY LANGUAGES (SQL) 415
datatype, ..... columnN datatype,

Table 8.4 | Clauses in SQL
PRIMARY KEY( one or more columns ) );
For example, create table instructor Clause Description
(INST_ID char(5), name varchar(20),
From Equals to cross product
dept_name varchar(20), salary
numeric(8,2)); Where Selects the tuples which satisfies the
condition
••Syntax of alter command
Group by Groups the table based on specified
ALTER TABLE table_name ADD column_ attribute
name datatype;
Having Used to select groups based on condition
ALTER TABLE table_name DROP COLUMN
column_name;
ALTER TABLE table_name MODIFY COLUMN
2. Data Manipulation Language (DML):
column_name datatype;
DML is used to manipulate the data in database.
ALTER TABLE table_name ADD CONSTRAINT It allows to insert, update and delete data items.
Constraint UNI (column1, column2...); SELECT, INSERT, DELETE and UPDATE
ALTER TABLE table_name ADD CONSTRAINT are part of DML. Control statements BEGIN
Constraint CHECK (CONDITION); TRANSACTION, SAVEPOINT, COMMIT and
ALTER TABLE table_name ADD CONSTRAINT ROLLBACK are also part of DML.
PrimaryKey PRIMARY KEY (column1, ••Syntax of select command
column2...);
SELECT * FROM table_name;
For example, ALTER TABLE CUSTOMERS ADD
GRADE char(1); SELECT column1, column2, columnN
FROM table_name;
••Syntax of truncate command For example, SELECT empno, ename FROM
TRUNCATE TABLE table_name; emp;
For example, TRUNCATE TABLE emp; ••Syntax of insert command
••Syntax of drop command INSERT INTO TABLE_NAME (column1,
DROP TABLE table_name; column2, column3,...columnN)] VALUES
(value1, value2, value3,...valueN);
For example, DROP TABLE emp;
INSERT INTO TABLE_NAME VALUES
(value1,value2,value3,...valueN);
Table 8.3 | SQL commands

For example, INSERT
into emp (empno,
ename, sal, dept) VALUES (100, ABC,
Command Description 10000, Accounts);
CREATE Creates a new table, a view of a ••Syntax of delete command
table or other objects in the database
DELETE FROM table_name WHERE [condi-
ALTER Modifies an existing database (table) tion];
DROP Deletes a table, a view of a table or For example, DELETE FROM emp WHERE
other objects in the database empno = 8;
SELECT Retrieves data from one or more ••Syntax of update command
tables (database)
INSERT Inserts new data record in the UPDATE table_name SET column1 =
database value1, column2 = value2....,
columnN = valueN WHERE [condition];
UPDATE Modifies data in the database
For example, UPDATE mp SET sal = 12000
DELETE Deletes data from the database WHERE empid = 8;
GRANT Gives a privilege to the user
3. Data Control Language (DCL): To assign or
REVOKE Takes back privileges granted from
revoke access rights data control language is used.
the user
GRANT and REVOKE are used for DCL.
Chapter 8.indd 415 4/9/2015 10:03:51 AM

••Syntax of grant command in logarithmic time. B trees are a general form of binary
trees where a node can have more than one child.The
GRANT privilege_name ON object_name
B-trees are efficient for those systems that read and write
TO {user_name | PUBLIC | role_name}
large blocks of data, that is, databases and file systems.
[with GRANT option];
B trees are self-balancing trees. All the leaf nodes are
For example, GRANT SELECT ON emp TO
at the same level. A B tree with order p has:
user1;
1. Root node may contain minimum 1 key
••Syntax of revoke command  p
2. Minimum number of child =   − 1
REVOKE privilege_name ON object_name  2
3. Maximum number of children = p
FROM {User_name | PUBLIC | Role_name};
4. Maximum keys = p − 1
For example, REVOKE SELECT ON emp TO
user1; The order of B-tree can be found as follows:
p × P + ( p − 1) (K + Pr ) ≤ Block size
8.7 FILE STRUCTURES where p is order of the tree, P is the block pointer, Pr is
the record pointer and K is the key pointer.
File structure mainly deals with how files are stored on
the disk. Various file organisations are described below. Problem 8.7: The order of B-tree index is the maxi-
mum number of children it can have. Suppose that
8.7.1 Sequential Files a block pointer takes 6 bytes, the search field value
takes 9 bytes, record pointer takes 7 bytes and the
It is a file organisation system where every file record block size is 512 bytes. What is the order of the B
contains an attribute to uniquely identify a particular tree?
record. Records are placed in a sequential order with
some unique key. Solution:
Given that block size = 512 B; record pointer (Pr) =
8.7.2 Indexing 7 B; block pointer (PB) = 6 B; key pointer (K) = 9
B. So, we have
Indexing is a data structure mechanism to efficiently
retrieve records from the database, for example, book p ×P + ( p − 1) (K + Pr ) ≤ Block size
index. It is defined based on its indexing attributes.
Indexing is of three types: 6p + (p − 1)(9 + 7) = 512
1. Primary index: Indexing is based on ordering p = 24

key field of file.
2. Secondary index: Indexing is based on non-
ordering field of file. 8.7.4 B+ Trees
3. Clustering index: Indexing is based on ordering
non-key field of file. It supports multilevel indexing. The leaf nodes of B+
Ordering field is the field on which the records of file are tree represent actual data pointers. It ensures all leaf
ordered. Ordered indexing is of two types: dense index nodes are balanced, that is, at the same height. Leaf
and sparse index. nodes are linked using link list.
1. Dense index: For every search key value in the B+ tree structure is such that B+ tree is of order n
database, there is an index record. Index record has and it is fixed for every B+ tree.
two parts: search key value and the pointer. The The internal nodes contain at least [n/2] pointers,
pointer is pointed to the actual record. except the root node and at most n pointers. Leaf nodes
2. Sparse index: In this no index record is created contain at least [n/2] record pointers, at least [n/2] key
for every search key. values, at most n record pointers, at most n key values,
and every leaf node contains one block pointer P to point
8.7.3 B Tree to next leaf node and forms a linked list.
Order of internal nodes can be calculated as:
p ×PB + ( p − 1)×k ≤ Block size
A B tree is a data structure that stores data in such a
manner that search, insertions and deletions can be done
Chapter 8.indd 416 4/9/2015 10:03:53 AM

8.8 TRANSACTIONS AND CONCURRENCY CONTROL 417
Order of leaf nodes: (no transaction will affect the existence of any
Pleaf ×[k + Pr ] + PB ≤ Block size other transaction).
4. Durability: Persistence of data even if the system
where p is the order of internal nodes, Pr is record fails and restarts.
pointer, PB is block pointer, k is key pointer, Pleaf is the
order of leaf nodes.
8.8.2 Schedule
Problem 8.8: The order of an internal node in a B+
A chronological execution sequence of transactions is called
tree index is the maximum number of children it can
schedule. It is a list of actions reading, writing, aborting
have. Suppose that a block pointer takes 6 bytes, the
or committing from a set of transactions. A schedule can
search field value takes 9 bytes, record pointer take
have many transactions. Schedule can be further divided
7 bytes and the block size is 512 bytes. What is the
into two types–serial schedule and concurrent schedule.
order of the internal node and leaf node?
Solution: 8.8.2.1 Serial Schedule

Given that block size = 512 B, record pointer (Pr)
A schedule is called serial schedule if all transactions
= 7 B, block pointer (PB) = 6 B and key pointer
are executed in a non-interleaving manner. Let there are
(k) = 9 B.
two transactions T1 and T2 schedule S then schedule S
Order of internal nodes: is called serial schedule if one transaction executes after
p ×PB + ( p − 1) k £ Block size another shown in Fig. 8.5.
6p + (p − 1)9 = 512
p = 34 T1 T2 T1 T2
Order of leaf nodes: Read (X) Read (X)

X=X−Z X=X+W
Pleaf × éëk + Pr ùû + PB £ Block size Write (X)
Write (X)
Pleaf [9 + 7] + 6 = 512 Read (X)
Read (Y )
Y=Y+Z X=X−Z
Pleaf = 31 Write (X)
Write (Y )
Read (X) Read (Y )
8.8 TRANSACTIONS AND X=X+W Y=Y+Z
CONCURRENCY CONTROL Write (X) Write (Y )
Figure 8.5 | Serial schedule.

Transactions are series of read and write operations on
database. When many users access same database at the 8.8.2.2 Concurrent Schedule
same time, some control mechanism is required by data-
bases to stay in consistent state. Following sections will A schedule is called concurrent schedule if transactions
provide description about transactions and their control are executed in an interleaving manner or simultaneous
mechanism. execution of two or more transactions. Let there be two
transactions T1 and T2 in a schedule S, then schedule S
8.8.1 Transactions is called concurrent schedule if both the transactions exe-
cute parallel. One of the scenarios is shown in Fig. 8.6.
It is a series of reads and writes of database objects.
It maintains the integrity of a database while running T1 T2
multiple concurrent operations. Transactions have four
properties that a DBMS must ensure to maintain data. Read (X)
These are known as ACID properties: X = X−Z
Read (X)
1. Atomicity: Either all actions are carried out or X=X+W
none (no partial transaction). Write (X)
2. Consistency: After the execution of transaction, Write (X)
Read (Y)
the database must be in a consistent state (no con- Y =Y+Z
current execution of other transactions). Write (Y)
3. Isolation: All the transactions are carried out
and executed as the only transaction in the system Figure 8.6 | Concurrent schedule.
Chapter 8.indd 417 4/9/2015 10:03:54 AM

8.8.2.3 Comparison between Serial and Example: Let value of data item A = 1000, X =
Concurrent Schedule 100 and Y = 200, then transaction Ti will update
the database with value 1100. Transaction Tj will
Table 8.5 shows the comparison between serial and con- read data item value as 1100. Transaction Tj will
current schedule. update database with value 1300. As transaction Ti
fails, the database value should be 1200. Therefore,
Table 8.5 | Serial schedule vs. concurrent schedule this problem is known as uncommitted read prob-
lem or dirty read problem.
Serial Schedule Concurrent Schedule
2. RW problem (Write after Read): RW
All serial schedules are All serial schedules are problem is also known incorrect summary prob-
consistent schedules not consistent schedules lem or unrepeatable read problem (Fig. 8.8). Let
Less throughput More throughput there be two transactions Ti and Tj of schedule S.
Poor resource utilization Good resource utilization Transaction Ti read a data item and Tj also read
similar data item. Transactions Ti and Tj both
More response time Less response time have write operation on that data item. If in both
For the given For the given transactions, read operation occurs before commit
transactions, the number transactions, the number of the other transaction and before writing back
of serial schedule is of concurrent schedule is that data item in database by other transaction
very much less than the more than the number of then RW problem will arise.
number of concurrent serial schedule
schedule Transaction Ti Transaction Tj
Read (A)
8.8.2.4 Problems Occurring due to If (A > 0)
Concurrent Schedule A=A+1
Read (A)
In concurrent schedule, more than one transaction is
If (A > 0)
executed simultaneously; and due to this some problem
A=A+1
arises with concurrent schedule given as follows:
Write (A)
1. WR problem (Read after Write): WR problem Commit Write (A)
is also known as dirty read problem or uncommit- Commit
Figure 8.8 | Incorrect summary problem

ted read problem. Let there be two transactions Ti
and Tj of schedule S. If transaction Tj reads a data
(RW problem)
item which is written by Ti, but till that time trans-
action Ti is not committed, then WR problem can
occur. If in transaction Ti rollback or failure occurs, Example: Let value of data item A = 100.
then database will be inconsistent due to uncom- Transaction Ti will read data item with value 100
mitted read by Tj transaction (Fig. 8.7). and update database with value 101. Transaction
Tj will read data item value as 100 because it will
Transaction Ti Transaction Tj read before updating value of data item A by trans-
action Ti . Transaction Tj will update database
Read (A)
with value 101. But data item À’ value should
A=A+X
be 102 because both the transaction have commit
Write (A) Dirty Read successfully. Therefore, this problem is known as
unrepeatable read problem.
Read (A)
Write (B) 3. WW Problem (Write after Write): WW
A=A+Y problem in concurrent schedule is also known as
Read (B)
lost update problem (Fig. 8.9). Let there are two
transactions Ti and Tj of schedule S. Transaction
Ti reads a data item and updates it. Now, trans-
action Tj also writes similar data item with some
Failure or rollback
other value and transaction Tj commits success-
Write (A) fully. After commit of transaction Tj, transaction
Figure 8.7 | Uncommitted read problem
Ti will rollback. Thus, updated value of data item
by transaction Tj will be lost.
(WR problem).
Chapter 8.indd 418 4/9/2015 10:03:56 AM

Transaction Ti Transaction Tj may exist in cascading rollback recoverable schedule.

Problems of no recoverability, Uncommitted Read prob-
Read (A) lem (WR problem) and cascading rollback problem do
Write (A) not occur in such schedule.
Write (A) 8.8.3.4 Strict Recoverable Schedule

Commit
Let a schedule S have two transactions Ti and Tj . If S is
Failure and Rollback
called a strict recoverable schedule then it satisfies these
Figure 8.9 | Lost update problem (WW problem). two conditions:
1. Schedule S should be cascading rollback recover-
Example: Let the value of data item A=100. able schedule.
Transaction Ti read data item and update data- 2. If one transaction Ti writes a data item À’, then
base with value 200. Let transaction Tj updated the other transaction Tj is not allowed to write on
database with value 250 and commit successfully. that data item.
After committing transaction Tj , transaction Ti Only incorrect summary problem (RW problem) may
rollback which set data item value to its initial arise in strict recoverable schedule. Irrecoverable,
value 100. So, this problem is known as lost update Uncommitted Read problem (WR problem), lost update
problem. and cascading rollback problems do not occur in such
schedule.
8.8.3 Classification of Schedule Based
on Recoverability 8.8.3.5 Summary of Schedules Based
on Recoverability
8.8.3.1 Irrecoverable Schedule
Let there be two transactions Ti and Tj of schedule S. If Schedule Problem Exists Problem
transaction Tj reads a data item which is updated by Remove
transaction Ti and transaction Tj commits before the
commit (or rollback) of transaction Ti , then the given Irrecoverable All None
schedule S is called irrecoverable schedule. Recoverable Incorrect summary Irrecoverable
problem (RW),
Uncommitted Read
8.8.3.2 Recoverable Schedule problem (WR
problem), lost
Let a schedule S have two transactions Ti and Tj . If
update (WW) and
transaction Tj reads a data item which is updated by
cascading rollback
transaction Ti and transaction Tj is not allowed to
problem
commit (or rollback) before the commit (or rollback) of
transaction Ti , then the given schedule S is called recov- Cascading Incorrect summary Irrecoverable,
erable schedule. Recoverable schedule may suffer from rollback problem (RW) and Uncommitted
uncommitted read, lost update and incorrect summary recoverable lost update (WW). Read problem
problem. (WR problem)
and cascading
rollback
8.8.3.3 Cascading Rollback Recoverable problem.
Schedule
Strict Incorrect summary Irrecoverable,
Let there be four transactions (T1 , T2 , T3 , T4 ) in a given recoverable problem (RW) Uncommitted
schedule S. In schedule S, if rollback of a transaction (T1 ) Read problem
results in the rollback of the other transactions (T2 , T3 , T4 ) (WR problem),
(because of their dependency on each other), then this is lost update
called cascading rollback. (WW) and
cascading
If a schedule is recoverable and has no cascading roll- rollback
back, then it is called cascading rollback recoverable problem.
schedule. Incorrect summary and lost update problems
Chapter 8.indd 419 4/9/2015 10:04:00 AM

Problem 8.9: A concurrent schedule S has three Step 3: Check for strict recoverable. There
transactions T1 , T2 , T3. Transactions execute read/write should be commit operation between write operations
operation in the following sequence: on similar data items by two different transactions. In
the given problem, transaction T2 and transaction T3
Read1(x), Read2 (z), Read3 (x), Read1(z), Read2 (y), have write operations on data item (y), but no commit
Read3 (y), Write1(x), Commit1 , Write2 (z), Write3 (y), operation is performed by transaction T3 before data
item is updated by transaction T2. Hence, the given
Write2 (y), Commit3 , Commit2 .
schedule is not strict recoverable.
Find out the type of recoverable schedule?
Solution: 8.8.4 Classification of Schedule Based

on Serializability
Transaction Transaction Transaction
T1 T2 T3 A concurrent transaction schedule is said to be a serial-
izable schedule if its outcome (resulting database state)
Read1(x) ----- ----- is equal to the outcome of the serial execution of trans-
actions in the given concurrent schedule. Serializable
----- Read2 (z) ----- schedule can be divided into two types: conflict serializ-
ability and view serializability.
----- ----- Read3 (x)
Read1(z) ----- ----- 8.8.4.1 Conflict Serializability

-----
----- Read2 (y) A concurrent schedule S is said to be conflict serial-
izable schedule if S is conflict equivalent to a serial
----- ----- Read3 (y) schedule.
1. Conflict equivalent schedule: Let there be
Write1 ( x ) ----- ----- two schedules S1 and S2. If after swapping con-
secutive non-conflict pairs of operation of transac-
Commit1 ----- ----- tions in schedule S1 results in schedule S2, then
S1 and S2 schedules are called conflict equivalent
----- Write2 (z) ----- schedules.
2. Conflict and non-conflict pairs: A pair of
----- ----- Write3 (y) operations in a schedule is called conflict schedule
if they have all three conditions:
----- Write2 ( y ) ----- •• At least one of the operations should be write
operation.
----- ----- Commit3 •• Both the operations should be performed on the
same data item.
----- Commit2 ----- •• Both operations should be performed by differ-
ent transactions.
Step 1: Check for recoverable. By using the defi- If a pair of operations do not fulfill the above three con-
nition of recoverable schedule we can find that given ditions, then the pair is call a non-conflict pair.
schedule is recoverable or not. (In this problem, no
transaction has read operation on a data item after
that data item is written by another transaction, so Problem 8.10: A schedule S1 with two transactions
the given schedule is recoverable.) T1 and T2 is given as:
Step 2: Check for cascading recoverable. In Read1(x), Write1(x), Read2 (x), Read1(y), Write1(y),
cascading recoverable schedule, no uncommitted read
Read2 (y)),....
is allowed. In the given problem, there is no uncommit-
ted read so the given schedule is cascadless schedule. Find conflict equivalent schedule to S1 .
Chapter 8.indd 420 4/9/2015 10:04:07 AM

Solution: (i) Vi : Read(x) operation and V j : Write(x) operation

Schedule S1: (ii) Vi : Write(x) operation and V j : Read(x) operation
Transaction T1 Transaction T2 (iii) Vi : Write(x) operation and V j : Write(x) operation
Read1 (x)
Conflict Pair
Write1 (x) Problem 8.11: Consider the given schedule S and
draw precedence graph for S.
Read2 (x)
Read1 (y) Transaction T1 Transaction T2

Non-Conflict
Read1 (x)
Pair
Write1 (y) ------
Read2 (y) Write1 (x) ------
Conflict and non-conflict pairs in S1. ------ Read2 (x )

After swapping non-conflict pair in schedule S1 we get Write1 (y ) ------
another schedule called S2 . The schedules S1 and S2
are conflict equivalent schedules. ------ Read2 (y )
Schedule S2:
Solution:
Transaction T1 Transaction T2
Vertices V1 and V2 represent transactions T1 and T2.
Read1 (x) ------ Edge 1 for
V2 : Read(x) operation and V1 : Write (x) operation
Write1 (x) ------
Edge 2 for
Read1 (y )
V1 : Write(y) operation and V2 : Read (y) operation
------
Read2 (x )
1
------
Write1 (y ) ------ V2
Read2 (y )
V1
------
2
Conflict equivalent schedules.
Note: If two schedules S1 and S2 are given then they

3. Testing method of conflict serializable will be conflict equivalent if they satisfy following
schedule: To check that a given schedule is con- conditions:
••Precedence
flict serializable or not, we use precedence graph. If
graph of S1 and S2 should be equal.
precedence drawn from a schedule does not contain
••Each transaction operation should be equal in
a cycle then only it is a conflict serializable sched-
ule. If precedence graph contains a cycle then it is S1 and S2. (The two transaction will be equal of
not a conflict serial schedule. A precedence graph they have same number of operations and type
G contains: of operation are also same.)
(a) Vertex set V: denotes the transactions of the
schedule.
(b) Edge between two vertex from Vi to V j will
be draw where i ¹ j and Vi’s operation occur
Problem 8.12: Consider the given schedule S and
identify that schedule S is conflict serial schedule or
before V j’s operation, if Vi and V j have follow-
not by the use of precedence graph.
ing cases:
Chapter 8.indd 421 4/9/2015 10:04:14 AM

Transaction Transaction Transaction 1. If transaction Ti reads the initial value of data
T1 T2 T3 item (x) in schedule S1, then transaction Ti must
read the initial value of (x ) in schedule S2 .
Read (x) ------ ------ 2. In schedule S1 , if transaction Ti reads the value
Write (x) ------ ------ of data item (x ) produced by transaction Tj then
Ti must reads the value of (x ) that produced by
------ Read (x) ------ transaction Tj in the schedule S2 .
------ ------ Write (x) 3. If transaction Ti write the final value of data item
(x ) in schedule S1 , then transaction Ti must write
------ Write (y) ------ the final value of (x ) in schedule S2 .
Write (y) ------ ------
Problem 8.13: Identify whether the given concurrent
------ ------ Read (y) schedule is a view serializable schedule or not?
Solution: T1 T2 T3
Vertices V1, V2 and V3 represent transactions T1, T2 ------ ------ Read (x)
and T3.
------ Read (x) ------
Edge 1 for
V1 : Write(x) operation and V2 : Read (x) operation ------ ------ Write (x)
Edge 2 for Read (x) ------ ------
V2 : Read (x) operation and V3 : Write(x) operation Write (x) ------ ------
Edge 3 for
V2 : Write(y) operation and V1 : Write(y) operation
Solution: We have a concurrent schedule S1 with
Edge 4 for three transactions, so the number of possible serializ-
V1 : Write(y) operation and V3 : Read (y) operation able schedules with three transactions is 8 (23 = 8).
Edge 5 for So, we will pick one serializable schedule S2 and if
V2 : Write(y) operation and V3 : Read (y) operation the given concurrent schedule is be view equivalent
to that schedule, only then would the given schedule
will be view serializable.
3
V1
T1 T2 T3
1
V2 ------ ------ Read (x)
4
------ Read (x) ------
2 and 5 ------ ------ Write (x)
Read (x) ------ ------
V3
Write (x) ------ ------
Precedence graph drawn from the given schedule con- Schedule S1

tains cycle with vertices V1 and V2. So, the given
schedule is not a conflict serializable schedule. T2 T3 T1
Read (x)
8.8.4.2 View Serializability Read (x)

If a given concurrent schedule is view equivalent to one Write (x)
of the possible serial schedule, then it is called a view
serializable schedule. Let there be two schedules S1 and Read (x)
S2 with the similar set of transactions. Schedules S1 and Write (x)
S2 will be view equivalent if they satisfy following three
conditions: Schedule S2
Chapter 8.indd 422 4/9/2015 10:04:21 AM

1. Two Phase Locking Protocol (2PL): Two

Step 1: Initial Reads. Transactions T3 and T2 read
phase locking protocol is used by the transaction
initial value of data item (x ) in schedule S1 (will not
to lock a data item before reading and writing to
consider T1 because it reads updated value of x by
maintained consistency in the database. The proto-
T3). T3 and T2 also read the initial value of (x ) in
col uses two phases to apply the locking scheme as:
•• Expanding phase: Transaction acquired locks in
serial schedule S2. So, the first condition is true.
Step 2: W-R Conflicts. In the given schedule S1, this phase and no locks are released.
there is only one W-R conflict (between T3 and T1). •• Shrinking phase: Transaction releases locks in
Transaction T1 reads the value of x produced by this phase and no locks are acquired.
transaction T3 in schedule. In schedule S2, transaction Note: Two phase locking protocol ensure conflict
T1 also reads the value of x produced by transaction serializability, but may not be free from irrecover-
T3 in schedule. So, the second condition is also true. able, deadlock and starvation.
Step 3: Final Write. Transaction T1 writes the 2. Two Phase Locking Protocol with Lock
final value of data item (x ) in schedules S1 and S2 Up-gradation: This technique allows transaction
both. So, the third condition is also satisfied. holding a shared lock on a data item can upgrade
shared lock into the exclusive lock on the data item
All the three conditions are satisfied, so the given con-
without performing unlock for shared lock.
current schedule S1 is view equivalent to one of the
3. Strict Two Phase Locking Protocol: In this
possible serial schedule S2. Therefore, the given con-
protocol, if a transaction has exclusive lock on a data
current schedule is a view serializable schedule.
item, then it cannot release the lock until commit
or rollback is performed by that transaction.
4. Rigorous Two Phase Locking Protocol:
8.8.5 Concurrency Control Protocol In this protocol, if a transaction has any lock
(exclusive or shared) on a data item, then it cannot
Concurrency control protocols are used to ensure serial- release the lock until commit or rollback performed
izability of transactions in a concurrent transaction exe- by that transaction.
cution environment. They can be lock based protocols
and Time stamp ordering protocols.
8.8.5.2 Time Stamp Ordering Protocol
8.8.5.1 Lock Based Protocol
The time stamp ordering protocol uses time stamp values
(unique) of the transactions assigned by database to ensure
Lock based protocols uses a mechanism through which a
serializability among transactions. A time stamp is a
transaction has to obtain a read lock (for read operation)
unique value assigned by the database in ascending order.
and write lock (for write operation) on a data item (x)
Suppose we have three transactions T1, T2 and T3 arriving
without obtaining a lock transaction not allowed to per-
in database, then assigned time stamp values will be as
form any operation. There are two types of lock mecha-
T1 < T2 < T3 . A data item has two types of time stamp.
nisms supported by lock based protocol.
1. Shared lock: If a transaction T has shared lock 1. Read Time Stamp (x): It is the highest trans-
on a data item (x), then it can only perform Read action time stamp that has performed Read opera-
operation on that data item. tion successfully on a data item (x). It is also
2. Exclusive lock: If a transaction T has exclusive denoted by RTS(x). Initial value of RTS(x) is zero.
lock on a data item (x), then it can perform Read 2. Write Time Stamp (x): It is the highest
or Write any operation on that data item. transaction time stamp that has performed Write
operation successfully on a data item (x). It is also
Lock Compatibility Matrix is given in Table 8.6, which
denoted by WTS(x). Initial value of WTS(x) is
shows that if a transaction has shared/exclusive lock
zero.
then for which lock it can make a request.
3. Basic Time Ordering Protocol
Table 8.6 | Lock compatibility matrix •• Transaction Ti issues Read(x) operation
Lock Hold Shared Exclusive (a) If (WTS(x) > Time Stamp( Ti ))
Lock Lock Lock Rollback Ti
Request (b) Otherwise allowed execution of Read(x) by
Shared Lock Yes No transaction Ti
Exclusive Lock No No Set RTS(x) = Maximum (RTS(x),
Time Stamp ( Ti ))
Chapter 8.indd 423 4/9/2015 10:04:23 AM

•• Transaction T issues Write(x) operation to inspect all the operation performed by trans-
(a) If (RTS(x) > Time Stamp( Ti )) actions in a schedule. There are various deadlock
Rollback Ti prevention scheme which uses time stamp ordering
mechanism.
(b) If (WTS(x) > Time Stamp( Ti )) 6. Wait-Die Protocol: Assume TimeStamp (T1 ) < … < TimeSt
Rollback Ti TimeStamp (T1 ) < … < TimeStamp (Tn ) and Ti and Tj are any two
Note: Basic Time Ordering protocol ensures seri- transactions and transaction Ti is older than trans-
alizability, equivalent serial schedule based on time action Tj . This implies
stamp Ordering. It is deadlock free protocol, but
basic time ordering protocol can have starvation TimeStamp (Ti ) < … < TimeStamp (Tj ).
and irrecoverable schedule (because it does not •• If transaction Ti required a lock which is hold
ensure order of commit). by transaction Tj, then transaction Ti is allowed
4. Strict Time Ordering Protocol: In this pro- to wait.
tocol, an additional condition is added in the Basic •• If transaction Tj required a lock which is hold by
Time Ordering protocol. This condition is as follows: transaction Ti then transaction Tj will rollback.
Let a schedule have two transactions Ti and Tj 7. Wound-Wait Protocol: Let TimeStamp (T1 ) < … < TimeSt
where time stamp of ( Ti ) is less than timeTimeStamp
stamp (T1 ) < … < TimeStamp (Tn ) and Ti and Tj are any two
of Tj . If transaction Tj issues Read(x)/ Write(x) transactions and transaction Ti is older than trans-
operation with WTS(x) < Time Stamp (Tj ) then action Tj . This implies
(Tj ) has to be delayed Read(x)/ Write(x) opera-
tion until commit or rollback of transaction Ti TimeStamp (Ti ) < TimeStamp (Tj )
that has performed Write(x). •• If transaction Ti required a lock which is hold
Strict Time Ordering Protocol ensures serializability by transaction Tj then rollback transaction Tj.
and deadlock free. But it may suffer from starvation. •• If transaction Tj required a lock which is hold
5. Deadlock Prevention: To prevent deadlock by transaction Ti then transaction Tj is allowed
situation in a database system it is very important to wait.
IMPORTANT FORMULAS
1. All the 3NF and BCNF decomposition guarantee 5. CP: Child Pointer
lossless join.
6. KV: Key Value
2. All 3NF and BCNF decomposition may not guar-
7. Order of internal node in B+ tree is
antee dependency preservation.
PI . CP + (PI − 1)KV ⇐ Block size
3. Any relation with two attributes is in BCNF.
4. Functional dependency F: X → Y implies that for
any two tuples if t1[X ] = t2[X ], they must have
t1[Y ] = t2[Y ].
Inference Rules:
1. Reflexivity 5. Decomposition
If Y ⊆ X, then X → Y If X → YZ, then X → Y and X → Z
2. Augmentation 6. Pseudo-transitivity
If X → Y, then XZ → YZ for any Z If A → B and BC → D, then AC → D
3. Transitivity 7. Number of rows in cross product = r1 × r2
If X → Y and Y → Z, then X → Z
8. Number of attributes in cross product = m + n
4. Union
If X → Y and X → Z, then X → YZ 9. Complete set of operations are {s, p, ∪, −, ×}
Chapter 8.indd 424 4/9/2015 10:04:27 AM

SOLVED EXAMPLES 425
10. In clustering index, file is ordered on non-key field. 12. Different kind of keys — Primary Key, Secondary
Key, Candidate Key, Alternate key, Foreign Key,
11. Secondary indexes are created on unordered file on Superkey, Compound key (or Composite key)
either key or non-key field. (or concatenated key).
SOLVED EXAMPLES
1. Which of the following operations does not modify Solution: Lost update, inconsistent data and
the database? uncommitted (WR) problems are overcome by
database lock.
(a) Sorting (b) Insertion at the beginning
Ans. (d)
(c) Append (d) Modify
6. Database language may consist of the following?
Solution: Sorting does not involving addition/
(a) DDL and DML
insertion or modification of any element.
(b) DML
Ans. (a)
(c) Query language
2. Which of the following operations is based on rela- (d) DDL, DML, query language
tional algebra?
Solution: DDL, DML, query language are all data
(a) rename and union base languages.
(b) select and union Ans. (d)
(c) only union
7. Which of the following is not a function of DBA?
(d) rename, select and union
(a) Network maintenance
Solution: Rename, select and union are some of (b) Routine maintenance
the operations of relational algebra. (c) Defining the schema
Ans. (d) (d) Data access through authorization
3. ACID properties of a transaction are Solution: The role of database administrator is to
(a) atomicity, constant, integrity, durability. manage the DBMS.
(b) atomicity, consistency, isolation, durability. Ans. (a)
(c) atomicity, compact, in-built, durability. 8. In a relational database the category type of infor-
(d) automatically, consistency, isolation, database. mation is represented in
Solution: ACID properties are atomicity, consis- (a) tuple. (b) field.
tency, isolation and durability. (c) primary key. (d) database name.
Ans. (b) Solution: Field shows type of information in rela-
tional database.
4. Which normal form is concerned with removing
Ans. (b)
transitive dependency?
9. Which key is used to identify a tuple uniquely?
(a) 1NF (b) 2NF
(c) 3NF (d) BCNF (a) Primary key (b) Tuple key
(c) Unique key (d) Domain key
Solution: 3NF is also the highest normal form.
Solution: Primary key.
Ans. (c) Ans. (a)
5. The concept Database Lock is used to overcome
10. An association of the information is represented by
the problem of
(a) an attribute. (b) a relationship.
(a) lost update and inconsistent data.
(c) a normal form. (d) the records.
(b) uncommitted dependency and inconsistent data.
(c) inconsistent data only. Solution: Relationship shows association of
(d) lost updates, inconsistent data and uncommitted information.
dependency. Ans. (b)
Chapter 8.indd 425 4/9/2015 10:04:27 AM

11. In which stage of database design, all the necessary Solution: It provides data storage and responsive-
fields and their types of a database are listed? ness to queries.
Ans. (a)
(a) Data definition (b) Data field definition
(c) E-R diagram (d) User definition 17. If D1, D2, … Dn are domains in a relational
model, then the relation is a table, which is a
Solution: In data definition stage, all the fields
subset of
and their types are listed.
(a) D1 ⊕ D2 ⊕ … ⊕ Dn (b) D1 × D2 × … × Dn
(c) D1 ∪ D2 ∪ … ∪ Dn (d) D1 ∩ D2 ∩ … ∩ Dn
Ans. (a)
12. The maximum length of an attribute of type text is

Solution: D1 × D2 × …× Dn
(a) 127. (b) 255. Ans. (a)
(c) 256. (d) It is a variable.
18. Which normal form is considered adequate for
Solution: 255 normal relational database design?
Ans. (b)
(a) 2NF (b) 5NF
13. When a transaction has completed all its opera- (c) 4NF (d) 3NF
tions and is waiting for commit or rollback action,
the state of transaction is Solution: Decomposition in BCNF is not always
(a) partially commit. (b) ready commit. lossless and dependency preserving. So, 3NF is
(c) half commit. (d) commit and rollback. considered to be most adequate form in relational
database.
Solution: The state of that transaction is par- Ans. (d)
tially commit.
19. Let R = (A, B, C, D, E, F) be a relation scheme with
Ans. (a) the following dependencies C → F, E → A, EC → D,
14. Which of the following statement is not false? A → B. Which of the following is a key for R?
(a) Time stamp protocols avoid deadlock. (a) CD (b) EC (c) AE (d) AC
(b) Locking technique is used to avoid deadlock.
(c) 2Phase locking does not provide serializability. Solution: To find key, we try to find closure.
(d) 2Phase deals with input and storing phase. (CD)+ = {C, D, F}
Solution: Time stamp protocol is a deadlock free (EC)+ = {A, B, C, D, E, F}
protocol. (AE)+ = {A, B, E}
(AC)+ = {A, B, C, F}
Ans. (a)
15. Data integrity control makes use of for
maintaining the integrity of database. Only EC satisfies the closure property.
Ans. (b)
(a) specific alphabets (uppercase and lowercase
alphabets) 20. Give the following relation instance:
(b) passwords
(c) relational algebra x y z
(d) storing on a backup hard disk 1 4 2
1 5 3
Solution: It promotes and enforces some integ- 1 6 3
rity rules for the reducing data redundancy and 3 2 2
increasing data consistency.
Ans. (a) Which of the following functional dependencies are
16. Data warehouse provides satisfied by the instance?
(a) transaction responsiveness. (a) XY → Z and Z → Y

(b) storage, functionality responsiveness to queries. (b) YZ → X and Y → Z
(c) demand and supply responsiveness. (c) YZ → X and X → Z
(d) storage of transactions. (d) XZ → Y and Y → X
Chapter 8.indd 426 4/9/2015 10:04:27 AM

Solution: Functional dependency should uniquely maximum and minimum sizes of the join, respec-
identify the value. From the given options only the tively, are
following satisfies this condition:
(a) m + n and 0. (b) mn and 0.
XY → Z, YZ → X, Y → Z AND Y → X (c) m + n and |m − n|. (d) mn and m + n.
Ans. (b)
Solution: Number of tuples in their join gets mul-
21. Relation R is decomposed using a set of functional tiplied with number of tuples in both the relations.
dependencies F, and relation S is decomposed So, the maximum and minimum tuples will be mn
using another set of functional dependencies G. and 0, respectively.
One decomposition is definitely BCNF, the other is Ans. (b)
definitely 3NF, but it is not known which is which.
To make a guaranteed identification, which on one 23. Given the relations
of the following tests should be used on the decom-
Employee(name, salary, deptno) and depart-
positions? (Assume that the closures of F and G
ment (deptno, deptname, address), which
are available.)
of the following queries cannot be expressed
(a) Dependency-preservation using the basic relational algebra operations
(b) Lossless-join (s, p, , E, ∩, −)?
(c) BCNF definition (a) Department address of every employee
(d) 3NF definition (b) Employees whose name is the same as their
department name
Solution: All the 3NF and BCNF decomposition (c) The sum of all employees’ salaries
guarantee lossless join. All 3NF and BCNF decom- (d) All employees of a given department
position does not guaranty dependency preservation.
According to this, only option (c) can be correct. Solution: Aggregate functions such as sum, avg,
Ans. (c) min and max cannot be expressed in terms of
basic relational algebra operations. These require
22. Consider the join of a relation R with a relation extended relational algebra.
S. If R has m tuples and S has n tuples then the Ans. (c)
1. Which of the following scenarios may lead to an 2. Consider the following SQL query:
irrecoverable error in a database system?
SELECT distinct a1,a2,K,an a
(a) A transaction writes a data item after it is read FROM r1,r2,L,rm r
by an uncommitted transaction. WHERE P
(b) A transaction reads a data item after it is read
by an uncommitted transaction. For an arbitrary predicate P, this query is equiva-
(c) A transaction reads a data item after it is lent to which of the following relational algebra
written by a committed transaction. expressions?
(d) A transaction reads a data item after it is (a) Πa1 ,a2 ,…,a n s p (r1 × r2 ×  × rm )
written by an uncommitted transaction.
(GATE 2003: 1 Mark) (b) Πa ,a , K ,an s p (r1  r2  L  rm)
1 2
Solution: In option (d), a problem will occur. (c) Πa ,a

1 2
, K , an s p (r1 ∪ r2 ∪ L ∪ rm )
Suppose first transaction A is reading data written
by second transaction B, but B has not committed (d) Πa ,a
1 2
, K , an s p (r1 ∩ r2 ∩ L ∩ rm )
yet. If A acts on that data and eventually commits,
but B rollback. A will have “acted on" data that (GATE 2003: 1 Mark)
was never committed. So it will lead to irrecover-
able error in database. Solution: P is projection that by default selects
Ans. (d) distinct value of attributes. Cross product will
Chapter 8.indd 427 4/9/2015 10:04:29 AM

consider all the tables r1, r2, … rn. Row (s ) selects Grades: (Roll_number, Course_number, Grade)
all the rows satisfying predicate P. SELECT distinct Name
Ans. (a) FROM Students, Courses, Grades
3. Consider the following functional dependencies in a WHERE Students.Roll_number = Grades.
database: Roll_number
Date_of_birth → Age and Courses.Instructor = Shyam
Age → Eligibility and Courses.Course_number = Grades.
Name → Roll_number Course_number
Roll_number → Name and Grades.grade = A
Course_number → Course_name Which of the following sets is computed by the
above query?
Course_number → Instructor
(Roll_number, Course_number) → Grade (a) N ames of students who have got an A grade in
all courses taught by Shyam.
The relation (Roll_number, Name, Date_of_
(b) Names of students who have got an A grade in
birth, Age) is
all courses.
(a) In Second normal form but not in Third normal (c) Names of students who have got an A grade in
form at least one of the courses taught by Shyam.
(b) In Third normal form but not in BCNF (d) None of the above.
(c) In BCNF (GATE 2003: 2 Marks)
(d) In none of the above
(GATE 2003: 2 Marks) Solution: Use of distinct selects the name only
once. Due to distinct selected names will be unique,
Solution: The applicable FDs are: so one person will be selected only once, in spite of
Date_of_birth → Age getting grade A in any number of courses taught
Age → Eligibility by Shyam.
Ans. (c)
Name → Roll_number
Roll_number → Name 5. Consider three data items D1, D2 and D3, and the fol-
The candidate keys are lowing execution schedule of transactions T1, T2 and
(Roll_number, Date_of_birth) and (Name, T3. In the diagram, R(D) and W(D) denote the actions
Date_of_birth) reading and writing the data item D, respectively.
For candidate key the FD Date_of_birth → Age
T T2 T3
is a partial dependency.
So, the relation is in 1NF. R(D3);
Candidate keys for given relation are: (date of R(D2);
birth, name) and (date of birth, roll_number)
W(D2);
Data of Birth → Age
R(D2);
Age (non-prime) is partially dependent upon Date
of Birth which is prime attribute. According to R(D3);
2NF definition, every non-prime attribute should R(D1);
be fully functional dependent on key, so it is not in W(D1);
2NF. Therefore, it will not be in 3NF and BCNF
W(D2);
also.
Ans. (d) W(D3);
R(D1);
4. Consider the set of relations shown below and the
SQL query that follows: R(D2);
W(D2);
Students: (Roll_number, Name, Date_of_
birth) W(D1);
Courses: (Course number, Course_name,
Instructor) Which of the following statements is correct?
Chapter 8.indd 428 4/9/2015 10:04:30 AM

(a) The schedule is serializable as T2; T3; T1. present in (Student * Enroll), where `*’ denotes
(b) The schedule is serializable as T2; T1; T3. natural join?
(c) The schedule is serializable as T3; T2; T1.
(a) 8, 8 (b) 120, 8
(d) The schedule is not serializable.
(c) 960, 8 (d) 960, 120
(GATE 2004: 1 Mark)
Solution: Let us draw a flow diagram between
the three schedules. An edge is drawn between two Solution: The maximum and minimum number
schedules if there exists read-write or write-write of tuples presented in (Student * Enroll) would be
dependency between them. represented by the minimum of these 120 and 8,
that is, min(120, 8) = 8
Natural join has inbuilt condition of equality. So in
T2 both the relations, minimum and maximum tuples
T3 will be 8.
Ans. (a)
8. It is desired to design an object-oriented employee
record system for a company. Each employee has
a name, unique id and salary. Employees belong
T1 to different categories and their salary is deter-
mined by their category. The functions getName,
getId and compute salary are required. Given the
The schedule has cycle between T1 and T3. So, the class hierarchy below, possible locations for these
schedule is not serializable. functions are:
Ans. (d)
(i) getId is implemented in the superclass
6. Let R1(A, B, C, D) and R2(D, E) be two relation (ii) getId is implemented in the subclass
schema where the primary keys are shown (iii) getName is an abstract function in the
underlined, and let C be a foreign key in R1 refer- superclass
ring to R2. Suppose there is no violation of the (iv) getName is implemented in the superclass
above referential integrity constraint in the corre- (v) getName is implemented in the subclass
sponding relation instances r1 and r2. Which one of (vi) getSalary is an abstract function in the
the following relational algebra expressions would superclass
necessarily produce an empty relation? (vii) getSalary is implemented in the superclass
(a) ΠD (r2 ) − ΠC (r1 ) (b) ΠC (r1 ) − ΠD (r2 ) (viii) getSalary is implemented in the subclass

(c) ΠD (r1  C ≠D (d) ΠD (r1  C = D r2 )
r2 )

(GATE 2004: 1 Mark)
Employee
Solution: C is foreign key in relation R1 refer-

ring to relation R2. So, C would not contain any
additional value that is not present in D. So ΠC(r1) Manager Engineer Secretary
− ΠD(r2) will return no value.
Returns all the tuples which are in ΠC(r1) but not
in ΠD(r2). Choose the best design
Ans. (b)
(a) (i), (iv), (vi), (viii) (b) (i), (iv), (vii)
7. Consider the following relation schema pertaining (c) (i), (iii), (v), (vi), (viii) (d) (ii), (v), (viii)
to a student’s database:
Students(rollno, name, address) Solution: getid and getname can be placed in base
class. Because all the subclasses will have the same
Enroll(rollno, courseno, coursename)
implementation for these functions. But getsalary
where the primary keys are shown underlined. The is dependent on category of employee. So, abstract
number of tuples in the student and Enroll tables function can be placed in base class and its imple-
are 120 and 8, respectively. What are the maxi- mentation in the subclasses.
mum and minimum number of tuples that can be Ans. (a)
Chapter 8.indd 429 4/9/2015 10:04:31 AM

9. The relation scheme Student Performance (Name, 11. The order of an internal node in a B+ tree index
CourseNo, RollNo, Grade) has the following func- is the maximum number of children it can have.
tional dependencies: Suppose that a child pointer takes 6 bytes, the
Name, CourseNo → Grade search field value takes 14 bytes, and the block
RollNo, CourseNo → Grade size is 512 bytes. What is the order of the internal
Name → RollNo node?
RollNo → Name (a) 24 (b) 25
The highest normal form of this relation scheme is (c) 26 (d) 27
(a) 2NF (b) 3NF (GATE 2004: 2 Marks)
(c) BCNF (d) 4NF
Solution: Order of internal node is PI . CP + (PI − 1)
(GATE 2004: 2 Marks) KV <= Block Size
Child pointer (CP) = 6 B
Solution: Name and RollNo are candidate keys. Key value (KV) = 14 B
The attributes are not repeated. So, the relation Block size = 512 B
schema is in 1 NF. 6 + (PI − 1) × 14 = 512
There is no partial dependency. So, the relation is PI = 26
in 2NF. Ans. (c)
Grade is not fully dependent upon all candidate
keys, so it is not in 3NF. 12. The employee information in a company is stored
Candidate keys for given relation are: (rollNo, in the following relation:
courseNo) and (name, courseNo) there is no partial Employee(name, sex, salary, deptName)
and transitive dependencies. So the highest normal
form is 3NF. name → rollNo and rollNo → name Consider the following SQL query:
are violating the condition of BCNF.
SELECT deptName
Ans. (b)
FROM Employee
10. Consider the relation Student (name, sex, marks), WHERE sex = ‘M’
where the primary key is shown underlined, per-
taining to students in a class that has at least one GROUP by deptName
boy and one girl. What does the following rela- HAVING avg(salary)>
tional algebra expression produce? (SELECT avg (salary) FROM Employee)
(Note: r is the rename operator).
It returns the names of the department in which
Π name (rsex = female (Student) − Π name (Studentsex = female Λx = maleΛmarks≤m rn ,x ,m (Student))
(a) the average salary is more than the average
) − Π name (Studentsex = female Λx = maleΛmarks≤m rn ,x ,m (Student)) salary in the company.
(b) the average salary of male employees is more
(a) Names of girl students with the highest marks than the average salary of all male employees
(b) Names of girl students with more marks than in the company.
some boy student (c) the average salary of male employees is more
(c) Names of girl students with marks not less than than the average salary of employees in the
some boy student same department.
(d) Names of girl students with more marks than (d) the average salary of male employees is more
all the boy students than the average salary in the company.
Solution: The query is Part 1 − Part 2. Part 2 Solution: Inner query selects the average salary
of the query results in names of girls having marks of employees in the company. Outer query selects
less than or equal to marks of all the boys. And average salary of male employees. So, the whole
Part 1 selects the names of all the girls in class. So, query finds the department name of whose average
Part 1 − Part 2 will result in names of girl students salary of male employees is more than the average
having more marks than all the boy students. salary in the company.
Ans. (d) Ans. (d)
Chapter 8.indd 430 4/9/2015 10:04:32 AM

13. Which one of the following is a key factor for pre- Table r1 Table r2
ferring B+ trees to binary search trees for indexing
A B C A D
database relations?
1 10 100 1 1000
(a) Database relations have a large number of
records. 1 20 200 1 2000
(b) Database relations are sorted on the primary key.
(c) B+ trees require less memory than binary Table s (natural join of r1 and r2)
search trees.
(d) Data transfer from disks is in blocks. A B C D
(GATE 2005: 1 Mark)
1 10 100 1000
Solution: Indexing performs well for large data 1 20 200 1000
blocks. 1 20 100 200
B+ trees are preferred over the binary search
1 20 200 2000
trees.
B+ trees transfer data from disk to primary Ans. (c)
memory in form of data blocks.
In case of B+ trees, data moves in terms of blocks. 16. Let E1 and E2 be two entities in an E/R diagram
Ans. (d) with simple single-valued attributes. R1 and R2 are
two relationships between E1 and E2, where R1 is
14. Which one of the following statements about one-to-many and R2 is many-to-many. R1 and R2
normal forms is FALSE? do not have any attributes of heir own. What is the
(a) BCNF is stricter than 3NF. minimum number of tables required to represent
(b) Lossless, dependency-preserving decomposition this situation in the relational model?
into 3NF is always possible. (a) 2 (b) 3
(c) Lossless, dependency-preserving decomposition (c) 4 (d) 5
into BCNF is always possible. (GATE 2005: 2 Marks)
(d) Any relation with two attributes is in BCNF.
(GATE 2005: 1 Mark) Solution:
R2
Solution: It is not always possible to decompose E1 E2 E1 E2
a table in BCNF and preserve dependencies. For
l x L x
example, a set of functional dependencies {AB → C,
C → B} cannot be decomposed in BCNF. m y L y
Ans. (c) n z M y
15. Let r be a relation instance with schema R = The one-to-many relationships are represented
(A, B, C, D). We define r1 = PA, B, C(R) and r2 = with entity set from one side. In one-to-many rela-
PA, D(r). Let s = r1 * r2 where * denotes natural tionships, each entity in the entity set can be asso-
join. Given that the decomposition of r into r1 and ciated with at most one entity of the other. So, the
r1 is lossy, which one of the following is TRUE? table is not formed for R1. Hence, the tables are
formed for R2, E1 and E2.
(a) s ⊂ r (b) r ∪ s = r
Using cross-reference technique, minimum of 3
(c) r ⊂ s (d) r * s = s
tables are required.
(GATE 2005: 1 Mark)
Ans. (b)
Solution: 17. The following table has two attributes A and C
where A is the primary key and C is the foreign key
Table r
referencing A with on-delete cascade.
A B C D
A C
1 10 100 1000
2 4
1 20 200 1000
3 4
20 200 2000
(Continued)
Chapter 8.indd 431 4/9/2015 10:04:32 AM

Continued {A → B, BC → D, E → C, D → A}.
A C What are the candidate keys of R?
4 3 (a) AE, BE (b) AE, BE, DE
5 2 (c) AEH, BEH, BCH (d) AEH, BEH, DEH
7 2 (GATE 2005: 2 Marks)
9 5
Solution: Let S be a candidate key of relation R
6 4 if the closure of S is all attributes of R and there
is no subset of S whose closure is all attributes
The set of all tuples that must be additionally of R.
deleted to preserve referential integrity when the Closure of AEH: AEH+ = {ABCDEH}
tuple (2, 4) is deleted is
Closure of BEH: BEH+ = {ABCDEH}
(a) (3, 4) and (6, 4)
(b) (5, 2) and (7, 2) Closure of DEH: DEH+ = {ABCDEH}
(c) (5, 2), (7, 2) and (9, 5) Ans. (d)
(d) (3, 4), (4, 3) and (6, 4)
20. Consider the following log sequence of two transac-
(GATE 2005: 2 Marks) tions on a bank account, with initial balance 12000,
that transfer 2000 to a mortgage payment and then
Solution: When (2,4) is deleted, as C is a foreign apply a 5% interest.
key referring A with delete on cascade, all entries
with value 2 in C must be deleted. 1. T1 start
So, (5, 2) and (7, 2) are deleted. As a result of 2. T1 B old = 1200 new = 10000
this, 5 and 7 are deleted from A, which causes
(9, 5) to be deleted. 3. T1 M old = 0 new = 2000
Ans. (c) 4. T1 commit
5. T2 start
18. The relation book (title, price) contains the titles
and prices of different books. Assuming that no 6. T2 B old = 10000 new = 10500
two books have the same price, what does the fol- 7. T2 commit
lowing SQL query list?
Suppose the database system crashes just before log
SELECT title record 7 is written. When the system is restarted,
FROM book as B which one statement is true of the recovery procedure?
WHERE (SELECT count(*) (a) We must redo log record 6 to set B to 10500.
FROM book as T (b) We must undo log record 6 to set B to 10000
and then redo log records 2 and 3.
WHERE T.price > B.price) < 5
(c) We need not redo log records 2 and 3 because
(a) Titles of the four most expensive books transaction T1 has committed.
(b) Title of the fifth most inexpensive book (d) We can apply redo and undo operations in arbi-
(c) Title of the fifth most expensive book trary order because they are idempotent.
(d) Titles of the five most expensive books (GATE 2006: 1 Mark)
Solution: When a transaction is committed, no
Solution: The outer query selects all titles from need to redo or undo operations.
book table. For every selected book, the sub-query Ans. (c)
returns count of those books, which are more expen-
sive than the selected book. The where clause of outer 21. Consider the relation account (customer, balance)
query will be true for five most expensive books. where customer is a primary key and there are
Ans. (d) no null values. We would like to rank customers
according to decreasing balance. The customer with
19. Consider a relation scheme R = (A, B, C, D, E, H ) the largest balance gets rank 1. Ties are not broke
on which the following functional dependencies hold: but ranks are skipped: if exactly two customers
Chapter 8.indd 432 4/9/2015 10:04:33 AM

have the largest balance they each get rank 1, and (a) All queries return identical row sets for any database
rank 2 is not assigned. (b) Query 2 and Query 4 return identical row sets for
all databases but there exist databases for which
Query 1: SELECT A.customer, count(B. Query 1 and Query 2 return different row sets.
customer) FROM account A, account B (c) There exist databases for which Query 3 returns
WHERE A.balance ⇐ B.balance group by strictly fewer rows than Query 2
A.customer (d) There exist databases for which Query 4 will
Query 2: SELECT A.customer, 1+count(B. encounter an integrity violation at runtime.
customer) FROM account A, account B
WHERE A.balance < B.balance group by (GATE 2006: 2 Marks)
A.customer Solution: The output of Query 2, Query 3 and
Consider these statements about Query 1 and Query 2. Query 4 will be identical. Query 1 produces dupli-
cate rows. But row set produced by all the queries
I. Q
uery 1 will produce the same row set as Query will be same. For example,
2 for some but not all databases.
II. Both Query 1 and Query 2 are correct imple- Table: Enrolled Table: Enrolled
mentation of the specification. Student Course Student Amount
III. Q uery 1 is a correct implementation of the
specification but Query 2 is not. A C1 A 100
IV. Neither Query 1 nor Query 2 is a correct imple- B C1 B 100
mentation of the specification. A C2 D 200
V. Assigning rank with a pure relational query
C C1
takes less time than scanning in decreasing
balance order assigning ranks using ODBC.
Output of Query 1: A B
Which two of the above statements are correct? Output of Query 2: A B
(a) II and V (b) I and III Output of Query 3: A B
(c) I and IV (d) III and V
Output of Query 4: A B
(GATE 2006: 2 Marks) Ans. (a)
Solution: Both query1 and query2 perform the 23. Consider the relation enrolled (student, course) in
same task, but none of them sort the customer which (student, course) is the primary key, and the
according to the decreasing balance. So, option (c) relation paid (student, amount) in which student
is correct. is the primary key. Assume no null values and no
Ans. (c) foreign keys or integrity constraints. Assume that
amounts 6000, 7000, 8000, 9000 and 10000 were
22. Consider the relation enrolled (student, course) in
each paid by 20% of the students. Consider these
which (student, course) is the primary key, and
query plans (Plan 1 on left, Plan 2 on right) to “list
the relation paid (student, amount) where student
all courses taken by students who have paid more
is the primary key. Assume no null values and no
than x”.
foreign keys or integrity constraints. Given the fol-
lowing four queries: A disk seek takes 4 ms, disk data transfer band-
Query 1: SELECT student FROM enrolled WHERE width is 300 MB/s and checking a tuple to see if
student in (SELECT student FROM paid) amount is greater than x takes 10 µs. Which of the
following statements is correct?
Query 2: SELECT student FROM paid WHERE
student in (SELECT student FROM enrolled) (a) P lan 1 and Plan 2 will not output identical row
Query 3: SELECT E.student FROM enrolled sets for all databases.
E, paid P WHERE E.student = P.student (b) A course may be listed more than once in the
Query 4: SELECT student FROM paid WHERE output of Plan 1 for some databases.
exists (c) For x = 5000, Plan 1 executes faster than Plan
2 for all databases.
(SELECT * FROM enrolled WHERE enrolled. (d) For x = 9000, Plan 1 executes slower than Plan
student = paid.student) 2 for all databases.
Which one of the following statements is correct? (GATE 2006: 2 Marks)
Chapter 8.indd 433 4/9/2015 10:04:33 AM

Enrolled Paid Enrolled Paid
Probe index Sequential Probe index Sequential scan

on student scan select on student
′
amount >x
Indexed nested loop join Indexed nested loop join
Project on course Select on amount > x
Project on course
Solution: Plans need to load both tables’ courses The closure of {AF}+ is {A F D E}. It cannot
and enrolled. So, disk access time is the same for drive all the members of set given in option (c). So,
both plans. option (c) is false.
Plan 2 does lesser number of comparisons com- Ans. (c)
pared to Plan 1.
25. Information about a collection of students is given
So, join operation will require more comparisons
by the relation studinfo(studId, name, sex). The
as the second table will have more rows in Plan 2
relation enroll(studId, courseId) gives which stu-
compared to Plan 1.
dent has enrolled for (or taken) what course(s).
The joined table of two tables will have more rows,
Assume that every course is taken by at least
so, more comparisons are needed to find amounts
one male and at least one female student. What
greater than x.
does the following relational algebra expression
Plan 1 executes faster than Plan 2. First tuples are
represent?
filtered then joined, whereas in plan 2 first tuples
are joined and then filtered, which takes more time. Π courseId (Π studId (s sex ="female"(studInfo)) × Π courseId (enroll))
Π courseId (Π studId (s sex ="female"(studInfo)) × Π courseId (enroll))
Ans. (c)
(a) C ourses in which all the female students are
24. The following functional dependencies are given:
enrolled.
AB → CD, AF → D, DE → F, C → G, F → E, G (b) Courses in which a proper subset of female
→A students are enrolled.
(c) Courses in which only male students are
Which one of the following options is false? enrolled.
(a) {CF}+ = {ACDEFG} (d) None of the above
(b) {BG}+ = {ABCDG} (GATE 2007: 2 Marks)
(c) {AF}+ = {ACDEFG} Solution:
(d) {AB}+ = {ABCDFG}
(i) Select studId of all female students and select
(GATE 2006: 2 Marks) all courseId of all courses.
(ii) The query performs a Cartesian product of the
Solution: Closure of AF or AF+ = {ADEF}, above select two columns in Step 1.
closure of AF does not contain C and G. (iii) It subtracts enroll table from the result of Step 2.
Chapter 8.indd 434 4/9/2015 10:04:33 AM

This will remove all the (studId, courseId) pairs, Solution: Q1 selects an employee but does not
which are present in enrol table. compute the result as after the `Where not exists’,
If all female students have registered in courses, it does not have statement to produce the result.
then this course will not be there in the subtracted Q2 selects an employee who gets higher salary
result. than anyone in the department 5. It correctly finds
So, the complete expression returns courses in employees who get higher salary than anyone in
which a proper subset of female students is enrolled. the department 5.
Ans. (b) Ans. (b)
26. Consider the relation employee (name, sex, 28. Which one of the following statements is FALSE?
supervisorName) with name as the key. super-
(a) Any relation with two attributes is in BCNF.
visorName gives the name of the supervisor of
(b) A relation in which every key has only one
the employee under consideration. What does
attribute is in 2NF.
the following Tuple Relational Calculus query
(c) A prime attribute can be transitively dependent
produce?
on a key in a 3NF relation.
{e.name|employee(e) ∧{(∀x) (d) A prime attribute can be transitively depen-
[¬employee(x) ∨ x.supervisorName ≠ dent on a key in a BCNF relation.
e.name ∨ x.sex = “male”)]}/ inserted on (GATE 2007: 2 Marks)
the left of expression
Solution: According to the definition of 3NF, a
(a) Names of employees with a male supervisor.
prime attribute can be transitively dependent on a
(b) Names of employees with no immediate male
key. Option (d) is incorrect because this does not
subordinates.
satisfy the condition of BCNF.
(c) Names of employees with no immediate female
Ans. (d)
subordinates.
(d) Names of employees with a female supervisor. 29. The order of a leaf node in a B+ tree is the maxi-
(GATE 2007: 2 Marks) mum number of (value, data record pointer) pairs
Solution: The query selects all those employees it can hold. Given that the block size is 1 KB, data
whose immediate subordinate is “male”. record pointer is 7 bytes long, the value field is
Ans. (b) 9 bytes long and a block pointer is 6 bytes long,
what is the order of the leaf node?
27. Consider the table employee (empId, name, depart- (a) 63 (b) 64
ment, salary) and the two queries Q1, Q2 below. (c) 67 (d) 68
Assuming that department 5 has more than one (GATE 2007: 2 Marks)
employee, and we want to find the employees who
get higher salary than anyone in the department Solution: Let the order of the leaf node be n.
5, which one of the statements is TRUE for any Block size = 1 KB = 1024 bits
arbitrary employee table? 6 + 7n + (n − 1)9 = 1024
Q1: SELECT e.empId 1 n = 64
FROM employee e Ans. (b)
WHERE not exists (SELECT * FROM
employee s WHERE s.department = “5”
30. Consider the following schedules involving two
and s.salary >= e.salary) transactions. Which one of the following statements
is TRUE?
Q2: SELECT e.empId
FROM employee e S1: r1(X ); r1(Y ); r2(X ); r2(Y ); w2(Y ); w1(X )
WHERE e.salary > Any S2: r1(X ); r2(X ); r2(Y ); w2(Y ); r1(Y ); w1(X )
(SELECT distinct salary FROM em-
ployee s WHERE s.department = “5”) (a) Both S1 and S2 are conflict serializable.
(b) S1 is conflict serializable and S2 is not conflict
(a) Q1 is the correct query. serializable.
(b) Q2 is the correct query. (c) S1 is not conflict serializable and S2 is conflict
(c) Both Q1 and Q2 produce the same answer. serializable.
(d) Neither Q1 nor Q2 is the correct query. (d) Both S1 and S2 are not conflict serializable.
Chapter 8.indd 435 4/9/2015 10:04:33 AM

Solution: Solution: Using negation theorem, we find option

Schedule S1 (c) is true.
Ans. (c)
T1 T2
33. Let R and S be two relations with the following
r1(X)
schema:
r1(Y) R(P, Q, R1, R2, R3)
r2(X) S(P, Q, S1, S2)
r2(Y) where {P, Q} is the key for both schemas. Which
of the following queries are equivalent?
w2(Y)
w1(X) I. Π P (R  S )
II. Π P (R)  Π P (S )
( )
The schedule is not conflict serializable.
III. Π P Π P ,Q (R) ∩ Π P ,Q (S ) .
( ))
Schedule S2
T1 T2
(
IV. ΠP ΠP ,Q (R ) − ΠP ,Q (R ) − ΠP ,Q (S )
r1(X) (a) Only I and II (b) Only I and III

(c) Only I, II and III (d) Only I, III and IV
r2(X)
r2(Y)
w2(Y) Solution:
r1(Y) In I, P from natural join of R and S are selected.
In III, all P from intersection of (P, Q) pairs present
w1(X)
in R and S.
IV is also equivalent to III because (R − (R − S))
The schedule is conflict serializable to T2T1. = R ∩ S.
Ans. (c)
II is not equivalent as it may also include P, where
31. A clustering index is defined on the fields which are Q is not same in R and S.
of type Queries I, II and III performs the same operations,
that is, they select attribute P.
(a) Non-key and ordering
Ans. (c)
(b) Non-key and non-ordering
(c) Key and ordering
34. Consider the following relational schemes for a
(d) Key and non-ordering
library database:
(GATE 2008: 1 Mark)
Book(Title, Author, Catalog_no, Publisher, Year,
Solution: Clustering index is defined on the fields. Price)
If the records of the file are physically ordered on Collection(Title, Author, Catalog_no)
a non-key field, it will not have a distinct value for within the following functional dependencies:
each record. So, the clustering index is defined on I. Title Author → Catalog_no
the fields of type non-key and ordering. II. Catalog_no → Title Author Publisher Year
Ans. (a) III. Publisher Title Year → Price
32. Which of the following tuple relational calculus Assume {Author, Title} is the key for both schemes.
expression(s) is/are equivalent to ∀t ∈r (P(t))? Which of the following statements is true?
Ι. ¬∃t ∈r (P(t) ) (a) Both Book and Collection are in BCNF.
ΙΙ. ∃t ∉r (P(t) ) (b) Both Book and Collection are in 3NF only.
(c) Book is in 2NF and Collection is in 3NF.
ΙΙΙ. ¬∃t ∈ r(¬P (t) ) (d) Both Book and Collection are in 2NF only.
ΙV. ∃t ∉r (¬P(t) ) (GATE 2008: 2 Marks)
(a) I only (b) II only Solution: Collection is in BCNF as there is only one
(c) III only (d) III and IV only functional dependency Title Author → Catalog_no
(GATE 2008: 1 Mark) and {Author, Title} is key for collection.
Chapter 8.indd 436 4/9/2015 10:04:35 AM

Book is not in BCNF because Catalog_no is not a Book is in 2NF because every non-prime attribute
key and there is a functional dependency Catalog_ of the table is dependent on either the key [Title,
no → Title Author Publisher Year. Author] or another non-prime attribute.
Book is not in 3NF because non-prime attributes Ans. (c)
(Publisher Year) are transitively dependent on key Linked Answer Questions 35 and 36: Consider the
[Title, Author]. following ER diagram:
M1 M2 M3 P1 P2 N1 N2
M R1 P R2 N
35. The minimum number of tables needed to repre- S1: R1[x]R2[x]R2[y]W1[x]W1[y]W2[y]

sent M, N, P, R1, R2 is
S2: R1[x]R2[x]R2[y]W1[x]W2[y]W1[y]
(a) 2 (b) 3 S3: R1[x]W1[x]R2[x]W1[y]R2[y]W2[y]
(c) 4 (d) 5
(GATE 2008: 2 Marks) S4: R2[x]R2[y]R1[x]W1[x]W1[y]W2[y]
Which of the above schedules are conflict-
Solution: Many-to-one and one-to-many relation- serializable?
ship sets that are total on the many-side can be
(a) S1 and S2 (b) S2 and S3
represented by adding an extra attribute to the
(c) S3 only (d) S4 only
“many” side, containing the primary key of the
“one” side. (GATE 2009: 2 Marks)
Since R1 is many to one and participation of M is
total, M and R1 can be combined to form the table Solution: The serial schedule T1T2 has the follow-
{M1, M2, M3, P1}. N is a week entity set, so it can ing sequence of operations:
be combined with P.
Ans. (a) R1[x]W1[x]W1[y]R2[x]R2[y]W2[y]
36. Which of the following is a correct attribute set And the schedule T2T1 has the following sequence
for one of the tables for the correct answer to the of operations:
above question? R2[x]R2[y]W2[y]R1[x]W1[x]W1[y]
(a) {M1, M2, M3, P1}
The schedule S2 is conflict-equivalent to T2T1 and
(b) {M1, P1, N1, N2}
S3 is conflict-equivalent to T1T2.
(c) {M1, P1, N1}
Only S2 and S3 are conflict serialzable schedules.
(d) {M1, P1}
To test for conflict serializability, make the wait
for graph. If there is a cycle in graph, it means not
Solution: conflict serializable.
For R1 correct attribute set: M1, M2, M3, P1 Ans. (b)
For R2 correct attribute set: N1, N2, P1, P2 with P1
38. The following key values are inserted into a B+ tree
as primary key and N1 as weak entity set.
in which order of the internal nodes is 3, and that of
Ans. (a)
the leaf nodes is 2, in the sequence given below. The
37. Consider two transactions T1 and T2, and four order of internal nodes is the maximum number of
schedules S1, S2, S3, S4 of T1 and T2 as given below: tree pointers in each node, and the order of leaf
nodes is the maximum number of data items that
T1: R1[x]W1[x]W1[y] can be stored in it. The B+ tree is initially empty.
T2: R2[x]R2[y]W2[y] 10, 3, 6, 8, 4, 2, 1
Chapter 8.indd 437 4/9/2015 10:04:35 AM

The maximum number of times leaf nodes would FROM Catalog C

get split up as a result of these insertions is WHERE C.pid NOT IN (SELECT P.pid
(a) 2 (b) 3 FROM Parts P
(c) 4 (d) 5 WHERE P.color <> ‘blue’))
Assume that relations corresponding to the above
Solution: Total number of splits here are 5. schema are not empty. Which one of the following
is the correct interpretation of the above query?
4 (a) F ind the names of all suppliers who have sup-
plied a non-blue part.
2 3 6 (b) Find the names of all suppliers who have not
supplied a non-blue part.
(c) Find the names of all suppliers who have sup-
1 2 3 4 6 8 10 plied only blue parts.
(d) Find the names of all suppliers who have not
Ans. (a) supplied only blue parts.
39. Let R and S be relational schemes such that R =
{a, b, c} and S = {c}. Now consider the following Solution: The sub-query “SELECT P.pid FROM
queries on the database: Parts P WHERE P.color <> ‘blue’” returns
I. Π R − S (r) − Π R − S (Π R − S (r) × s − Π R − S ,S (r))
pids of parts, which are not blue.
The sub-query “SELECT C.sid FROM Catalog
II. { t | t ∈ ΠR −S (r) ∧ ∀u ∈ s(∃v ∈ r(u = v[s] ∧ t = v[R − S ]))} C WHERE C.pid NOT IN (SELECT P.pid FROM
∀u ∈ s(∃v ∈ r(u = v[s] ∧ t = v[R − S ]))} Parts P WHERE P.color <> ‘blue’)” returns
III. {t | t ∈ ΠR −S (r) ∧ ∀v ∈r(∃u ∈ s(u = v[s] ∧ t = v[R − S ]))} sid of all those suppliers who have supplied blue
parts.
∧ ∀v ∈r(∃u ∈ s(u = v[s] ∧ t = v[R − S ]))}
The complete query returns the names of all sup-
IV. SELECT R.a, R.b FROM R, S WHERE R.c pliers who have supplied a non-blue part.
= S.c
Inner query SELECT P. pid FROM Parts P
Which of the above queries are equivalent? WHERE P. color <> ‘blue’ will select pid
(a) I and II (b) I and III of non-blue parts. Outer query of this will pro-
(c) II and IV (d) III and IV duce sid of those suppliers whose pid is not in
the result of inner query. The query will result
(GATE 2009: 2 Marks) the supplier names who have supplied non-blue
Solution: I and II describe the division operator parts
in relational algebra and tuple relational calculus, Ans. (a)
respectively.
Ans. (a) 41. Assume that, in the suppliers relation above, each
supplier and each street within a city has a unique
Common Data Questions 40 and 41: Consider the name, and (sname, city) forms a candidate key.
following relational schema: No other functional dependencies are implied other
Suppliers(sid: integer, sname: string, than those implied by primary and candidate keys.
city: string, street: string) Which one of the following is TRUE about the
above schema?
Parts(pid: integer, pname: string,
color: string) (a) The schema is in BCNF.
Catalog(sid: integer, pid: integer, (b) The schema is in 3NF but not in BCNF.
cost: real) (c) The schema is in 2NF but not in 3NF.
(d) The schema is not in 2NF.
40. Consider the following relational query on the (GATE 2009: 2 Marks)
above database:
SELECT S.sname Solution: Candidate key: sname, city
FROM Suppliers S Primary key: sname
WHERE S.sid NOT IN (SELECT C.sid Alternate key: sid, sname
Chapter 8.indd 438 4/9/2015 10:04:37 AM

city → street EXISTS(SELECT * FROM Passenger

sname → city WHERE age > 65 AND Passenger.pid =
Transitivity: sname → city → street Reservation.pid)
So, it is in 3NF.
(a) 1, 0 (b) 1, 2
Ans. (b) (c) 1, 3 (d) 1, 5
42. Consider a B+ tree in which the maximum number (GATE 2010: 1 Marks)
of keys in a node is 5. What is the minimum number
of keys in any non-root node? Solution: The outer query selects four entries
(a) 1 (b) 2 (with pid as 0, 1, 5, 3) from Reservation table. Out
(c) 3 (d) 4 of these selected entries, the sub-query returns non-
(GATE 2010: 1 Mark) null values only for 1 and 3.
Ans. (c)
Solution: In B+ tree, root node has minimum
two block pointers and maximum p block pointer, 44. Which of the following concurrency control proto-
where p = order and key = order − 1. cols ensure both conflict serializability and freedom
In the non-root node, the minimum number of keys from deadlock?
= p/2 − 1 I. Two-phase locking
Here, key = 5, order = 6 II. Timestamp ordering
So, minimum number of keys in non-root node (a) I only (b) II only
= 6/2 − 1 = 2 (c) Both I and II (d) Neither I nor II
43. A relational schema for a train reservation data-
base is given below: Solution: Two-phase locking is a concurrency
control method, which guarantees serializability.
Passenger(pid, pname, age) The protocol utilizes locks, applied by a transac-
Reservation(pid, class, tid) tion to data, which may block other transactions
Table: Passenger from accessing the same data during the transac-
tion’s life. 2PL may be lead to deadlocks which
pid pname Age result from the mutual blocking of two or more
transactions.
0 `Sachin’ 65
Timestamp ordering algorithm is a non-lock con-
1 `Rahul’ 66 currency control method. In Timestamp based
2 `Saurav’ 67 method, deadlock cannot occur as no transaction
3 Ànil’ 69 ever waits.
Ans. (b)
Table: Reservation 45. Consider the following schedule for transactions
pid Class Tid T1, T2 and T3:
0 ÀC’ 8200
T1 T2 T3
1 ÀC’ 8201
Read(X )
2 `SC’ 8201
Read(Y )
5 ÀC’ 8203
Read(Y )
1 `SC’ 8204
Write(Y )
3 ÀC’ 8202
Write(X )
What pids are returned by the following SQL query Write(X )
for the above instance of the tables?
Read(X )
SELECT pid
FROM Reservation WHERE class = ‘AC’ AND Write(X )
Chapter 8.indd 439 4/9/2015 10:04:37 AM

Which one of the schedules below is the correct multiple accounts or joint accounts. This attri-
serialization of the above? bute stores the primary account number.
(a) T1 → T3 → T2 (b) T2 → T1 → T3 IV. Name: Name of the Student.
(c) T2 → T3 → T1 (d) T3 → T1 → T2 V. Hostel_Room: Room number of the hostel.
(GATE 2010: 2 Marks) Which of the following options is INCORRECT?
(a) BankAccount_Number is a candidate key.
Solution: T1 can complete before T2 and T3 as
(b) Registration_Number can be a primary key.
there is no conflict between Write(X) of T1 and
(c) UID is a candidate key if all students are from
the operations in T2 and T3 which occur before
the same country.
Write(X) of T1.
(d) If S is a superkey such that S ∩ UID is NULL
then S ∪ UID is also a superkey.
T3 can complete before T2 as the Read(Y) of T3
does not conflict with Read(Y) of T2. Similarly,
(GATE 2011: 1 Mark)
Write(X) of T3 does not conflict with Read(Y) and
Write(Y) operations of T2. Solution: When two students hold joint account
After topologically sorting, the sequence is T1 → in that case BankAccount_Num will not uniquely
T3 → T2. determine other attributes.
Ans. (a) Ans. (a)
46. The following functional dependencies hold for 48. Database table by name Loan_Records is given
relations R(A, B, C) and S(B, D, E) below:
B → A,
Borrower Bank_Manager Loan_Amount
A→C
Ramesh Sunderajan 10000.00
The relation R contains 200 tuples and the rela-
tion S contains 100 tuples. What is the maximum Suresh Ramgopal 5000.00
number of tuples possible in the natural join R  S? Mahesh Sunderajan 7000.00
(a) 100 (b) 200
(c) 300 (d) 2000 What is the output of the following SQL query?
(GATE 2010: 2 Marks) SELECT count(*) FROM(
(SELECT Borrower, Bank_Manager FROM
Solution: B is a candidate key of R.
Loan_Records) AS S NATURAL JOIN
So, all 200 values of B must be unique in R. (SELECT Bank_Manager, Loan_Amount
There is no functional dependency given for S. FROM Loan_Records) AS T );
For the maximum number of tuples in output,
(a) 3 (b) 9
there are two ways.
(c) 5 (d) 6
(i) All 100 values of B in S are same and there is (GATE 2011: 2 Marks)
an entry in R, which matches with this value.
So, we are having 100 tuples in output.
Solution: Output as Table S:
(ii) All 100 values of B in S are different and these
values are present in R also. So, we are having Borrower Bank_Manager
100 tuples.
Ans. (a) Ramesh Sunderajan
Suresh Ramgopal
47. Consider a relational table with a single record
for each registered student with the following Mahesh Sunderajan
attributes:
Output as Table T:
I. Registration_Number: Unique registration
number for each registered student. Bank_Manager Loan_Amount
II. UID: Unique Identity number, unique at the Sunderajan 10000.00
national level for each citizen.
III. B ankAccount_Number: Unique account Ramgopal 5000.00
number at the bank. A student can have Sunderajan 7000.00
Chapter 8.indd 440 4/9/2015 10:04:37 AM

Natural Join of S and T: 50. Which of the following statements are TRUE
about an SQL query?
Borrower Bank_Manager Loan_Amount
I. An SQL query can contain a HAVING clause
Ramesh Sunderajan 10000.00 even if it does not have a GROUP BY clause.
Ramesh Sunderajan 7000.00 II. An SQL query can contain a HAVING clause
only if it has GROUP BY clause.
Suresh Ramgopal 5000.00 III. A ll attributes used in the GROUP BY clause
Mahesh Sunderajan 10000.00 must appear in the SELECT clause.
IV. Not all attributes used in the GROUP BY
Mahesh Sunderajan 7000.00 clause need to appear in the SELECT clause.
(a) I and III (b) I and IV
Ans. (c)
(c) II and III (d) II and IV
49. Consider a database table T containing two col- (GATE 2012: 1 Mark)
umns X and Y each of type integer.
Solution:
After the creation of the table, one record (X = 1,
Y = l) is inserted in the table. I. HAVING clause can also be used with aggregate
function. If we use a HAVING clause without
Let MX and MY denote the respective maximum
a GROUP BY clause, the HAVING condi-
values of X and Y among all records in the table at
tion applies to all rows that satisfy the search
any point in time. Using MX and MY, new records
condition.
are inserted in the table 128 times with X and Y
values being MX + 1, 2 * MY + 1, respectively. II. To verify S,
It may be noted that each time after the insertion, CREATE TABLE temp (id INT, name
values of MX and MY change. VARCHAR(100));
What will be the output of the following SQL INSERT INTO temp VALUES (1, "abc");
query after the steps mentioned above are carried INSERT INTO temp VALUES (2, "abc");
out?
INSERT INTO temp VALUES (3, "bcd");
SELECT Y FROM T WHERE X = 7;
INSERT INTO temp VALUES (4, "cde");
(a) 127 (b) 255 SELECT Count(*) FROM temp GROUP BY name;
(c) 129 (d) 257
(GATE 2011: 2 Marks) Result:
count(*)
Solution: 2
The value of X is calculated using MX + 1 and Y
1
is calculated using 2 × MY + 1
1
For example, X = 1 and Y = 1
Ans. (b)
at the next step, X = MX + 1 = 1 + 1 = 2,
Y = 2 × MY + 1 = 2 × 1 + 1 = 3 51. Given the basic ER and relational models, which of
the following is INCORRECT?
X Y (a) A n attribute of an entity can have more than

one value.
1 1 (b) An attribute of an entity can be composite.
2 3 (c) In a row of a relational table, an attribute can
have more than one value.
3 7 (d) In a row of a relational table, an attribute can
4 15 have exactly one value or a NULL value.
5 31 (GATE 2012: 1 Mark)
6 63 Solution: In relational table, more than one value
7 127 for an attribute will violate the condition of first
normal form. So, option (c) is incorrect.
Ans. (a) Ans. (c)
Chapter 8.indd 441 4/9/2015 10:04:37 AM

52. Which of the following is TRUE? Solution: Two schedules are conflict-serializable
if the actions belong to different transactions, one
(a) Every relation is 3NF is also in BCNF.
of the actions is a write operation, the actions
(b) A relation R is in 3NF if every non-prime attri-
access the same object (read or write).
bute of R is fully functionally dependent on
every key of R. Two schedules are conflict-equivalent if both sched-
(c) Every relation in BCNF is also in 3NF. ules involve the same set of transactions, and the
(d) No relation can be in both BCNF and 3NF. order of each pair of conflicting actions in both
(GATE 2012, 1 mark) schedules is the same.
Take the following schedule:
Solution: Any relation that is in `x’NF, will auto-
matically be in `x-1’ NF. Now, let us see the order of S: r1(P); r1(Q); r2(Q); r2(P); w1(Q);w2(P)
NF as we discussed in the text, it is — 1NF, 2NF, 3NF,
BCNF, 4NF and 5NF. Therefore, option (c) is correct.
Ans. (c)
53. Suppose R1(A, B) and R2(C, D) are two relation
schemas. Let r1 and r2 be the corresponding rela- T1 T1
tion instances. B is a foreign key that refers to C in
R2. If data in r1 and r2 satisfy referential integrity
constrains, which of the following is ALWAYS
TRUE?
(a) ΠB(r1) − ΠC(r2) = ∅
Cycle exists between two transactions, so T1 and
(b) ΠC(r2) − ΠB(r1) = ∅
T2 are not conflict serializable.
(c) ΠB(r1) − ΠC(r2) A schedule is conflict-serializable when the schedule
(d) ΠB(r1) − ΠC(r2) ≠ ∅ is conflict-equivalent to one or more serial sched-
(GATE 2012: 2 Marks) ules. There are two possible serial schedules–T1
followed by T2 and T2 followed by T1. In both,
Solution: B is a foreign key in r1, which refers to one of the transactions reads the value written by
C in r2. another transaction as a first step.
r1 and r2 satisfy referential integrity constraints. Ans. (b)
So, every value that exists in column B of r1 must Common Data Questions 55 and 56: Consider the
also exist in column C of r2. following relations A, B and C:
Ans. (a)
A
54. Consider the following transactions with data Id Name Age
items P and Q initialized to zero:
12 Arun 60
T1: read (P); 15 Shreya 24
read (Q); 99 Rohit 11
if P=0 then Q:=Q+1;
write (Q). B
T2: read (Q); Id Name Age
read (P) 15 Shreya 24
if Q=0 then P:=P+1; 25 Hari 40
write (P). 98 Rohit 20
Any non-serial interleaving of T1 and T2 for con-
99 Rohit 11
current execution leads to
(a) a serializable schedule C
(b) a schedule that is not conflict-serializable
(c) a conflict-serializable schedule Id Phone Area
(d) a schedule for which precedence graph cannot 10 220 02
be drawn
99 2100 01
Chapter 8.indd 442 4/9/2015 10:04:37 AM

55. How many tuples does the result of the following 57. An index is clustered, if
relational algebra expression contain? Assume that
(a) It is on a set of fields that form a candidate key.
the schema of A ∪ B is the same as that of A.
(b) It is on a set of fields that include the primary key.
(A ∪ B) A.Id> 40  C.Id <15 C (c) The data records of the file are organized in the
same order as the data entries of the index.
(a) 7 (b) 4 (d) The data records of the file are organized not in
(c) 5 (d) 9 the same order as the data entries of the index.
(GATE 2013: 1 Mark)
Solution: Solution: In clustered index, the leaf nodes

A∪B contain the data page of the table concerned. Each
index row contains a key value and a pointer to
Id Name Age either the intermediate level (other than root and
12 Arun 60 leaf nodes) or data row in leaf level. Therefore, they
are organised in the same order.
15 Shreya 24
Ans. (c)
99 Rohit 11
58. Consider the following relational schema:
25 Hari 40
Students(rollno: integer, sname: string)
98 Rohit 20
Courses(courseno: integer, cname: string)
(A ∪ B)  A.Id> 40  C.Id <15 C Registration(rollno: integer, courseno;
integer, percent: real)
Id Name Age Id Phone Area Which of the following queries are equivalent to
12 Arun 60 10 2200 02 this query in English?
15 Shreya 24 10 2200 02 “Find the distinct names of all students who score
99 Rohit 11 10 2200 02 more than 90% in the course numbered 107”
25 Hari 40 10 2200 02 I. S
ELECT DISTINCT S.sname FROM Students
98 Rohit 20 10 2200 02 as S, Registration as R WHERE
R.rollno=S.rollno AND R.Courseno=107
AND R.percent>90
99 Rohit 11 99 2100 01
98 Rohit 20 99 2100 01
II. P sname (s courseno =107 L percent > 90(Registration students))
Ans. (a)
P sname (s courseno =107 L percent > 90(Registration students))
56. How many tuples does the result of the following
SQL query contain? III. { T |∃S ∈ Student, ∃R ∈ Registration (S.rollno =
R.rollno ∧ R.courseno = 107 ∧ R.percent > 90 ∧
SELECT A.Id T. sname = S.name)}
FROM A {<SN> | ∃SR∃RP(<SR, SN> ∈ Student∧ < SR,
WHERE A.Age > ALL (SELECT B.Age 107, RP> ∈ RP > 90)}
FROM B (a) I, II, III and IV (b) I, II and III only

(c) I, II and IV only (d) II, III and IV only
WHERE B.Name = ‘Arun’)
(a) 4 (b) 3 (c) 0 (d) 1
Solution: All of the query statements are equiva-
lent to the given query.
Solution: ALL returns the values to be greater Ans. (a)
than all the values returned by the sub-query. Linked Answer Questions 59 and 60: Relation R
There is no entry with the name Arun in Table B. has eight attributes ABCDEFGH. Fields of R con-
So, the sub-query will return NULL. If a sub-query tain only atomic values. F = {CH → G, A → BC,
returns NULL, then the condition becomes true for B → CFH, E → A, F → EG} is a set of functional
all rows of A. dependencies (FDs) so that F+ is exactly the set of
Ans. (b) FDs that hold for R.
Chapter 8.indd 443 4/9/2015 10:04:38 AM

59. How many candidate keys does the relation R have? (c) In 3NF, but not in BCNF
(d) In BCNF
(a) 3 (b) 4
(c) 5 (d) 6 (GATE 2013: 2 Marks)
Solution: A → BC, B → CFH and F → EG are
Solution: Candidate keys AD, BD, ED and FD
partial dependencies.
Ans. (b)
So, they are in 1NF but not in 2NF.
60. The relation R is Ans. (a)
(a) In 1NF, but not in 2NF
(b) In 2NF, but not in 3NF
PRACTICE EXERCISES
Set 1
1. A relational database can be defined as 7. Which of the following is a unary operation?

(a) a place to store relational information. (a) Join (b) Project
(b) a database that relates to different operations. (c) Intersection (d) Cartesian product
(c) a database to relate the relationship of different
databases. 8. What is the least and ceiling number of rows in the
(d) a database related to storing of information join selection of two tables having m rows and n
about humans. rows, respectively?
2. The minimal superkeys are called (a) 1 and m × n (b) 0 and m + n
(c) 1 and infinite (d) 0 and mn
(a) minimum keys. (b) candidate keys.
(c) domain keys. (d) optimal keys. 9. An attribute of one table having the same value of
a primary key of another table is termed as
3. Which of the following combinations represents a
valid data model? (a) foreign key. (b) alternate key.
(c) candidate key. (d) transitive key.
(a) E-R model, network data model, normal form
data model 10. A database has five attributes defined as {A, B, C,
(b) Relational data model, hierarchical data model, D, E}. The functional dependency is defined by
SQL data model {A → BC, E → FG}
(c) Object-based data model, table data model
The candidate key of this table is
(d) E-R model, relational data model, object-based
data model (a) A. (b) AE.
(c) E. (d) BC and FG.
4. Which of the following is not a database model?
11. Consider the following set of functional dependency
(a) Network database model
(b) Relational database model {AB → CD, CD → E, E → A} AB, CD are
(c) Object-oriented database model candidate keys.
(d) Normal form data model It is in which normal form?
5. Which one of the following allows accessing or (a) 2NF (b) BCNF (c) 3NF (d) 4NF
maintaining data in a database?
12. The term physical data independence can be
(a) DCL (b) DML
defined as
(c) DDL (d) None of the above
(a) If we modify the conceptual schema, then there
6. Which of the following is a binary operation? is no need to modify the physical schema.
(a) Select (b) Project (b) If we modify the physical schema, then there is
(c) Rename (d) Union no need to modify the conceptual schema.
Chapter 8.indd 444 4/9/2015 10:04:39 AM

(c) If we modify the conceptual schema, then there 20. The value of particular field should be less than 50
is no need to modify the external schema. if it is
(d) If we modify the middle level schema, then
(a) An integrity constraint.
any one schema (internal/external) may be
(b) A referential constraint.
modified.
(c) A primary key constraint.
13. Relational algebra can be defined as (d) A normalization constraint.
(a) Procedural. (b) Non-procedural. 21. The ER model of a database includes

(c) User dependent. (d) Domain dependent.
I. Different entity types.
14. Which of the following is true? II. Attributes for each entity type.
III. Relationships among entity types.
(a) A relation in BCNF is always in 3NF. IV. Semantic integrity constraints are not defined.
(b) A relation in 3NF is always in BCNF.
(c) PJNF and 3NF are same. Which of the above statements is correct?
(d) A relation in PJNF is not in 3NF.
(a) All of the above (b) Only I, II and III
15. RAD stands for . (c) III and IV (d) II and III
(a) Read and destroy 22. The cardinality ratio and participation
(b) Read and develop of a table helps us to implement either the cross
(c) Rapid application development referencing design or mutual referencing design.
(d) Rapid application design
(a) Functionality (b) Constraints
16. In DML, RECONNCT command cannot be used (c) SQL queries (d) Domains
with
23. Consider the following schemas:
(a) CLEAN set. (b) FIXED set.
(c) OPTIONAL set. (d) USER set. Branch = (Branch-name, Assets, Branch-city)
Customer = (Customer-name, Bank name,
17. The UWA is used to communicate the contents of
Customer-city)
individual records between
Borrow = (Branch-name, loan number,
(a) Host program and present record. customer account-number)
(b) Host program and host record. Deposit = (Branch-name, Accountnumber,
(c) Host program and DBMS. Customer-name, Balance)
(d) Host program and host language.
Using relational algebra, the query that finds cus-
18. Referential integrity is concerned with tomers having deposits have accountnumber >
23456 is
(a) Primary key. (b) Foreign key.
(c) Alternate key. (d) Project_Join key. (a) Πcustomer-name (s account-number > 23456 (Deposit)
(b) s customer-name (s account-number > 23456 (Deposit)
(c) Πcustomer-name (s account-number > 23456 (Borrow)
19. Match the following:
(d) s customer-name (Π account-number > 23456 (Borrow)
List — I List — II
I. DDL A. LOCK TABLE 24. What deletes the entire file except the file structure?
II. DML B. COMMIT (a) DELETE
III. TCL C. Natural Difference (b) DELETE RECORDS
(c) ZAP
IV. BINARY D. REVOKE
(d) RETAIN STRUCT
Operation
25. The function of a Transaction Manager is to
Codes:
I. Maintain a log of transactions.
I II III IV II. Maintain database images before and after a
(a) A D B C transaction.
(b) A B D C III. Maintain appropriate concurrency control.
(c) D B A C IV. Avoid deadlocks.
(d) D A B C
Chapter 8.indd 445 4/9/2015 10:04:39 AM

(a) I, II, III and IV (b) II and III 32. A network schema is used to
(c) I, III and IV (d) I, II and III
(a) restrict to many-to-many relationship.
(b) permit many-to-many relationship.
26. The term granularity relates to
(c) permit to store data in a database.
(a) the size of table. (d) help in storing one-to-many relationship.
(b) the size of data item.
(c) the size of tuple. 33. Storing under unified scheme at a single site is
(d) the size of primary key called as
(a) data mining. (b) universal database.
27. Match the following: (c) data warehousing. (d) advanced database.
34. The task of correcting and cleaning data is called as
I. OLAP A. Back propagation
(a) data organization. (b) pre-processing.
II. OLTP B. Data warehouse
(c) data mining. (d) data structuring.
III. Decision tree C. RDBMS
35. A clustering index consists of .
IV. Neural network D. Classification
(a) declared and ordered primary key
(b) both grouped and ordered primary key and for-
I II III IV eign key
(c) ordered foreign key
(a) D C A B
(d) grouped and ordered candidate key and alter-
(b) B C D A nate key
(c) A B C D
36. Consider the following schema:
(d) D B A C
Emp(Empcode, Name, Sex, Salary, Deptt)
A simple SQL query is executed as follows:
28. Which of the following statements is TRUE?
SELECT Deptt FROM Emp
(a) A prime attribute can be transitively depen-
dent on a key in 5NF relation. WHERE sex = ‘M’
(b) A prime attribute can be transitively depen- GROUP by Dept
dent on a key in 3NF relation.
HAVING avg (Salary) > {select avg (Salary)
(c) A prime attribute can be transitively depen-
from Emp}
dent on a key in PJNF relation.
(d) A prime attribute cannot be transitively depen- The output will be
dent on a key in BCNF relation.
(a) a verage salary of male employee is the average
29. Which normal form is considered as adequate for a salary of the organization.
simple database design? (b) average salary of male employee is less than the
average salary of the organization.
(a) 2NF (b) 3NF
(c) average salary of male employee is equal to the
(c) BCNF (d) PJNF
average salary of the organization.
30. Which of the following is not a type of DBMS? (d) average salary of male employees in a depart-
ment is more than the average salary of the
(a) Hierarchical (b) Network organization.
(c) Relational (d) Graphical
37. Consider a relation X(p, q, r), and the domains are
31. Manager’s salary details are to be hidden from represented by their atomic values. Further, the
Employee Table. This technique is called as functional dependency p → q, q → r is observed,
the relation is in
(a) internal level datahiding.
(b) physical level datahiding. (a) 1NF, not in 2NF
(c) external level datahiding. (b) 2NF, not in 3NF
(d) logical level datahiding. (c) 3NF
(d) 1NF, not in 3NF
Chapter 8.indd 446 4/9/2015 10:04:39 AM

38. Match the following: Name Age Occupation Category

I. Foreign keys A. Domain constraint Jenifer 28 DOC B
II. Private key B. Referential integrity Maya 32 SER D
III. E
vent control C. Password Dev 24 MUS C
action model
There is an index file associated with this and it
IV. Data security D. Trigger
contains the values 1, 3, 2, 5 and 4. Which one of
the fields in the index built from?
Codes:
(a) Age (b) Name
I II III IV (c) Occupation (d) Category
(a) C B A D 4. Consider the following database relations contain-
(b) B A D C ing the attributes
(c) C D A B Book_id
(d) A B C D Subject_Category_of_books
Name_of_Author
Set 2 Nationalilty_of_Author
With Book_id as the primary key
1. For a database relation R(a, b, c, d), where the
domains of a, b, c, d include only atomic values, (a) T
he highest normal form satisfied by this rela-
only the following functional dependencies and tion is – NF.
those that can be inferred from them holds. (b) Suppose the attributes Book_title and Author_
address are added to the relation, and the pri-
a→c mary key is changed to {name_of_Author,
b→d Book_title}, the highest normal form satisfied
This relation is by the relation is – NF.
(a) In First normal form but not in Second normal 5. Consider the scheme R = (S T U V) and the
form. dependencies S → T, T → U, U → V and V → S.
(b) In Second normal form but not in Third normal Let R = (R1 and R2) be a decomposition such that
form. R1 ∩ R2 = f. The decomposition is
(c) In Third normal form. (a) Not in 2NF
(d) None of the above. (b) In 2NF but not 3NF.
2. Let R(a, b, c) and S(d, e, f) be two relations in (c) In 3NF but not 2NF.
which d is the foreign key of S that refers to the (d) In both 2NF and 3NF.
primary key of R. Consider the following four oper- 6. Consider a schema R(A, B, C, D) and functional
ations on R and S: dependencies A → B and C → D. Then the decom-
I. Insert into R II. Insert into S position of R into R1(AB) and R2(CD) is
III. Delete from R IV. Delete from S (a) Dependency preserving and lossless join.
Which of the following is true about the referential (b) Lossless join but not dependency preserving.
integrity constraint above? (c) Dependency preserving but not lossless join.
(d) Not dependency preserving and not lossless join.
(a) None of I, II and IV can cause its violation.
(b) All of I, II, III and IV can cause its violation. 7. Relation R with an associated set of functional
(c) Both I and IV can cause its violation. dependencies, F, is decomposed into BCNF. The
(d) Both II and III can cause its violation. redundancy (arising out of functional dependencies)
in the resulting set of relation is
3. There are five records in a database
(a) Zero.
Name Age Occupation Category (b) More than zero but less than that of an equiva-
lent 3NF decomposition.
Rama 27 CON A
(c) Proportional to the size of F+.
Abdul 22 ENG A (d) Indeterminate.
Chapter 8.indd 447 4/9/2015 10:04:39 AM

8. From the following instance of a relation scheme 13. Which of the following relational calculus expres-
R(A, B, C), we can conclude that sions is not safe?
A B C (a) {t|∃ ∈ R1(t[A] = u[A]) ∧ ¬ ∃s ∈ R2(t[A] = s[A])}

1 1 1 (b) {t|∀u ∈ R1(u[A] = “x”⇒ ∃s ∈ R2 (t[A] = s[A]
∧ s[A] = u[A]))}
1 1 0
(c) {t|¬ (t ∈ R1)
2 3 2
(d) {t|∃ ∈ R1(t[A] = u[A]) ∧ ∃s ∈ R2(t[A] = s[A])}
2 3 2
14. With regard to the expressive power of the formal
(a) A functionally determines B and B functionally relational query languages, which of the following
determines C. statement is true?
(b) A functionally determines B and B does not
(a) R elational algebra is more powerful than rela-
functionally determine C.
tional calculus.
(c) B does not functionally determine C.
(b) Relational algebra has the same power as rela-
(d) A does not functionally determine B and B
tional calculus.
does not functionally determine C.
(c) Relational algebra has the same power as safe
9. Which of the following is/are correct? relational calculus.
(d) None of the above.
(a) An SQL query automatically eliminates
duplicates. 15. For the schedule given below, which of the follow-
(b) An SQL query will not work if there are no ing is correct?
indexes on the relations. 1 Read A
(c) SQL permits attribute names to be repeated in
2 Read B
the same relation.
(d) None of the above. 3 Write A
4 Read A
10. Given relation r(w, x) and s(y, z), the result of
select distinct w, x from r, s is guaranteed to be 5 Write A
same as r, provided 6 Write B
(a) r has no duplicates and s is non-empty. 7 Read B
(b) r and s have no duplicates. 8 Write B
(c) s has no duplicates and r is non-empty.
(d) r and s have the same number of tuples. (a) T his schedule is serializable and can occur in a
scheme using 2PL protocol.
11. Which of the following query transformations (b) This schedule is serialzable but cannot occur in
(i.e. replacing the LHS expression by the RHS a scheme using 2PL protocol.
expression) is incorrect? R1 and R2 are relations, (c) This schedule is not serializable but can occur
C1 and C2 are selection conditions and A1 and A2 in a scheme using 2PL protocol.
are attributes of R1. (d) This schedule is not serializable and cannot
occur in a scheme using 2PL protocol.
(a) s c (s c (R1 )) → s c (s c (R1 ))
1 2 2 1
(b) s c (ΠA (R1 )) → ΠA (s c (R1 )) 16. Which of the following is correct?

1 1 1 1
(c) s c (R1ER2 ) → s c (R1 ) E s c (R1 ) (a) B

trees are for storing data on disk and B+
1 1 1
(d) ΠA (s c (R1 )) → s c (ΠA (R1 )) trees are for main memory.

2 1 1 2 (b) Range queries are faster on B+ trees.
(c) B trees are for primary indexes and B+ trees
12. The relational algebra expression equivalent to the
are for secondary indexes.
following tuple calculus expression:
(d) The height of a B+ tree is independent of the
{t|t ∈ r ∧ (t[A] = 10 ∧ t[B] = 20)} is number of records.
(a) s(A = 10 ∨ B = 20)(r) 17. B+ trees are preferred to binary trees in database
(b) s(A = 10(r) E sB = 20(r) because
(c) s(A = 10(r) ∩ sB = 20(r) (a) d
isk capacities are greater than memory
(d) s(A = 10(r) − sB = 20(r) capacities.
Chapter 8.indd 448 4/9/2015 10:04:40 AM

(b) disk access is much slower than memory access. student names are of length 8 bytes, disk blocks are
(c) d
isk data transfer rates are much less than of size 512 bytes and index pointers are of size 4
memory data transfer rates. bytes. Given this scenario, what would be the best
(d) disks are more reliable than memory. choice of the degree (i.e., the number of pointers
per node of the B+ tree?
18. A B+ tree index is to be built on the name attri-
bute of the relation STUDENT. Assume that all
Set 1
1. (a) 8. (d) 15. (c) 22. (b) 29. (b) 36. (d)
2. (b) 9. (a) 16. (b) 23. (a) 30. (d) 37. (a)
3. (d) 10. (b) 17. (c) 24. (c) 31. (c) 38. (b)
4. (d) 11. (c) 18. (b) 25. (d) 32. (b)
5. (a) 12. (b) 19. (d) 26. (b) 33. (c)
6. (d) 13. (a) 20. (a) 27. (b) 34. (b)
7. (b) 14. (a) 21. (a) 28. (d) 35. (a)
Set 2
1. (a) The relation has the following functional Candidate keys for this relation are {S, T, U, V }
dependencies: As given R1 ∩ R2 = f
a→c So, R1 and R2 might have two attributes. A rela-
b→d tion with two attributes satisfies both 2NF and
{a b} is key for the given relation. 3NF conditions.
Here prime attributes are a and b. 6. (c) Dependency is preserved because there is no
loss of functional dependencies.
c and d are dependent on partial keys, this is not Since R1(AB) and R2(CD) do not have any common
allowed in 2NF. So, the given relation is in 1 NF. attribute. So, it is lossy join.
2. (d) Insertion into S and deletion from R can cause 7. (a) Redundancy is zero when we decompose into
inconsistency. So, II and III can cause violation. BCNF.
3. (c) The index is built from field occupation where 8. (b, c) F: X → Y implies that for any two tuples if
column is sorted in alphabetic order CON, DOC, t1[X] = t2[X], they must have t1[Y] = t2[Y].
ENG, SER and MUS. Index is assigned according
9. (d) DISTINCT key word is used to remove dupli-
to alphabetic position.
cate rows along with SELECT keyword. SQL does
4. (a) 2NF, (b) 3NF not permit attribute names to be repeated in the
same relation.
5. (d) The relation schema R = (S T U V) has follow-
ing functional dependencies 10. (a) Cross join with an empty relation produces the
S→T overall result as empty. Therefore, S is not empty. `r’
should not have any duplicates to have same rows as `r’.
T→U
U→V 11. (a) s c (s c (R)) and (s c (s c (R))) will not be equiv-
1 1 2 1
V→S alent if conditions c1 and c2 are not equivalent.
Chapter 8.indd 449 4/9/2015 10:04:41 AM

12. (c) Tuple calculus expression select tuples where Leaf nodes are linked together in B+ trees, hence
A = 10 and B = 20 of relation R. So, the equiva- range queries are faster.
lent relational algebra expression is s(A = 10(r) ∩
sB = 20(r). 17. (b) In B+ trees, disk access is slow and numbers of
hits are less. B+ trees have very high fan-out (typi-
13. (c) The query −t|¬ (t ∈ R1) is syntactically cor- cally on the order of 100 or more), which reduces
rect. But it requests for all tuples that are not in the number of I/O operations required to find an
R1. That set of such t tuples is obviously infinite, element in the tree.
in the context of such as the infinite domain set of 18. (43) Let n be the degree.
all integers. Therefore, this is unsafe query.
Given k key size (length of the name = 8 byte attri-
14. (c) Both relational algebra and relational calculus bute of student)
has same expressive powers. Every query written
in relation algebra can be expressed in relational Disk block size, B = 512 bytes
calculus. Index pointer size, b = 4 bytes
15. (d) Given schedule has W1(A), R2(A) and W2(B), Degree of B+ tree can be calculated if we know
R1(B) conflicting pairs. So the schedule is not seri- the maximum number of key an internal node can
alizable. Hence, it cannot occur in scheme using have; the formula for that is
2PL protocol. (n − 1)k + n × b = Block size
16. (b) Most database systems use indexes built on (n − 1)8 + n × 4 = 512
some form of a B+ tree due to its many advan- 12n = 520
tages, in particular its support for range queries. n = 520/12 = 43
Chapter 8.indd 450 4/9/2015 10:04:41 AM

UNIT IX : INFORMATION SYSTEMS AND SOFTWARE ENGINEERING
4
Number of questions
Marks 1
2 Marks 2
1
0
2015 2014 2013 2012 2011 2010
Year Concept
2015 Function point metric, Seeded faults, McCabe's complexity, COCOMO
Model, SRS document, Code walkthrough and inception, Testing
2014 Cohesion and coupling
2013 Cohesion and coupling
2012 IEEE floating-point representation
2011 Lines of code, COCOMO model, SRS, black-box testing
2010 Software development life cycle
Chapter 9.indd 451 4/9/2015 10:06:35 AM

Chapter 9.indd 452 4/9/2015 10:06:35 AM
CHAPTER 9
INFORMATION SYSTEMS AND SOFTWARE ENGINEERING
Syllabus: Information gathering, Requirement and feasibility analysis, Data flow diagrams, Process specifications,
Input/output design, Process life cycle, Planning and managing the project, Design, Coding, Testing, Implementation,
Maintenance.
9.1 INTRODUCTION 9.2 INFORMATION SYSTEMS
Software engineering is the study of design, develop- Information system consists of components such
ment and maintenance of software. Real-time problems as storing and processing data. All business firms
are so large and complex that they cannot be solved and organisations rely on them to carry out and
by single developer. Software engineering defines teams manage their operations, interact with their custom-
and its quality. It is concerned about all the aspects of ers and suppliers, and compete in the market. For
software development from feasibility of software to its example, these organisations could reach their cus-
maintenance. The team in software engineering consists tomers through Internet for the processing of finan-
of developers, testers, designers, customers, managers, cial accounts and manage their human resources.
etc. Every team member works to fulfil task assigned Other users can rely on information system through
to him. online services such as social networking, auctions,
Chapter 9.indd 453 4/9/2015 10:06:35 AM

454 Chapter 9: Information Systems and Software Engineering
Executive
information
systems
Decision
support systems
Management
reporting systems
Management Support
Professional Collaboration Knowledge

support systems systems management
systems
Support of Knowledge Work
Supply chain Customer Transaction

management relationship processing systems
management
Operational Support
Figure 9.1 | An information system.
banking, entertainment, shopping, etc. Information communications among employees. OIS is also known as
and knowledge have become vital economic resources. office automation. With the help of OIS, employees or
Information systems support operations, knowledge users can perform tasks electronically using computers
work and management in organisations (Fig. 9.1). and other electronic devices, instead of manually pro-
cessing data. For example, a registration department can
9.2.1 Components of Information Systems post the class schedule via email to students, whereas
in a manual system, the registration department would
The main components of information systems are computer have to photocopy the schedule and mail it to each stu-
hardware and software, telecommunications, databases and dent’s house.
data warehouses, human resources and procedures (Fig. 9.2).
OIS supports a large range of business office activities
The hardware, software and telecommunications constitute
such as creating and distributing graphics and/or docu-
information technology (IT), which is now ingrained in the
ments, sending messages, scheduling and accounting. All
operations and management of organisations.
levels of users utilise and benefit from the features of
an OIS.
Environment
Effector The software used in OIS includes word processing,
(output) spreadsheets, databases, presentation graphics, email,
Processor Memory Web browsers, Web-page authoring, personal informa-
Receptor tion management and groupware. OIS uses communica-
(input)
tion technology such as voice mail, fax, video conferencing
and electronic data interchange (EDI) for the electronic
Figure 9.2 | Structure of an information system.
exchange of text, graphics, audio and video. The hard-
ware includes computers, modems, video cameras, speak-
ers and microphones, scanners and fax machines.
9.2.2 Types of Information Systems
9.2.2.1 Office Information System 9.2.2.2 Transaction Processing Systems
Office information system (OIS) uses hardware, soft- A transaction processing system (TPS) captures and pro-
ware and network to enhance the workflow and facilitate cesses data generated during an organisation’s day-to-day
Chapter 9.indd 454 4/9/2015 10:06:36 AM

9.2 INFORMATION SYSTEMS 455
transactions. A transaction is a business activity such as 1. Internal sources of data: A data from an organ-
a deposit, payment, order, reservation, etc. isation such as manufacturing, inventory, sales or
Transaction processing includes the following activities: financial data.
2. External source of data: A data from other
1. Recording a business activity such as registration, organisations such as population trends, interest
order, payment, etc. rates, construction rates, price of raw material, etc.
2. Confirming an action or triggering a response such
as printing documents, sending a response, gener- DSS may also include query language, statistical analy
ating pay cheque, issuing receipts, etc. sis capabilities, spreadsheets and graphics for the
3. Maintaining data such as adding new data, modi- extraction and evaluation of data, if it is very large.
fying data or deleting data. DSS includes some kind of capabilities for the creation
of a model of the factors affecting the decision. In these
For the processing of business data, TPS was the first models the user can change the values of all or few
computerised system developed as a function called data parameters and project the results. There are lots of
processing. It is an existing manual system, which allows application software packages which are used for the
faster processing, reduced clerical costs and improved decision-making process such as executive information
customer service. system (EIS), which supports the information needs of
Online transaction processing (OLTP) is an example of executive management. EIS uses data in the form of
TPS. The administrative users use OLTP to enter data and charts or tabular forms to show trends, ratios and other
take print out of the entered data in the form of receipts. managerial statistics.
9.2.2.3 Management Information Systems

9.2.2.5 Expert Systems
Management information system (MIS) generates accu-
rate, timely and organised information, and based on An expert system is an information system, which is
that user can make decisions, solve problems, supervise responsible for capturing and storing human knowledge
activities and track progress. MIS generates reports on a and imitating human reasoning and decision-making pro-
regular basis, a management information system called cesses. It consists of two main components: knowledge
management reporting system (MRS). MIS interacts base and inference rules.
with TPS. On the basis of activities performed by TPS,
1. Knowledge base: It is a combination of subject
MIS generates reports of daily activities. The three basic
knowledge and experiences of human experts.
types of information it produces is as follows:
2. Inference rules: It is a set of logical judgement
1. Detailed information: It confirms transaction applied to knowledge base when a user describes a
processing activities, for example, detailed order situation to the expert system.
report.
2. Summary information: It consolidates data into It is used in decision-making at any level in an organisa-
a format, which can be reviewed quickly and easily tion. It is also used to resolve situations such as diagnos-
by a user, for example, an inventory summary ing illnesses, searching for oil, making soup, etc.
report, which includes totals, tables or graphs. It is a part of an artificial intelligence. Artificial intel-
3. Exception information: It filters data to report ligence (AI) is an application of human intelligence to
information, which is outside of a normal condition computers. It senses human actions based on logical
called the exception criteria. An exception criterion assumptions and prior experience, then take appropriate
is defined as the range of what is considered normal action to complete the task. It can perform various func-
activity or status; for example, exception report, tions such as speech recognition, logical reasoning and
which notifies the items it needs to reorder. It helps creative responses.
in saving time.
9.2.2.4 Decision Support Systems 9.2.2.6 Integrated Information Systems
Decision support system (DSS) is a computer-based As technology widens its horizons, it becomes difficult
information system designed to support decision- to classify a system belonging uniquely to one of the
making activities such as business or organisation. above five types of information system. Today’s applica-
A variety of DSSs exist having support of wide range tions supporting processing of transactions, management
of decisions. It uses data from two sources: internal of information and decision-making need an integrated
and external. information system.
Chapter 9.indd 455 4/9/2015 10:06:36 AM

Types of Information Systems 2. Software does not wear out, but deteriorates. In
hardware design, at early development, the failure
The following are six major types of information sys-
rate is high due to undiscovered errors. After cor-
tems corresponding to each organisational level (the four
recting the defects, the failure rate is decreased and
levels shown in Fig. 9.1):
the component starts running at a steady state.
1. Transaction processing systems (TPS) to serve the However, over a period of time, due to external
operational level of an organisation. conditions, such as temperature, air, dust and envi-
2. Knowledge work systems (KWS) to keep an orga- ronmental maladies, the failure rate again increases
nization up-to-date in knowledge in terms of tech- and the hardware starts to wear out. When it
nology, science, art and social thoughts. undergoes wear out, we can replace the component
3. Office automation systems (OAS) to serve the with a new one. In software design, at early devel-
knowledge level of an organisation. opment, the failure rate is high due to undiscov-
4. Decision-support systems (DSS) to analyze the ered errors. After correcting the defects the failure
existing information to forecast the effects of their rate is decreased and the software starts running
decisions in the near future. within the constrained level. After a period of time,
5. Management information systems (MIS) to serve another change is injected in the existing soft-
the management level of the organisation. ware due to which the failure rate again increases,
6. Executive support systems (ESS) to serve the stra- and after correction the failure rate decreases. So,
tegic level of an organisation. during the lifetime of software, when the require-
ments keep changing, the software undergoes dete-
9.3 SOFTWARE rioration (Fig. 9.3).
3. The industry is moving towards component-based
development, but software is still being custom-
Software is a program, and when it is executed the built. As components are error-free codes, they are
desired functionality and performance must be satis- reusable. With component-based development for
fied. Software is a data structure used to manipulate developing some hardware, time and development
the information. It is a document that describes the cost are reduced as the design consists of IC num-
operation and use of a program. Finally, it is a logical bers which are already available. So, at the time
entity rather than physical entity. of implementation we can directly use them. For
softwares also, we should use reusable components.
9.3.1 Characteristics of Software Software development also uses component names
such as:
The following are the characteristics of a software:
1. Off the shelf: This component is used in the proj-
1. Software is developed or engineered, but not manu- ect from the third party, that is, the component is
factured in classical sense. Although the develop- not available in the library of the developing organ-
ment approach for both the hardware design and isation.
software design is same, but after implementation 2. Fully experienced and partially experi-
of the hardware design we get the physical com- enced: This component is derived based on the pre-
ponent and after implementation of the software vious project, so can be directly used in the current
design we get the logical component. In both cases project.
quality is an important factor, that means when 3. New component: This component is developed
the output is operational the associated functional- from the base line, means from the initial stage or
ity must be satisfied. scratch.
F
a Deteriorate change
i Reengineer the software
l
u Wear out
r
e
r Steady change
a
t
e Actual curve
Time Time
Figure 9.3 | Wear out and deterioration of software.
Chapter 9.indd 456 4/9/2015 10:06:36 AM

9.3 SOFTWARE 457
9.3.2 Software Engineering 1. Corrective support: In this support, unidentified

errors are going to be managed. The corresponding
Software engineering is the systematic, disciplined and maintenance is called as corrective maintenance.
quantitative approach for the development, operation 2. Adaptive support: Over a period of time, if
and maintenance of the software. the customer wants to change the platform, that
Software engineering is described in the following two is, CPU, memory, operating system or external
approaches: interfaces, there is a need of adaptive maintenance.
3. Perfective (enhancement) support: Over
1. Layered approach: The layered approach con- a period of time, if the customer or the end user
sists of the following four layers (Fig. 9.4): wants to introduce new functionality into the exist-
•• Quality: Confirmation to explicitly stated form ing software to maintain the new software, perfec-
as mentioned in the document. tive maintenance is required.
•• KPA: Key process areas such as planning, 4. Preventive support: When a customer keeps
schedule, performance and quality attribute. on changing the requirement, the software under-
•• Methods: “How to” process model. goes deterioration. To maintain such a case,
•• Tools: Automated and semi-automated tool there is a need of preventive maintenance or
available to develop the software. re-engineering.
9.3.2.2 Software Process
A process gives the framework to develop efficient soft-

ware. On the basis of the process used to develop the
software, the organisation undergoes assignment to
assess if there is a need of a maturity model.
The Software Engineering Institute (SEI) has intro-
duced a maturity model to assess an organisation. This
Tools model is called as the capability maturity model (CMM).
Method The CMM consists of the following five (5) maturity
Process levels:
Quality
1. Level 0 (Initial): In this level, there is no stan-
Figure 9.4 | Layered approach of software engineering. dard maintained to develop the software. This
level uses the adhoc procedure, and development
2. Generic approach: The generic approach con- becomes chaotic. Success depends on the individual
sists of the following three stages: levels.
•• Definition: This stage is focused on `What’, that 2. Level 1 (Repeatable): In this level, based on the
is, what information is processed, what functional- knowledge of previous project management, activi-
ity and performance is desired, what system behav- ties are developed to trace the cost, schedule and
iour is expected, what constraints are imposed and performance. There is no guarantee to the success
what validation criteria is used. In this stage, anal- scenario of the current project.
ysis and requirement operations are performed. 3. Level 2 (Defined): In this level, project man-
•• Validation: Àre we building the right prod- agement and engineering activities are defined.
uct?’. The right condition is injected in the test- All the standards are there in the diagrammati-
ing phase of the development to cover the errors. cal approach, they are not prepared for anticipated
•• Development: This stage is focused on `how’, that situation.
is, how to process the information, how to define the 4. Level 3 (Managed): In this level, project man
data structure, how to maintain the interface, how agement and engineering activities are quan-
to convert procedural description to machine read- titatively collected and documented. Quality
able code, and how to develop test case etc. This attributes are also defined. Because of the lack
case performs design, coding and testing operations. of continuous improvement, minimum quality is
assured.
9.3.2.1 Support 5. Level 4 (Optimised): In this level, more profit is
gained with minimum effort while maintaining the
After delivering, the project support stage is required to continuous improvement process model, innovative
maintain the software. The following four types of sup- ideas, tools and techniques, and qualitative feed-
ports or maintenance are required: back from the customers.
Chapter 9.indd 457 4/9/2015 10:06:37 AM

9.4 PROCESS MODELS •• Identifying the interfaces: It represents the

interconnection between different modules pres-
ent in the software.
There are two types of process models to develop the •• Defining procedural details: It converts the
software (Fig. 9.5). high-level abstraction into low-level abstraction
(algorithmic approach).
3. Coding: It translates the procedural details into
9.4.1 Conventional Process Model
machine readable form.
4. Testing: In this stage, different test cases are
In order to develop a software, a team of software engi-
implemented to cover the logical errors and func-
neers follow a development strategy specifying the process,
tional errors.
methodology, tools and phases. This is referred to as process
5. Supports: After delivering the software to the
model. The different types of software models are as follows:
customer, support stage is involved to maintain the
software and to adapt to the changes.
9.4.1.1 Waterfall Model
Note: When the requirements are clear, use this model
Waterfall model, also known as linear sequential to develop the software. But the final product will be
model, provides a systematic and sequential approach available after a long time schedule.
for the development of a software starting from anal-
ysis phase to designing, coding, testing and support 9.4.1.2 Prototype Model
(Fig. 9.6).
1. Requirement gathering: On the basis of the When customer requirements are not clear, the pro-
business objective, the input domain and output totype model is used to finalise the requirement. The
domain are defined. According to the IEEE stan- framework of prototype model is shown in Fig. 9.7.
dard, requirement is defined as a condition or a In this model, listening to the customer is the entry
capability required by the user to interact with the point to prepare the software requirement specification
system or “A documented representation of condi- (SRS). On the basis of the SRS, the designer performs the
tional capability is called as requirement”. quick design, quick code. Quick test mock up. The proto-
2. Design: In this stage, the following four opera- type model is evaluated by the customer, and based on the
tions are performed: feedback of the customer its iteration continues until the
•• Creating the data structure: It identifies the customer is satisfied. After finalising the SRS, the same
data objects and attributes and creates the rela- process model is used to build the final product, which is
tionship between the data objects. called as open-end prototyping or evaluating prototyping
•• Creating the software architecture: It creates model. After finalising the SRS, the developer can choose
the blue print, that is, the skeleton to represent another process model to build the final product, which is
the inflows and outflows of the software. called as throwaway prototyping or close-end prototyping.
Process model
Conventional Evolutionary
(Static) (Dynamic)
Component-based
Waterfall Prototype RAD Incremental Spiral
development
Figure 9.5 | Classification of process model.
Requirement Design Coding Testing Support

analysis
Figure 9.6 | Waterfall model.
Chapter 9.indd 458 4/9/2015 10:06:37 AM

9.4 PROCESS MODELS 459
Framework of the RAD model is shown in Fig. 9.8.

1. Business modelling: It identifies the business
objective. On the basis of business objective, we
can gather data objects.
Customer communication Quick design
2. Data modelling: It creates the data structure,
OR and
that is, identifies the data objects and attributes
Listen to customer Testing makeup
and relationship.
3. Process modelling: It creates the information
flow in the project by modelling, adding or deleting
the data object.
4. Code generation: By using a fourth-generation
technique, it translates the procedural details into
machine-readable format.
Customer Test case: Different test cases are developed and
evaluation implemented to cover the error. At the end of the test
state, all the modules are available with error-free code,
which can be integrated to build the final product. RAD
Figure 9.7 | Prototype model. model is suitable only when the organisation has efficient
manpower.
9.4.1.3 Rapid Application Development
9.4.2 Evolutionary Process Model
The rapid application development (RAD) model is used
when the requirements are clear but the time schedule In the evolutionary process model, the software is devel-
is very short. In this model, the project is divided into oped in version basis because requirements are not static.
modules and develops the model simultaneously. To When requirement changes, that change is reflected in
maintain effective modularity make sure that the model the version. Under this case, different models are used to
has high cohesion and low coupling. develop effective software.
1. Coupling: It is a quantitative measure of func-
tional strength of a module. 9.4.2.1 Incremental Model
2. Cohesion: It is a quantitative measure of to what
extent the module is independent from another In this model, classic life cycle is used to develop the
module. software. See Fig. 9.9.
Team 1 Team 2 Team 3
Business Business Business

P modelling modelling modelling
R
Data modelling Data modelling Data modelling
O
C
Process modelling Process modelling Process modelling
E
S
Code generation Code generation Code generation
S
Test turnover Test turnover Test turnover
Time
Figure 9.8 | RAD model.
Chapter 9.indd 459 4/9/2015 10:06:38 AM

Analysis Design Coding Testing Support Release 1
Analysis Design Coding Testing Support Release 2
Analysis Design Coding ….. Release m
Figure 9.9 | Incremental model.
9.4.2.2 Spiral Model Step 3: Risk analysis

Step 4: Engineering
Steps involved in spiral model (Fig. 9.10) are as follows:
Step 5: Construction and release
Step 1: Customer communication
Step 6: Customer evaluation
Step 2: Planning
Planning
Customer
communication Risk analysis
Entry point
Engineering
Customer
evaluation
Construction and release

Figure 9.10 | Spiral model.
1. In the spiral model, the entry point is the customer 5. In the spiral model, customer satisfaction is high,
communication to finalise the SRS. In the planning and at the same time the designer is able to design
stage, we can prepare the estimation model to esti- efficient design templates and deliver the code on
mate the cost schedule and performance. budget. So, this model is called win-win spiral
2. In the risk analysis stage, we can identify the tech- model.
nical and management risks. In the engineering
stage, we can select the effective process model to
develop the software. 9.4.2.3 Component-Based Development Model
3. In the construction and release stage, the proce
dural description is converted into machine- In the component-based model, the components are used
readable format and test cases also implemented. in the project development (Fig. 9.11).
Later, the software is delivered to the customer. After designing the project, before constructing the
4. In the customer evaluation stage, the software final software check the availability of the components
undergoes operation. During operation the cus- in the library. If the component is available, use it in
tomer identifies the change requirement and posts the current project; if not, construct the new compo-
it to the development team. The change is reflected nent and place it in the library, then move to the next
in the next release of the software. iteration.
Chapter 9.indd 460 4/9/2015 10:06:38 AM

9.5 MEASUREMENT OF METRICS 461
Planning
Customer
communication
Risk analysis
Customer Indentified
evaluation candidate
component
Engineering
Look up
Construction and release component
in library
Construct Put the Build the Extract
the nth component in component if component if
iteration the library not available available
Figure 9.11 | Component-based development (CBD) model.
9.5 MEASUREMENT OF METRICS 2. Project metrics: It describes the project char-
acteristics and execution. For example, effort cost
and schedule; effort in terms of man-months. Effort
Measurement is a quantitative indication which shows of a project is 3 man-months means one developer
the size or complexity of a process. works 3 months to develop the software.
3. Process metrics: It describes the characteristics
9.5.1 Metrics of the process model to develop and maintain the
software (Fig. 9.12). For example, calculates the
Metrics is a quantitative measure of an attribute of the defect rate and defect removal efficiency.
project or process. All the estimation models are empirical modules, that
Measurement is divided into two types: means, the models are developed based on the past expe-
1. Direct measurement: In the direct measure- rience without proof. Even though these models lead to
ment, the value is calculated based on direct success scenario of the software development.
approach. Example: Size of a project can be calcu-
lated based on the LOC (line of code).
2. Indirect measurement: Indirect measurement Size
SRS estimation Some other
means the value can be calculated based on the parameters
other parameters. Example: Size of a project can be
calculated by using the function point.
Effort
Software supports with four kind of P’s: estimation
1. People Schedule Cost
2. Product estimation estimation
3. Project
E = A + B * (ev)c
4. Process LOC (KLOC)
Size
On the basis of the four P’s, the software metrics are
classified into the following three types: A = Adding factor PP
B = Scaling factor Empirical
1. Product metrics: This metrics describes the constants
C = Exponential factor
characteristics of the product such as size, com-
plexity, performance and quality level. Figure 9.12 | Process metrics – an illustration.
Chapter 9.indd 461 4/9/2015 10:06:39 AM

9.5.2 Size-Oriented Metrics Table 9.1 | Function point estimation
The size of a project can be calculated using the follow- Attribute Value Weighing Factor
ing two ways: Simple Average Complex
1. Line of code (LOC)
2. Function point analysis (FPA) I/P files 3 4 5
O/P files 4 5 7
9.5.2.1 Line of Code
Enquiries 3 4 6
If we want to estimate the size of a project, there is a
need of taking the past project experience. Assume the
Files 7 10 15
LOC into three categories:
1. Optimistic LOC (Sopt) External
5 7 10
2. Most likely LOC (Sm) interfaces
3. Permissible LOC (Spess)
Count value = Values × Weighing factor
After identifying three levels of the source code, we can
FP = Count value × VAF
estimate the size of the project by using this following
formula: Apart from the functional requirements, some other
parameters are also involved in the project. So, we need
Sopt + 4Sm + Spess
S= to consider the effects of these factors on the project.
∑
6
VAF = 0.65 + 0.01 × Fi
P =1 to 14
Problem 9.1: Consider a three-dimensional (3D) geo- where VAF is value adjustment factor.
metric application. The range of LOC estimated is
Sopt = 5600, Sm = 7900 and Spess = 9600. What is the Note: On the basis of the knowledge of the previous
expected size? project, the weighing table is maintained as a simple/
average/complex case. Calculate the count value by pro-
Solution: viding the multiplication between attribute values with
5600 + 31600 + 9600 46800 the corresponding complexity levels. Default value of
S=
6
=
6
= 7800
∑ FFi == 42
∑ 42, if value adjustment complexity values
i
P =1 to 14
So, the estimated size = 7800 LOC = 7.8 KLOC P =1 to 14
are not given, use the default value.
Problem 9.2: Consider the software project with the

9.5.2.2 Function Point Analysis number of input and output, enquiries, files and inter-
faces as 30(4), 25(5), 20(4), 10(10), 5(7). Consider () is
Function point analysis (FPA) is an indirect measure- a weighing factor. Assume all complexity adjustment
ment to calculate the size of the project. Therefore, there factor are average = 3. Calculate the FP for the project.
is a need to consider the other parameters. Solution:
The function points are calculated based on the fol-
lowing five attributes: Values Weighing Total
Factor
1. Number of user input
2. Number of user output 40 4 160
3. Number of user enquiries 20 5 100
4. Number of files
25 4 100
5. Number of external interface
15 10 150
Every input is associated with the weighing factor. The
weighing factor is defined based on the experience of the 5 7 35
past project. Grand total = 545
If the project is simple, the corresponding values are Value of weighing factor = 545
maintained in a table (Table 9.1). If project is complex,
those values are also maintained in table. FP = 545 × [(0.65) + 0.01 × 42] = 583.15
Chapter 9.indd 462 4/9/2015 10:06:40 AM

Problem 9.3: Consider 3D geometry application. Problem 9.5: Software development is expected to
The size of the project can be measured in terms of involve eight parameters of efforts. Calculate the
FP with the following parameters: following using SEL and WF models.

USER (a) The number of lines of code that can be produced.
I/O 3 × 4 = 12 (b) The duration of the development
(c) The productivity in LOC per person-year
O/P 2 × 5 = 10 (d) The average manning (person)
Enquiries 2 × 4=8
Solution:
Files 1 × 10 = 10
By SEL method:
External interfaces 4 × 7 = 28
(a) E = 1.4 (KLOC)0.93
Assume the project complexity is average and value 8P − 4 = 1.4 (KLOC)0.93
adjustment complexity is 1. Calculate the FP.
(8 × 12) 96
Solution: (KLOC)0.93 = =
1.4 1. 4
FP = 68 × (0.65 + 0.01 × 14)
 96 1/ 0.93
68 × (0.65 + 0.14) KLOC =  
1.4 
= = 90.0414
= 68 × (0.79) So, LOC = 90041
= 53.73 (b) D = 4.1 (KLOC)0.36 = 4.6 × (90.0414)0.26
= 4.6 × 3.22 = 14.8 ∼ 15 months
90041
Problem 9.4: The effort of the above project is (c) Productivity = = 11255
8
5 person-months. What is the productivity? (d) Manning = Effort/Duration = (8 × 12)/15 person-
Solution: months = 6.4 persons
Given that 5 persons-month ⇒ 53.72
By WF method:
(a) E = 5.2 (KLOC)0.91
5 person-months means that the work done by 5 persons
in 1 month. 8P − 4 = 5.2 (KLOC)0.91
Person 5 → 53.72 (8 × 12) 96
(KLOC)0.91 = = = 18.46
Person 1 → 53.72/5 = 10.74 5. 2 5. 2
 96 1/ 0.91
Productivity = 10.74 KLOC =  
 5.2 
= 24.63
or we can say, 10.74 functional point developed by
1 person-month. So, LOC = 24630
(b) D = 4.6 (KLOC)0.26 = 4.6 × (24.63)0.26
= 4.6 × 2.3 = 10.6 ∼ 11 months
9.5.3 Effort and Schedule (Duration) Estimation 24630
(c) Productivity = = 3078.75
8
Effort and duration can be estimated by using different (d) Manning = Effort/Duration = (8 × 12)/11 person-
empirical model structure. months = 8.7 persons
1. According to the software engineering laboratory
(SEL) empirical model, the effort can be estimated
9.5.3.1 Constructive Cost Model
as follows:
E = 1.4 (KLOC)0.93 person-months (units) By using constructive cost model (COCOMO), we can
The duration can be calculated as calculate the effort and time, this model is also an empiri-
D = 4.6 (KLOC)0.26 months (units) cally derived model. Therefore, the structure of the effort
calculation is
2. According to Walstonald-Felio method (W-F) (IBM):
Effort = c1 × EAF ×(size)c2 (person-months)
The effort can be calculated as
E = 5.2 (KLOC)0.91 person-months Duration = c3 ×(Effort)c4 (months)
Duration can be calculated as where c1 and c3 are the empirically derived scaling fac-
D = 4.1 (KLOC)0.36 tors and c2 and c4 are empirically derived exponential
Chapter 9.indd 463 4/9/2015 10:06:41 AM

factors. Size indicates in terms of KLOC or functional

point. EAF (effort adjustment factor) (1 to 14 factors) S.No. Cost Driven Effect
rating (1—7) range.
1. Language experience 1.0
COCOMO model is applicable in three different kinds
2. Schedule constraints 1.0
of applications, as follows:
3. Database size 1.0
1. Organic mode (low complexity)
2. Semi-detached mode (average complexity) 4. Turnaround time 1.0
3. Embedded mode (high complexity) 5. Virtual machine exp. 1.0
On the basis of the type of project, the constants are 6. Virtual machine volatility 1.0
derived as follows:
7. Use of software tools 0.88
c1 c2 c3 c4 8. Modern programming 1.0
practices
Simple 2.4 1.05 2.5 0.38
9. Storage constraints 1.0
Average 3.0 1.12 2.5 0.35
10. Application experience 1.10
Complex 3.6 1.20 2.5 0.32
11. Timing constraints 1.0
12. Requirement reliability 1.15
13. Product complexity 1.15
Problem 9.6: A company needs to develop digi-
tal signal processing software for one of its newest 14. Team capability 1.0
invention. The software is expected to have 40000
LOC. The company needs to determine the efforts in
person-months needed to develop this software using Solution:
the basic COCOMO model. AF = Multiplication of all cost-driven effect
The multiplicative factor for this model is given (Multiple all) EAF = 1.28
as 2.8 for the software development on the embedded
system while exponential factor is given as 1.20. What E = c1 × EAF × (size)c2
is the estimated effort in person-months?
(a) 234.35 (b) 932.50 As c1= 2.0, so c2 = 1.2 = 2.8 × 1.28 (100)1.2 = 900
(c) 287.8 (d) 122.4 (person-months)
D = 2.5 × (Effort)0.32 = 2.5 × (900).32 = 22.04 months ∼
Solution: 23 months
Effort = 2.8 EAF × (40000)1.20 (EAF = by default =1)
= 2.8 × (40000)1.20 9.5.3.2 Defect Rate
= 2.8 × (40)1.20
It is a quality metric, defect is undiscovered error during
= 234.35 the development of the software. The defect rate is
defined as
Number of defect × Time
Defect rate =
Total or number of LOC
Problem 9.7: A company develops software for a
digital signal processing application. The software
size is expected to have 100000 LOC. LOC is some- 9.5.3.3 Defect Removal Efficiency
times called as DSE (delivery of source instructions).
The company needs to determine the effort in Defect removal efficiency (DRE) is also a quality metric.
person-months and time in months to develop the It is defined as
software using COCOMO model under embedded E
mode. The effort adjustment factor is calculated DRE% = ×100
E +D
based on the given table. What is the estimated effort
and time for the above project.
where E = error and d = defect.
Chapter 9.indd 464 4/9/2015 10:06:42 AM

9.5.3.4 Halstead Size-Oriented Metric The estimated level of program is defined as

2n 2
Halstead metric is used to count the number of lines L=
preset in the software. To calculate the size of the project N1 × N 2
in terms of LOC, we are supposed to count the number The difficulty is defined as
of lines in the software. But this value does not give
1
the correct estimation because comments and spaces and D=
non-executable lines are also counted. L
LOC estimation means only the count of executable Error is defined as
V
statements. To identify the correct executable statement E=
in the software, there is a need to divide the program in L
the form of lexemes. Estimated program length is calculated as
n1 × log2 n1 + n2 × log2 n2
Lexemes Token
For Keyword Problem 9.8: Consider the following program and
calculate the software matrix n, N, V, D, E
= Equal -op
var a, b, c, d, m: integer’
+ Add -op
Begin
* Mul -op
Read in (a, b, c, d)
3 Constant
a: = a + a;
After dividing the program into token, we can separate b:= b + b;
the operator and operands. Operators are special words c: = c * d;
and special symbol operators, etc. Operands are constant,
m: = a + b – c;
variable, identifier, etc.
write in (m);
Unique Occur- Unique Occur- End
Operator rence Operand rence Solution:
of the of the
Operator Operand Unique Occur- Unique Occur-
in the in the Operator rence Operator rence
Software Software of the of the
Σ = n1 Σ = N1 Σ = n2 Σ = N2 Opera- Operand
tor in the in the
Software Software
where n1 = the count of unique operator used in the
software = 4 a 6
n2 = count of unique operands used in the software + 3 b 6
N1 = c ount of occurrence of the operator in the * 1 c 5
software
- 1 d 3
N2 = c ount of occurrence of the operands in the
, 7
software
: 1
The following metrics are defined based on the above
database. ; 7
The vocabulary can be calculated by using n = n1 + n2 () 2
The size of the software can be calculated by . 1
N = N1 + N2 Read in 1
When we are considering machine language then Write in 1
N = 2 × LOC Integer 1
begin-end
The volume of the program can be calculated by using
n1 =13 N1 = 31 n2 = 5 N2 = 23
V = N log2 n
Chapter 9.indd 465 4/9/2015 10:06:43 AM

n = n1 + n2 = 13 + 5 = 18
9.6.1.2 Types of Risk
N = N1 + N2 = 31 + 23 = 54 1. Project risk: This risk threatens the project plan,
1 therefore schedule will be extended and develop-
D= ment cost will increase.
L
2. Technical risk: These risks threaten the quality
2n2 2×5 1 and timeliness of developing the software. When
L= = = = 30.30
n1 × N 2 13 × 23 0.033 this risk becomes real, we cannot develop the proj-
ect within the quality levels.
V = n⋅log2n = 54⋅log218 = 54 × 5 = 270
3. Business risk: These risks threaten the viability
of software risk. Under this category, four different
V 270
E= = = 8181.81 risks are defined:
L 0.033 •• Market risk: Build an excellent product but no users.
•• Strategic risk: Build a product that is no longer
fit into the business objective.
9.6 RISK ANALYSIS •• Management risk: Lack of high-level support to
inject the changes.
•• Budget risk: No synchronisation between cost
Risk means extreme condition when the product becomes and price.
a failure. Risk is associated with two characteristics: 4. Known risk: Uncover errors after a careful evalu-
ation of the project.
1. Uncertainty: It means may/may not occur, 100%
5. Predictable risk: It is extrapolation from the
probable risk.
past project experience.
2. Loss: It means when the risk becomes real there is
6. Unpredictable risk: It acts as a joker in the deck.
a loss, otherwise no loss.
7. Generic risk: It describes the characteristics of the
project such as planning, performance, cost, effort, etc.
9.6.1 Risk Identification 8. Product-specific risk: These risks describe the
characteristic of the product such as the size, qual-
After identifying the extreme condition, there is a need ity, performance, etc.
of analyses of the risk in terms of level of uncertainty
and degree of loss. The level of uncertainty is maintained Risk Components
in the following four ways: There are four different components present in the soft-
1. Negligible ware to maintain the probability of probable risk:
2. Marginal 1. Performance: It indicates the level of uncertainty
3. Critical to meet its specification.
4. Catastrophic 2. Cost: It indicates the level of uncertainty to main-
tain project budget.
We are going to identify critical and catastrophic risk
3. Support: It indicates the level of uncertainty to
condition and try to manage them in the software execu-
easily correct, adopt or enhance the existing software.
tion with the help of risk strategy.
4. Schedule: It indicates the level of uncertainty to
maintain the schedule and quality of the software.
9.6.1.1 Risk Strategy
9.6.2 Risk Projection or Risk Estimation
There are two types of risk strategies:
1. Reactive risk strategy: It means monitoring To estimate the cost of the risk, there is a need to iden-
the project for the probable risk. When the risk is tify the probability of the risk becoming real and the
present during execution of the project, it imple- associated impact.
ments the plans to control the risks. It is also called
as fire-lighting mode. Risk Name Category Probability Impact
2. Proactive risk strategy: It means risk plans and
actions are developed before the implementation of
the software. So, by using these strategies we can By using the above database we can estimate the risk
avoid the risk, if possible, or manage the risk, if not cost by using the following formula:
possible. By combining both proactive and reac- Risk exposure (RE) = P × C
tive strategies, the risk mitigation, monitoring and where P is the probability of risk becoming real and C
management (RMMM) plan is prepared. is the cost.
Chapter 9.indd 466 4/9/2015 10:06:44 AM

9.7 SOFTWARE DEVELOPMENT LIFE CYCLE 467
Problem 9.9: To develop an image-processing 9.7 SOFTWARE DEVELOPMENT

application, the software uses component-based LIFE CYCLE
development approach `60’. Reusable software com-
ponent are planned but 70% of the project reusable
component are used. `18’ components are developed The software development life cycle (SDLC) describes
from the baseline to complete the project. The cost the generic approach to develop the software. The frame-
component size is `100’ LOC, and the cost for each work of the SDLC is shown in Fig. 9.13.
LOC is $14. What is the risk RE when the risk prob-
ability is 80%? 9.7.1 Requirement
Solution: According to the IEEE standard, the requirement is
RE = P × C defined as a condition or capability needed by a user to
solve a problem or to achieve an objective.
P = 80%
There are five types of requirements classified in the
C = 18 × 100 × 14 = $252000
software engineering (Fig. 9.14):
RE = 0.8 × 25200 1. Business requirement: These requirements describe
= $20160 the high-level business requirements. These require-
ments are documented in vision and scope document.
Requirements Inception Requirement Engineering
Requirement Requirement
development management
process process
SRS Elicitation Change-
Analysis control
Specification Version
control
Validation Req tracing
Validation
Data level testing
Design Elaboration
Architecture level
Interface level Integration testing
Component level Top down
Bottom up
Regresser
Smoke
Code Construction Logical component Unit testing
Logical component
Functional
Behavioural
Test Transition Development test Black box
cases Grey box
Error free
α - testing — Graph based
β - testing logical component — Boundary value analysis
Corrective — Equivallence partioning
Maintenance Adaptive Structural
Perfective White box
Preventive Glass box
— Basis path
Recovery testing
System Security testing — Control structure
testing Stress testing
Performance testing
Figure 9.13 | The software development life cycle.
Chapter 9.indd 467 4/9/2015 10:06:44 AM

Functional requirement Non-functional requirement

Business
requirement
Vision scope document Business

rules
User Quality
requirement attribute
Use - case document External

interface
System Functional
requirement requirement
Constraints
SRS
Figure 9.14 | Requirement phase of SDLC.

2. User requirement: This requirement describes In order to develop the right system, the right require-
the user need, user goal and user capability towards ments need to be injected into the software development
the system to achieve their objective. Those require- process, and for developing the right requirement there
ments are documented in use-case document. is a need of requirement engineering.
3. Functional requirements: This requirement
describes the functionality of the software, which 9.7.1.2 Requirement Engineering
is developed into the system to satisfy the user
requirement. These requirements are documented Requirement engineering includes the systematic, dis-
in the SRS document. ciplined, quantifiable approach to the development,
4. System requirement: This requirement operation and maintenance of the requirement. The
describes the subsystem requirement on which following two processes take place in requirement
platform the developed product is running, that engineering:
is, CPU, memory capacity and operating system. 1. Requirement management process: These
These are the subset of the functional requirement. activities are used to manage the requirement
5. Non-functional requirement: This requirement throughout the project development and lifetime
describes the business rules (such as govt. acts, of the product.
govt. policies, account plans) and quality attributes 2. Requirement development process: The
(availability, usability, maintainability, robust- framework used to develop the requirement is
ness, flexibility). shown in Fig. 9.15.
9.7.1.1 External Interfaces On the basis of the goal of the software, elicit the
requirements from different sources and create the data
Like how many other systems are communicating with dictionary.
the software and constraints (risk condition). All these On the basis of the data dictionary, create the struc-
requirements are documented in the SRS. ture of the software. On the basis of the structure of the
Note: SRS document consists of goals of the software, software, prepare the technical specifications. All the cus-
functional requirement and non-functional requirements. tomer capabilities are maintained in the SRS document.
Validate
Elicitation Data Analysis Specification Validation SRS
dictionary
Data model
Functional model Control specification
Dynamic model Process specification
Figure 9.15 | Requirement process.
Chapter 9.indd 468 4/9/2015 10:06:44 AM

After reviewing the SRS document twice or thrice, diagram. E-R diagram consists of object, attribute
the customer and the developer sign on the SRS docu- and relationship.
ment. This document (SRS) is then forwarded to the Object is defined as a representation of com-
design face. posite elements. Composite elements must have
multiple properties. The property of the object
9.7.1.3 Elicitation is called as attribute. The relationship can be
maintained between the object by using two
In this state, by using different techniques, elicit the concepts.
requirement from different sources towards goal of •• Cardinality: It indicates in how many ways
the project. The different elicitation techniques are as one object can be communicated with another
follows: object. There are three types of cardinality:
1: 1
1. CORE (Controlled Requirement Expres
sion): In CORE technique, elicit the requirement 1: n
by using view point analysis. In this analysis, M: N
•• Modelling: It indicates “is the communication
maintain the sequence of data object, action and
consequence.
link between the object mandatory or optional”.
2. IBIS (Issue-based Information System): In
IBIS technique, elicit the requirement based on the 1 = Mandatory
question and answer analysis. This technique main-
tains the issue, position and justification. Modality
3. FODA (Future-oriented Domain Analy
sis): In FODA technique, elicit the requirement
0 = Optional
by creating different models to specify the struc-
ture of the end product. This technique maintains
the sequence context model – function model – Note: On the basis of the data model, the data
dynamic model (behaviour or data model). structure is created in the data level design.
4. QFD (Quality Function Deployment): In 2. Functional and information flow model:
QFD technique, elicit the requirement towards The functional and information flow model
quality deployment. describes the functionality of the end product, that
5. JAD (Joint Application Development): In means, what is the behaviour of the end product
JAD technique, elicit the requirement by conduct- with a proper set of input. The functionality can be
ing a quick meeting between customer, end user, represented by using the data flow diagram (DFD).
analyst and developers. Data flow diagram is also called as Bubble chart
6. Prototype: In prototype technique, elicit the or directed graph. Different levels of abstraction of
requirement by creating the quick design mock-ups the end product can be represented using different
and taking continuous feedback from the customer. levels of DFD.
•• Level-Zero DFD: It describes the high-level
Note: The output of the elicitation stage is data abstraction of the functionality by maintaining
dictionary, it consists of data objects related to the the single process bubble with multiple input-
software. and output-directed edges (Fig. 9.16).
9.7.1.4 Analysis
Input Slow Output
This stage describes the structure of the end product process
in terms of data object, functionality, information
flow and behaviour with respect to external event, by
creating three types of model: Figure 9.16 | Level-zero DFD.
1. Context or data model (E-R diagram)
2. Functionality and information flow model In this level of DFD, only the functionality is
(described by DFD) represented with absence of information flow.
3. Dynamic or behavioural model (described by STD) •• Level-1 DFD: It describes the functionality
1. Context or data model (E-R Diagram): This and information flow of the software by main-
model describes the structure of the end product in taining different process bubbles. It describes
terms of data object and relationship. Data model the project in low-level abstraction. Low-level
can be represented by using the entity, relationship abstraction is easy to code (Fig. 9.17).
Chapter 9.indd 469 4/9/2015 10:06:45 AM

9.7.2 Design
1 2
The design stage represents the project in the low-level
5 abstraction, that means, the customer description is
4 translated into technical description. To do this process,
3 the following four levels of design take place (Fig. 9.20):
1. Data level: Data level design describes the data
Figure 9.17 | Level-1 DFD. structure. It takes help from the E-R diagram and
data dictionary.
3. Dynamic model: It describes the behaviour of 2. Architecture level: It defines the structure of
the system with respect to external events. This the end product (skeleton). It takes help from the
model is represented using the state transition dia- DFD of the last stage.
gram (Fig. 9.18). 3. Interface level: Interface level design describes
the number of interfaces required in the system. It
Transition takes help from the state transition diagram.
4. Component level: It represents the procedural
State 1 State 2 description to implement the functionality in the
Figure 9.18 | State transition diagram.
software. It can take help from process specification
and control specification (Fig. 9.21).
•• Specification: In this stage, the models are
refined into control specification and process
specification
(a) Control specification: It shows the 4 Component
behaviour of the product. level
(b) Process specification: It refines the
mathematical expression, algorithm and 3 Interface level
constraints.
•• Validation: In this stage, various validation 2 Architecture level
criteria are implemented to validate the require-
ment towards the right condition. The following 1 Data level
reviews are done to validate the software:
Figure 9.20 | Levels.
(a) Configuration review
(b) Quick review
(c) Detailed review
“Validation means the black-box testing is per- Process specification
formed at the requirement stage itself”. 4
After validating the SRS document, it is forwarded
to the design stage. The output of the requirement
development process is shown in Fig. 9.19.
1 2
Process specification
Data
dictionary DFD
3
State transition
diagram
Data
dictionary 4
DFD
State transition
diagram Figure 9.21 | Component.
9.7.2.1 Designing Concept
The following concepts should be kept in mind while

Figure 9.19 | Validation. designing a document:
Chapter 9.indd 470 4/9/2015 10:06:46 AM

High
Integration cost
Cost
Software cost
Cost module
Low
Low High
Number of Modules
Figure 9.22 | Number of modules versus cost of software development.
1. Abstraction: There are different levels of abstrac- 1. Logical cohesion: Logical cohesion exists when
tion available to represent the statement. In high- a module that performs tasks are related logically.
level abstraction, the statement is represented in 2. Temporal cohesion: Temporal cohesion exists
concrete manner, that means, internal structure when a module contains tasks that are related in
is not provided. In low-level abstraction, all state- such a way that all the tasks must be executed
ments are defined in a single format with all details within the same time limit.
of implementation. 3. Procedure cohesion: The procedural cohesion
2. Refinement: It is a top-down strategy to sim- exists when the processing elements of a module
plify the statement structure. Refinement is also are related and must be executed in a specific
called as elaboration. After refinement of the order.
statement, we can get the direct implementation 4. Communication cohesion: The communication
statement. cohesion exists when all the processing elements
3. Modularity: Software development process uses concentrate on one area of a data structure.
the modularity concept to reduce the develop-
ment cost and time. Suppose the complexity of the
problem is denoted by C(P) and the effort of the
9.7.2.3 Coupling
problem is denoted by E(P). The following formats
Coupling is a measure of relative independence among
represent the advantage of C.
modules, that is, it is a measure of interconnection
C(P1) E(P1) among modules. Following types of coupling exist:
C(P2) E(P2) 1. Data coupling: Data coupling exists when a
C(P1) > C(P2) module accesses another module through an argu-
C(P1) > E(P2) ment or through a variable.
C(P1 + P2) > C(P1) + C(P2) 2. Stamp coupling: Stamp coupling exists when a
E(P1 + P2) > E(P1) + E(P2) module accesses another module through a data
structure.
When the number of modules is less, the cost
3. Control coupling: Control coupling exists when
module for developing the module is high, and at
the control is passed between modules using a con-
the same time the integration cost is low.
trol variable.
When the number of modules increases, the cost 4. External coupling: External coupling exists
per module is low but integration cost is high. This when modules are connected to an environment
relationship is shown in Fig. 9.22. external to the software.
Modules should be designed in such a way that each 5. Common coupling: Common coupling exists
module has specific functional requirements. Functional when the modules share the same global variable.
independence is measured using two terms cohesion and 6. Content coupling: Content coupling exists when one
coupling. module makes use of data or control information main-
tained in another module. Also, content coupling occurs
9.7.2.2 Cohesion when branches are made into the middle of a module.
It is a measure of relative functional strength of a module. Note: Effective modular design must have high cohesion
Following types of cohesion exist: and low coupling.
Chapter 9.indd 471 4/9/2015 10:06:46 AM

9.7.2.4 Software Architecture There are different advantages with horizontal partitioning:

1. Easy to develop
It shows the hierarchical structure of the software and 2. Easy to integrate
also describes the logical connections between the mod- 3. Easy to test
ules. In this process, it identifies the data model, func- 4. Easy to maintain
tional model and behaviour model.
Data Structure
Control Hierarchy
It shows the logical connection between the data objects.
It indicates the control sequence in the software project.
It represents the interaction between different modules
(Fig. 9.23). Software Procedure
It describes the customer requirements into programming
M1 language statements. Information hiding is used in effec-
tive modularity. It shows the independence of the module.
After the designing stage, we have the software design
M2 M3 document (SDD). The SDD represents the programming
description language oriented customer objective.
M4 M5 M6
9.7.3 Coding
M8 M9
M7 It translates the procedural-oriented description into
M10 machine-readable format by using appropriate program-
ming language.
Fan-in ⇒ number of control in Fan-in (M7) = 3
Fan-out ⇒ number of control out (M2) = 1 9.7.4 Testing
Fan-out (M7) = 0
The objective of testing to measure to what extent the
(M2) = 3 developer logical program becomes a failure. In the test-
Figure 9.23 | An illustration of control hierarchy. ing phase, different test cases are developed and imple-
mented to cover the structural and functional errors.
The control hierarchy defines various parameters of the Functional testing includes conducting tests to check the
software design: functionality of the product with all proper sets of input,
1. Depth: It is a measure of the number of levels which also covers errors. Structural testing includes con-
present in the software design. ducting tests to examine the internal structure of the
2. Width: It is a measure of the maximum number developed product to cover the error.
of modules present at any one of the level.
3. Fan-in: It is a measure of the number of modules
already controlling the given module.
9.7.4.1 Structural Testing
4. Fan-out: It is a measure of the number of modules
This testing is also called as white-box or glass-box test-
controlled by the given module.
ing. The objective of white-box testing is as follows:
5. Structural partitioning: It indicates the control
sequence of development, maintenance, and inte- 1. Guarantee the all the statements present in the
gration of the software. It is divided into the fol- developed program exercise at least once.
lowing two types: 2. Exercise logical expression towards true and false side.
•• Horizontal 3. Exercise all the loop boundaries within the opera-
•• Vertical tional limit and beyond the operational limit.
4. Exercise the data structure present in the program
Horizontal partitioning allows the level by level development towards validity.
and integration, that means, control flow is assigned to levels.
To reach the objectives of white-box testing, two differ-
Vertical portioning means the control sequence is for-
ent test operations are performed:
warded from root to leaf node, that is, depth-wise develop-
ment and integration take place. Because of dependencies, 1. Basis path testing
this approach is not suggested to software development. 2. Central structural testing
Chapter 9.indd 472 4/9/2015 10:06:46 AM

Basis Path Testing By using the cyclomatic complexity we can measure the
logical complexity of the program.
This test is used to cover all the paths present in the
development code. This testing guarantees that all the
statements in the program must be exercised at least Cyclomatic Complexity
once to identify the path in the program, convert the
program into graphical notation. On the basis of the It is defined as the quantitative measure of logical com-
control flow between the statements, the program is con- plexity of the program. Cyclomatic complexity can be
verted into graph notation called as flow graph or pro- calculated in three ways:
gram graph. Flow graph is a directed graph. 1. By using the number of edges and number of nodes
Each node in the flow graph represents the program we can measure the cyclomatic complexity as
statement. The sequence of the statement is represented V(G) = E − N + 2P
by using the following (Fig. 9.24):
E = Number of edges
1 2 N = Number of nodes
Figure 9.24 | Sequence statement.

P = Number of disconnected component
1. The if condition is represented by Fig. 9.25. 2. By using the region we can estimate the cyclomatic
complexity as
2 V(G) = Number of regions present in the graph
1 4 When we count the cyclomatic complexity of a pro-

gram by using basis path testing, outside of graph
3 is considered as one region.
Figure 9.25 | IF condition.

3. By using predictable nodes:
V(G) = P + 1
2. The while statement is represented by Fig. 9.26.
P = Number of predicate nodes
Predictable nodes are calculated by two or more
nodes expanding from it.
1 2
3
Problem 9.10: Consider the following flow graph and
calculate the cyclomatic complexity and also identify
Figure 9.26 | WHILE statement. the independent path.
3. Do —while statement is represented by using Fig. 9.27.

1
P1
1 2 3 R4
2,3 P2
Figure 9.27 | DO—WHILE statement. 6 P3 4.5
4. The switch case statement is represented by Fig. 9.28.
2 7 R1 8 R2 R3
1 3 6
9
4 10
5 11
Figure 9.28 | SWITCH-CASE statement. Flow graph depicting relation among different nodes.
Chapter 9.indd 473 4/9/2015 10:06:47 AM

Solution: Problem 9.12: Consider the following graph and

Cyclomatic complexity = E − N + 2 calculate the cyclomatic complexity.
= 11 − 9 + 2 = 4
1
Number of regions = 4 = R1, R2, R3, R4
So, cyclomatic complexity = 4
2
Cyclomatic complexity P + 1 = 3 + 1 = 4
R4
Path 1: 1-11 10 3
Path 2: 1-2-3-4-5-10-1-11
R5
12 11
Path 3: 1-2-3-6-8-9-10-1-11 4
13
Path 4: 1-2-3-6-7-9-10-1-11
5
R1
R2
Problem 9.11: Consider the following program and 6
calculate the cyclomatic complexity.
R3
8 7
1
9
Flow graph depicting relation among nodes.

2
Cyclomatic complexity = 6
3
4
Problem 9.13: Consider the following flow chart and
calculate the cyclomatic complexity.
5
6
10
7 1
8
9 11 2
Flow graph depicting relation among nodes. 3
Solution: 4
6
Rule 3: P + 1 = 4 + 1 = 5
Rule 2: Number of regions = 5 7 8 5
Rule 1 = E − n + 2 = 13 − 10 + 2 = 5 9
Path 1: 1-2-45-6-7-8-10 10
Path 2: 1-2-4-5-7-8-10
Path 3: 1-2-3-5-7-8-10 and more Flow chart depicting various operations of a program.
Chapter 9.indd 474 4/9/2015 10:06:48 AM

Cyclomatic complexity = N
umber of if statement
+1=3+1=4 1
The cyclomatic complexity of the program can be
calculated by using the graph matrices. a
P=4
3
Graph Matrices e
b
The flow graph is converted into graph matrices by
maintaining rows and columns. The number of rows and f
columns depends on the number of nodes in the flow 4 d
5
graph. Based on the communication link between the
nodes in the flow graph, the entry is reflected in corre- c
sponding position in the graph matrices. g
If any row consists of more than one entry that node 2
is called as predicate node, then take the weightage of
predicate node by subtracting one from the total. After
subtraction operation, add all the row values, it gives the Link
predicate nodes. Use the formula to calculate cyclomatic Mode 1 2 3 4 5
complexity V(G) = P + 1. 1 a
2
Example 9.1
3 d b
1 2 3 4 5 Connection 1 2 3 4 5
4 c f
1 b c 1 Matrix 1 0 1 0 1 1
g e h
5
2 a d 2 1 0 1 0 0
3 e f 3 0 1 0 1 0 A flow graph and graph matrix. Solution:
4 g 4 0 0 1 0 0 So, cyclomatic complexity = P + 1 = 5
5 h 5 1 0 0 0 0
Control Structure Testing
3−1=2 1 2 3 4 5 This testing is used to cover the error from the control
structure present in the program. There are three types
2−1=1 1 1 1 1 of expressions:
2−1=1 2 1 1 1. Arithmetic (= , − , *, /) = value
2. Relational (>, <, <=, >=, =, ≠) = True/False
1−1=0 3 1 1
(Boolean)
1−1=0 4 1 3. Boolean (&&, , ¬, /) = True/False (Boolean data type)
P = 04 5 1 Examples:
A graph matrix and its allocation. 1. a + b
2. (a + b) > (c + d) = T/F
3. ((a + b) > (c + d) && ((a × b) < ((c * d)) = T/F
After evaluating the control expression, the output may
Problem 9.14: Consider the following flow graph and be a Boolean data, that is, true/false.
create the graph matrix and calculate the cyclomatic In the control structure testing, examine all the logi-
complexity. cal expression towards true/false. During that evalua-
Note: If there are more than two outgoing nodes, tion, it covers the Boolean operator error, Boolean data
then the predicate formula will not work. or parenthesis symbol error, relational operator error
and arithmetic operator errors.
Chapter 9.indd 475 4/9/2015 10:06:48 AM

In this testing, two test cases are developed to cover 9.7.4.3 Comparison Testing
all the errors described in the above list.
1. Branch testing When the software is developed in the real-time applica-
2. Domain testing tion domain, there is a predefined value estimated based
on the product behaviour. After developing the soft-
ware, compare the behaviour of the new product with
9.7.4.2 Black-Box Testing
predefined values and remove the errors.
In this testing, various input test cases are developed
and the product functionality is examined, whether it 9.7.4.4 Life Cycle Testing
satisfies all possible inputs. This testing is also called
behavioural testing or grey-box testing. In this strategy, Various testing plans are implemented at each stage of
various testing techniques are used to cover the func- the software development life cycle.
tional errors such as
1. Equivalence partition
9.7.4.5 Validation Testing
2. Boundary value analysis
This testing is performed at requirement stage by con-
3. Comparison
ducting the black-box test plans on a prototype model.
1. Equivalence partition: In this testing, the input
domain is divided into equivalent partitions. Each 9.7.4.6 Integration Testing
position is called as class. One partition is the exact
valid range and another partition is above the This testing is performed at the design stage. Software
upper limit and another is below the lower limit. development uses the modularity concept and develops
the module independently. After developing the module,
Example 9.2 there is a need of integration to generate the final system.
In that regard, integration testing is required to check
If the input box consists of 1 to 1000 values, no need
the functionality. Integration testing is not focused on
of preparing the thousand (1000) test cases. By using
internal details of the modules.
equivalence partition, the above input box is divided
into three test cases. 1. Top-down integration: This testing focuses on
Test case is prepared between the ranges of {1—1000}
the function of the module integration either dep-
thwise or breadthwise. Depth-wise integration is a
Test case is prepared {>1000} vertical partitioning and widthwise integration is
Test case is prepared {<1} horizontal partitioning.
2. Bottom-up integration: In this testing, the
integration starts from leaf node and forwards to
2. Boundary value analysis: Most of the problems the root node. In this testing, the stubs and drivers
are possible at the extreme edges that are boundaries are maintained.
of input domain. Boundary value analysis focuses on
identified errors at the boundary rather than finding
9.7.4.7 Regression Testing
those existing at the centre of the input domain.
When a new module is added, there is a possibility of intro-
Example 9.3 ducing new functionality to the existing system. Due to the
new functionality, there is a possibility of errors occurring
Test case for input box accepting number (1–1000) in the existing system. To cover those errors, all test cases
using boundary value analysis is are re-conducted. This is called regression testing.
(a) Test cases with data exactly the input boundary of
input domain, that is, 1 and 1000. 9.7.4.8 Smoke Testing
(b) Test data with values at just below the extreme
edges of input domain, that is, 0 and 999. This testing is used in the wrap-shrink component devel-
(c) Test data with values just above the input boundary opment. Wrap-shrink component development means
of input domain, that is, 2 and 1000. within the short period of time the software is imple-
Note: When the product supports the N variable that mented. To perform this we use modularity and by
is tested using the boundary value analysis, that is, adding or deleting the existing module we can develop
4N + 1 test cases. the final product. In this way, smoke testing is used to
cover the error.
Chapter 9.indd 476 4/9/2015 10:06:49 AM

SOLVED EXAMPLES 477
9.7.4.9 Unit Testing 9.7.4.12 System Testing
This testing is performed at the coding stage. After After developing the product, various system-level test-
developing the module, use the white-box test plan to ing operations are performed to check the reliability of
cover the structural errors in the module. the product.
9.7.4.10 Alpha Testing 1. Recovery testing: This testing focuses on

the recovery of data due to crashes or power
After developing the product, this test is conducted by failure.
the customer at the developer side to cover the error. In 2. Security testing: This testing focuses on unau-
this testing, the developer’s presence is there. thorised accessing of product.
3. Stress testing: This testing is focused on load
9.7.4.11 Beta Testing
balancing. It is also known as volume testing.
After deploying the product to the customer side, this test 4. Performance testing: This testing is a user
is conducted by the customer or end user at the customer acceptance testing in terms of quality and
place. In this testing, the developer’s presence is not there. reliability.
IMPORTANT FORMULAS
Sopt + 4Sm + Spess

1. Size S = 5. COCOMO model
6
Effort = c1 × EAF ×(size)c2 (person-months)
2. FP = Count value × VAF
Duration = c3 ×(Effort)c4 (months)
where VAF = 0.65 + 0.01 × ∑ Fi
P =1 to 14 6. RE = P × C
3. According to the SEL empirical model, the effort where P is the probability of risk becoming real
and duration can be estimated as and C is the cost.
E = 1.4 (KLOC)0.93 person-month (units) 7. Cyclomatic complexity
D = 4.6 (KLOC) 0.26
months (units) (G) = E − N + 2P, where E = number of edges,
V
N = number of nodes and P = number of discon-
4. According to Walstonald-Felio method (W-F)
nected components.
(IBM):
V(G) = Number of regions present in the graph
E = 5.2 (KLOC)0.91person-months
D = 4.1 (KLOC)0.36 V(G) = P + 1, where P = number of predicate nodes
SOLVED EXAMPLES
1. The most important feature of spiral model is Solution: Cyclomatic complexity is a software metric
developed by Thomas J. McCabe. It is used to indi-
(a) Requirement analysis
cate the complexity of a program. It directly measures
(b) Risk management
the number of linearly independent paths through a
(c) Quality management
program’s source code. It is also known as conditional
complexity. It is defined as V (G) = e − n + 2.
(d) Configuration management
Solution: Risk management helps in developing a Ans. (b)
cost effective project where risks are analyzed and 3. Testing of software with actual data and in actual
risk strategies are decided upon. environment is called
Ans. (b)
(a) Alpha testing (b) Beta testing
2. Cyclomatic complexity of a flow graph G with n
(c) Regression testing (d) None of the above
vertices and e edges is
Solution: Users or an independent test team at the
(a) V(G) = e + n − 2 (b) V(G) = e − n + 2
developer’s site performs alpha testing. Alpha testing
(c) V(G) = e + n + 2 (d) V(G) = e − n − 2
Chapter 9.indd 477 4/9/2015 10:06:50 AM

is often employed as a form of internal acceptance 8. Superficially the term “object-oriented”, means
testing, before the software goes to beta testing. that, we organise software as a
Beta testing is performed after alpha testing for (a) c ollection of continuous objects that incorpo-
user acceptance testing. rates both data structure and behaviour.
Regression testing focuses on finding defects (b) collection of discrete objects that incorporates
after a major code change has occurred. both discrete structure and behaviour.
(c) collection of discrete objects that incorporates
Ans. (b)
both data structure and behaviour.
4. Object Request Broker (ORB) is (d) collection of objects that incorporates both dis-
crete data structure and behaviour.
I. A software program that runs on the client as
well as on the application server. Solution: This is the definition of object-oriented
II. A software program that runs on the client side according to software engineering.
only. Ans. (c)
III. A software program that runs on the application 9. Which one of the following is not a key process
server, where most of the components reside. area in CMM level 5?
(a) I, II and III (b) I and II (a) Defect prevention
(c) II and III (d) I only (b) Process change management
(c) Software product engineering
Solution: ORB manages interaction between clients
(d) Technology change management
and servers.
Ans. (d) Solution: The key process areas at level 5 covers
the following issues–Defect Prevention, Technology
5. Key process areas of CMM level 4 are also classi- Change Management and Process Change
fied by a process which is Management. Software Product Engineering has
(a) CMM level 2 (b) CMM level 3 been addressed in level 3.
(c) CMM level 5 (d) All of the above Ans. (c)
Solution: There are five maturity levels in CMM 10. Which of the following system components is
model–(1) Initial, (2) Managed, (3) Defined, (4) responsible for ensuring that the system is working
Quantitatively Managed and (5) Optimizing. The to fulfil its objective?
organization at level 2 includes level 1. So, level 4 (a) Input (b) Output
is included in level 5. (c) Feedback (d) Control
Ans. (c)
Solution: The components of the control system
6. Validation means ensure smooth functioning of the system to achieve
its objective.
(a) Are we building the product right Ans. (d)
(b) Are we building the right product
(c) Certification of fields 11. Modifying the software to match changes in the
(d) None of the above ever-changing environment is called
Solution: Verification means are we building the (a) Adaptive maintenance

product right, whereas validation means whether (b) Corrective maintenance
we are building the right product. (c) Perfective maintenance
Ans. (b) (d) Preventive maintenance
Solution: Corrective maintenance refers to modi-
7. Function points can be calculated by fications initiated by defects in the software.
(a) UFP × CAF (b) UFP × FAC Perfective maintenance refers to improving pro-
(c) UFP × Cost (d) UFP × Productivity cessing efficiency or performance, or restructuring
Solution: Function Point FP = UFP × CAF. UFP the software to improve changeability.
stands for Unadjusted Function Point, while CAF Preventive maintenance refers to prevention of
stands for Complexity Adjustment Factor. breakdowns.
Ans. (a) Ans. (a)
Chapter 9.indd 478 4/9/2015 10:06:50 AM

12. Consider the following flow graph and calculate 14. Which of the following metric does not depend on
the cyclomatic complexity and also identify the the programming language used?
independent path.
(a) Line of code (b) Function count
(c) Member of token (d) All of the above
1
Solution: Line of code and tokens depends upon
the programming language used for developing
2 programs.
Ans. (b)
3 4
15. If in a software project the number of user input,
user output, enquiries, files and external interfaces
are (15, 50, 24, 12, 8), respectively, with complex-
5 6 7 8 ity average weighing factor. Find productivity if
effort = 70 person-month.
9 10 Solution:
15 × 4 = 60
50 × 5 = 50
11 24 × 4 = 96
12 × 10 = 120
8 × 7 = 56
Solution: Cyclomatic complexity = E − N + 2
= 12 − 11 + 2 = 3 Count value = 382
OR Productivity = Count value × (0.65 + 42 × 0.01)
Number of regions = 3 = 382 × (0.65 + 0.42)
So, cyclomatic complexity = 3 = 382 × (1.07)
Ans. (3)
= 408.74
13. Software measurement is useful to Ans. (408.74)
(a) indicate quality of the product 16. 58000 LOC gaming software is developed, with the
(b) track progress effort of 3 person-year. What is the productivity of
(c) form a baseline for estimation and prediction person-month?
(d) all of the above.
(a) 1.9 KLOC (b) 1.6 KLOC
Solution: Software measurement is useful to indicate (c) 4.8 KLOC (d) 4.2 KLOC
quality of the product, track progress, assess productiv-
ity and form a baseline for estimation and prediction. Solution: 58000/36 = 1611.1 = 1.6 KLOC
Ans. (d) Ans. (b)
1. The coupling between different modules of a soft- (a) I-II-III-IV-V

ware is categorized as follows: (b) V-IV-III-II-I
(c) I-III-V-II-IV
I. Content coupling
(d) IV-II-V-III-I
II. Common coupling
III. Control coupling (GATE 2009: 1 Mark)
IV. Stamp coupling Solution: The types of coupling from least desir-
V. Data coupling able to most desirable are as follows:
Coupling between modules can be ranked in the Content coupling → Common coupling → Control
order of strongest (least desirable) to weakest coupling → Stamp coupling → Data coupling.
(most desirable) as follows: Ans. (a)
Chapter 9.indd 479 4/9/2015 10:06:50 AM

2. Which of the following statements are TRUE? (a) 19 (b) 21

(c) 20 (d) 10
I. The context diagram should depict the system
as a single bubble. (GATE 2010: 1 Mark)
II. External entities should be identified clearly at
Solution: Cyclomatic complexity = Number of
all levels of DFDs.
decision points + 1
III. Control information should not be represented
in a DFD. Number of decision points in A = 10 − 1 = 9
IV. A data store can be connected to either another Number of decision points in B = 10 − 1 = 9
data store or an external entity. Cyclomatic complexity is = (9 + 9) + 1 = 19
(a) II and III (b) II and III Ans. (a)
(c) I and III (d) I, II and III 5. What is the appropriate pairing of items in the two
(GATE 2009: 2 Marks) columns listing various activities encountered in a
software life cycle?
Solution: The context diagram depicts the system
as a single bubble. Control information should not
be represented in DFD. Column I Column II
Ans. (c) P. R
equirements 1. Module development
3. Consider the following statements about the cyc- capture and integration
lomatic complexity of the control flow graph of a Q. Design 2. Domain analysis
program module. Which of these are TRUE? R. Implementation 3. Structural and
I. The cyclomatic complexity of a module is equal behavioural modelling
to the maximum number of linearly indepen- S. Maintenance 4. Performance tuning
dent circuits in the graph.
II. The cyclomatic complexity of a module is the
number of decisions in the module plus one, (a) P-3, Q-2, R-4, S-1
where a decision is effectively any conditional (b) P-2, Q-3, R-1, S-4
statement in the module. (c) P-3, Q-2, R-1, S-4
III. The cyclomatic complexity can also be used as (d) P-2, Q-3, R-4, S-1
a number of linearly independent paths that (GATE 2010: 1 Mark)
should be tested during path coverage testing.
Solution: Requirement capture: Domain analysis
(a) I and II (b) II and III Design: Structural and behavioural modelling
(c) I and III (d) I, II and III Implementation: Module development and integration
(GATE 2009: 2 Marks) Maintenance: Performance tuning
Solution: The cyclomatic complexity of a module Ans. (b)
is not equal to the maximum number of linearly 6. The following program is to be tested for the state-
independent circuits in the graph. ment coverage:
Ans. (b)
begin
4. The cyclomatic complexity of each of the modules if (a==b) {S1; exit;}
A and B shown below is 10. What is the cyclomatic else if (c==d) {S2;}
complexity of the sequential integration shown on else {S3; exit;}
the right-hand side? S4;
end
The test cases T1, T2, T3 and T4 given below are

A expressed in terms of the properties satisfied by the
values of variables a, b, c and d. The exact values
are not given.
A B
T1: a, b, c and d are all equal
B T2: a, b, c and d are all distinct
T3: a = b and c != d
T4: a != b and c = d
Chapter 9.indd 480 4/9/2015 10:06:51 AM

Which of the test suites given below ensures cover- five years. Various parameters for the company are
age of statements S1, S2, S3 and S4? given in the table below.
(a) T1, T2, T3 (b) T2, T4 Parameter Language L1 Language L2
(c) T3, T4 (d) T1, T2, T4
(GATE 2010: 2 Marks) Man years LOC/10000 LOC/10000
needed for
Solution: development
T1 covers S1
Development cost $$1000000 $$750000
T2 covers S3
per man year
and T4 covers S2 and S4.
So, the test suit is T1,T2 and T4 Maintenance 5 years 5 years
Ans. (d) time
Cost of $$1000000 $$50000
7. Which one of the following is NOT desired in a
maintenance per
good SRS document?
year
(a) Functional requirements
(b) Non-functional requirements Total cost of the project includes cost of develop-
(c) Goals of implementation ment and maintenance. What is the LOC for L1
(d) Algorithms for software implementation for which the cost of the project using L1 is equal
(GATE 2011: 1 Mark) to the project using L2?
Solution: SRS document is prepared at the starting (a) 4000 (b) 5000 (c) 4333 (d) 4667
phase of software development cycle and only client (GATE 2011: 1 Mark)
software requirement is noted down. Algorithm
for software implementation is designed in later Solution: Number of lines of code for L1 = l
stages after the requirement phase. Number of lines of code for L2 =2l
Ans. (d) Total cost for L1 = Total cost for L2
l
8. A company needs to develop digital signal pro- × 1000000 + 5 × 100000
10000
cessing software for one of its newest inven-
2l
tions. The software is expected to have 40000 = × 750000 + 5 × 50000
lines of code. The company needs to determine 10000
the effort in person-months needed to develop Solving, we get l = 5000
this software using the basic COCOMO model. Ans. (b)
The multiplicative factor for this model is given 10. The following is the comment written for a C
as 2.8 for the software development on embed- function:
ded systems, while the exponentiation factor is
given as 1.20. What is the estimated effort in /*This function computes the roots of a
person-months? quadratic equation a.x^2 + b.x + c = 0.
The function stores two real roots in
(a) 234.25 (b) 932.50 *root1 and *root2, and returns the
(c) 287.80 (d) 122.40 status of validity of roots. It handles
(GATE 2011: 1 Mark) four different kinds of cases.
(i) When coefficient a is zero irre-
Solution: Estimated effort in person-months =
spective of discriminant
a1(KLOC)a2
(ii) When discriminant is positive
Given that a1 = 2.8 and a2 = 1.20
(iii) When discriminant is zero
So, KLOC = 40
(iv) When discriminant is negative.
2.8 (40)1.20 = 234.25 person-months
Only in cases (ii) and (iii), the stored
Ans. (a)
roots are valid. Otherwise 0 is stored
9. A company needs to develop a strategy for software in the roots. The function returns 0 when
product development for which it has a choice the roots are valid and –1 otherwise.
of two programming languages L1 and L2. The The function also ensures root1 > = root2.
number of lines of code (LOC) developed using L2 int get_QuadRoots (float a, float b, float
is estimated to be twice the LOC developed with c, float *root1, float *root2);
L1. The product will have to be maintained for */
Chapter 9.indd 481 4/9/2015 10:06:51 AM

A software test engineer is assigned the job of doing (c) A

verage cohesion goes down and coupling also
black box testing. He comes up with the following test reduces.
cases, many of which are redundant. (d) Average cohesion and coupling increase.
Test Input Set Expected Output Set
Solution: Coupling = Number of external links/
Case
a b c root1 root2 Return Number of modules
Value
Cohesion = Number of internal links/Number of
T1 0.0 0.0 7.0 0.0 0.0 −1 methods
Before modification, we have
T2 0.0 1.0 3.0 0.0 0.0 −1
Coupling = 2/2 =1
Cohesion for M1 = 8/4 =2
T3 1.0 2.0 1.0 −1.0 −1.0 0
T4 4.0 −12.0 9.0 1.5 1.5 0 After modification, we have
Coupling = 2/2 =1
T5 1.0 −2.0 −3.0 3.0 −1.0 0 Cohesion for M1 = 6/3 =2
T6 1.0 1.0 4.0 0.0 0.0 −1 Hence, there is no change.
Ans. (a)
Which one of the following options provide the set of Common Data Questions 12 and 13: The proce-
non-redundant tests using equivalence class partitioning dure given below is required to find and replace
approach from input perspective for black box testing? certain characters inside an input character
string supplied in array A. The characters to be
(a) T1, T2, T3, T6
replaced are supplied in array oldc, while their
(b) T1, T3, T4, T5
respective replacement characters are supplied in
(c) T2, T4, T5, T6
array newc. Array A has a fixed length of five
(b) T2, T3, T4, T5
characters, while arrays oldc and newc contain
(GATE 2012: 2 Marks) three characters each. However, the procedure is
Solution flawed.
T1 and T2 checks the same condition of a = 0
void find_and_replace (char *A, char *oldc, char *newc){
T3 and T4, both have D = 0 for (int i=0; i<5; i++)
These should not be in the same set. for (int j=0; j<3; j++)
Ans. (c) if (A[i] == oldc[j]) A[i] = newc[j];
}
11. The following figure represents access graphs of
two modules M1 and M2. The filled circles rep- The procedure is tested with the following four test
resent methods and the unfilled circles represent cases.
attributes. If method m is moved to module M2
(1) oldc = “abc”, newc = “dab”
keeping the attributes where they are, what can
(2) oldc = “cde”, newc = “bcd”
we say about the average cohesion and coupling
(3) oldc = “bca”, newc = “cda”
between modules in the system of two modules?
(4) oldc = “abc”, newc = “bac”
Module M1 Module M2 (GATE 2013: 2 Marks)
12. The tester now tests the program on all input

strings of length five consisting of characters ‘a’,
‘b’, ‘c’, ‘d’ and ‘e’ with duplicates allowed. If
the tester carries out this testing with the four test
m cases given above, how many test cases will be able
to capture the flaw?
(a) There is no change. (a) Only one (b) Only two
(b) Average cohesion goes up but coupling is reduced. (c) Only three (d) All four
Chapter 9.indd 482 4/9/2015 10:06:51 AM

Solution Solution
The test cases 3 and 4 are the only cases that Test case oldc = “abc”, newc = “bac” will
capture the flaw. The code does not work prop- detect the fault.
erly when an old character is replaced by a new Ans. (c)
character, and the new character is again replaced 14. In the context of modular software design, which
by another new character. This does not happen one of the following combinations is desirable?
in test cases (1) and (2), it happens only in cases
(3) and (4). (a) High cohesion and high coupling
(b) High cohesion and low coupling
Ans. (b)
(c) Low cohesion and high coupling
(d) Low cohesion and low coupling
13. If array A is made to hold the string “abcde”,
which of the above four test cases will be successful Solution: Cohesion measures the internal depen-
in exposing the flaw in this procedure? dency within module, whereas coupling presents
intermodule dependence. So, there should be high
(a) None (b) 2 only cohesion and low coupling.
(c) 3 and 4 only (d) 4 only
Ans. (b)
PRACTICE EXERCISES
Set 1
1. SRS is also known as specification of 5. All activities lying on critical path have slack time
equal to
(a) White box testing (b) Stress testing
(c) Integrated testing (d) Black-box testing (a) 0 (b) 1
(c) 2 (d) None of above
2. SRD stands for
6. A software agent is defined as
(a) Software requirements definition
(b) Structured requirements definition I. Software developed for accomplishing a given task.
(c) Software requirements diagram II. A computer program which is capable of acting
(d) Structured requirements diagram on behalf of the user in order to accomplish a
given computational task.
3. The symbol represents III. An open source software for accomplishing a
given task.
(a) mandatory 1 cardinality (a) I (b) II
(b) mandatory many cardinality (c) III (d) All of the above
(c) optional 0 or 1 cardinality
(d) optional zero-many cardinality 7. In terms of a system, finished products and infor-
mation are examples of:
4. Software testing is
(a) Input (b) Output
(a) The process of establishing that errors are not (c) Feedback (d) Control
present.
8. For a well-understood data processing applica-
(b) The process of establishing confidence that a
tion, it is best to use
program does what it is supposed to do.
(c) The process of executing a program to show (a) the waterfall model
that it is working as per specifications. (b) the prototyping model
(d) The process of executing a program with the (c) the evolutionary model
intent of finding errors. (d) the spiral model
Chapter 9.indd 483 4/9/2015 10:06:51 AM

9. A COCOMO model is 19. The feature of the object-oriented paradigm which

helps code reuse is
(a) Common Cost Estimation Model
(b) Constructive Cost Estimation Model (a) Object (b) Class
(c) Complete Cost Estimation Model (c) Inheritance (d) Aggregation
(d) Comprehensive Cost Estimation Model
20. The worst type of coupling is
10. IEEE 830-1993 is an IEEE recommended stan-
dard for (a) Data coupling (b) Control coupling
(c) Stamp coupling (d) Content coupling
(a) Software requirement specification
(b) Software design 21. Modules X and Y operate on the same input and
(c) Testing output data, then the cohesion is
(d) Both (a) and (b)
(a) Sequential (b) Communicational
11. All the modules of the system are integrated and (c) Procedural (d) Logical
tested as complete system in the case of
22. The desired level of coupling is
(a) Bottom-up testing (b) Top-down testing
(c) Sandwich testing (d) Big-Bang testing (a) No coupling (b) Control coupling
(c) Common coupling (d) Data coupling
12. If the objects focus on the problem domain, then
we are concerned with 23. Changes made to an information system to add
(a) Object-Oriented Analysis the desired but not necessarily the required fea-
(b) Object-Oriented Design tures is called
(c) Object-Oriented Analysis and Design (a) Preventative maintenance
(d) None of the above (b) Adaptive maintenance
13. A fault simulation testing technique is (c) Corrective maintenance
(d) Perfective maintenance
(a) Mutation testing (b) Stress testing
(c) Black-box testing (d) White-box testing 24. Output comparators are used in
14. If every requirement stated in the SRS has only (a) Static testing of single module
one interpretation, SRS is said to be (b) Dynamic testing of single module
(a) Correct (b) Unambiguous (c) Static testing of single and multiple modules
(c) Consistent (d) Verifiable (d) Dynamic testing of single and multiple modules
15. The model in which the requirements are imple- 25. Coupling and cohesion can be represented using a
mented by category is
(a) Cause-effect graph (b) Dependence matrix
(a) Evolutionary development model (c) Structure chart (d) SRS
(b) Waterfall model
(c) Prototyping 26. KPA in CMM stands for
(d) Iterative enhancement model (a) Key Process Area
16. Which one of the following is a fault base testing (b) Key Product Area
technique? (c) Key Principal Area
(d) Key Performance Area
(a) Unit testing (b) Beta testing
(c) Stress testing (d) Mutation testing 27. In the spiral model, `risk analysis’ is performed
17. Which of the following statements is true? (a) In the first loop
(a) Abstract data types are the same as classes. (b) In the first and second loop
(b) Abstract data types do not allow inheritance. (c) In every loop
(c) Classes cannot inherit from the same base class. (d) Before using spiral model
(d) Object have state and behaviour.
28. Each time a defect gets detected and fixed, the
18. If every requirement can be checked by a cost- reliability of a software product
effective process, then the SRS is
(a) Increases (b) Decreases
(a) Verifiable (b) Traceable (c) Remains constant (d) Cannot say anything
(c) Modifiable (d) Complete
Chapter 9.indd 484 4/9/2015 10:06:51 AM

29. The level at which the software uses scarce

c. Common coupling iii. W
hen complete data
resources is
structure is passed
(a) Reliability (b) Efficiency from one module to
(c) Portability (d) All of the above another
30. Alpha and beta testing are forms of d. Content coupling iv. W
hen the control is
passed from one
(a) Acceptance testing (b) Integration testing module to the middle
(c) System testing (d) Unit testing of another
31. Which one of the following is not a risk manage-
ment technique for managing the risk due to unre- Codes:
alistic schedules and budgets?
a b c d
(a) Detailed multisource cost and schedule (a) iii ii i iv
estimation (b) ii iii i iv
(b) Design cost (c) ii iii iv i
(c) Incremental development (d) iii ii iv i
(d) Information hiding
32. _______ of a system is the structure or structures 37. A process which defines a series of tasks that have
of the system which comprise software elements, the following four primary objectives is known as
the externally visible properties of these elements 1. To identify all items that collectively define the
and the relationship amongst them. software configuration.
(a) Software construction (b) Software evolution 2. To manage changes to one or more of these items.
(c) Software architecture (d) Software reuse 3. To facilitate the construction of different versions
of an application.
33. In function point analysis, the number of com-
plexity adjustment factors is To ensure that software quality is maintained as
the configuration evolves over time.
(a) 10 (b) 12 (c) 14 (d) 20
(a) Software Quality Management Process
34. Regression testing is primarily related to (b) Software Configuration Management Process
(a) Functional testing (b) Development testing (c) Software Version Management Process
(c) Data flow testing (d) Maintenance testing (d) Software Change Management Process
35. Which one of the following is not a definition of error? 38. Software Configuration Management is the disci-
pline for systematically controlling
(a) It refers to the discrepancy between a com-
puted, observed or measured value and the (a) T
he changes due to the evolution of work prod-
true, specified or theoretically correct value. ucts as the project proceeds.
(b) It refers to the actual output of software and (b) The changes due to defects (bugs) being found
the correct output. and then fixed.
(c) It refers to a condition that causes a system to fail. (c) The changes due to requirement changes
(d) It refers to human actions that result in soft- (d) All of the above
ware containing a defect or fault. 39. Which one of the following is not a step of require-
36. Match the following: ment engineering?
(a) Requirement elicitation
List − I List − II (b) Requirement analysis
a. Data coupling i. M
odule A and Module (c) Requirement design
B have shared data (d) Requirement documentation
b. Stamp coupling ii. D
ependency between 40. __________ is a process model that removes defects
modules is based before they can precipitate serious hazards.
on the fact they (a) Incremental model
communicate by only (b) Spiral model
passing of data (c) Clean room software engineering
(Continued) (d) Agile model
Chapter 9.indd 485 4/9/2015 10:06:52 AM

41. Equivalence partitioning is a __________ method (a) Debugging

that divides the input domain of a program into (b) Testing
classes of data from which test cases can be (c) Designing
derived. (d) Software quality assurance
(a) White-box testing 48. Are we building the right product?

(b) Black-box testing
(c) Orthogonal array testing This statement refers to
(d) Stress testing (a) Verification
(b) Validation
42. The following three golden rules are for: (c) Testing
(i) Place the user in control (d) Software quality assurance
(ii) Reduce the user’s memory load
(iii) Make the interface consistent 49. Main aim of software engineering is to produce
(a) User satisfaction (a) program
(b) Good interface design (b) software
(c) Saving system’s resources (c) within budget
(d) None of these (d) software within budget in the given schedule
43. Software safety is a __________ activity that Set 2
focuses on the identification and assessment of
potential hazards that may affect software nega- 1. Company needs to develop digital signal process-
tively and cause an entire system to fail. ing software for one of its newest invention. The
(a) Risk mitigation, monitoring and management software is expected to have 36000 LOC. The
(b) Software quality assurance company needs to determine the efforts in person-
(c) Software cost estimation month needed to develop this software using the
(d) Defect removal efficiency basic COCOMO model.
The multiplicative factor for this model is given as
44. While estimating the cost of software, LOC and
2.6 for the software development on the embedded
function points (FP) are used to measure which
system while exponential factor is given as 1.30.
one of the following?
The estimated effort in person-month is .
(a) Length of code
(b) Size of software 2. The relationship between a derived class (or sub-
(c) Functionality of software class) and base class is referred to as
(d) None of the above (a) Association
45. The factors that determine the quality of a soft- (b) Inheritance
ware system are (c) Polymorphism
(d) Instantiation
(a) Correctness, reliability (e) Aggregation
(b) Efficiency, usability, maintainability
(c) Testability, portability, accuracy, error toler- 3. A design is said to be a good design if the compo-
ances, expandability, access control, audit nents are
(d) All of the above
(a) Strongly coupled
46. ______ establishes information about when, why (b) Strongly coupled and weakly cohesive
and by whom changes are made in software. (c) Strongly coupled and strongly cohesive
(d) Strongly cohesive and weakly coupled
(a) Software configuration management
(b) Change control 4. If a control switch is passed as an argument, this
(c) Version control is an example of coupling.
(d) An audit trail
(a) Content (b) Common
47. ______ is an “umbrella” activity that is applied (c) Control (d) Stamp
throughout the software engineering process. (e) Data
Chapter 9.indd 486 4/9/2015 10:06:52 AM

5. The aim of software engineering is to produce (c) Functionality

software that is (d) Reliability
(e) Usability
(a) Fault-free
(b) Delivered on time 8. Which of the following is not a `concern’ during the
(c) Delivered within budget management of a software project?
(d) All of these are the aims of software engineering
(a) Product quantity
6. A systematic approach to software development, (b) Time
as epitomised by the various life-cycle models, is (c) Project/Product information
useful in (d) Money
(a) Helping us understand the nature of the soft- 9. Black-box testing is also called
ware product (a) Specification-based testing
(b) Convincing the customer that we know what (b) Structural testing
we are doing (c) Verification
(c) Filling texts on software engineering (d) Unit testing
(d) Managing the various activities necessary to (e) Stress testing
get the job done
10. Project risk factor is considered in?
7. A process view in software engineering would con-
(a) Spiral model
sider which of the following
(b) Waterfall model
(a) Product performance (c) Prototyping model
(b) Staffing (d) Iterative enhancement model
Set 1
1. (d) 10. (a) 19. (c) 28. (a) 37. (b) 46. (d)
2. (b) 11. (d) 20. (a) 29. (b) 38. (d) 47. (d)
3. (d) 12. (a) 21. (b) 30. (a) 39. (c) 48. (b)
4. (d) 13. (a) 22. (d) 31. (d) 40. (c) 49. (d)
5. (a) 14. (b) 23. (d) 32. (c) 41. (b)
6. (b) 15. (a) 24. (d) 33. (c) 42. (b)
7. (b) 16. (d) 25. (b) 34. (d) 43. (b)
8. (a) 17. (b) 26. (a) 35. (c) 44. (b)
9. (b) 18. (a) 27. (c) 36. (b) 45. (d)
Set 2 2. (b) A derived class inherits all the attributes and

property of parent class (base class).
1. (274.26) Effort = 2.6 EAF × (36k)1.30 (EAF = by
default =1) 3. (d) The aim is to minimize the interaction between
= 2.6 × (36k)1.30 modules and maximize interaction within a module.
= 2.6 × (36)1.20 4. (c) Two modules are control coupled if one passes
= 274.26 an element of control to another.
Chapter 9.indd 487 4/9/2015 10:06:52 AM

5. (d) The aim of software engineering is to produce 8. (a) Product quantity would not include during the
software that is fault-free, delivered on time, deliv- management of a software project.
ered within budget and satisfies users’ needs.
9. (a) Black-box testing is another name for specifica-
6. (d) A systematic approach to software devel- tion-based testing.
opment, as epitomised by the various life-cycle
10. (a) Aim of risk analysis phase in the spiral model
models, is useful in managing the various activities
is to eliminate the high-risk problems before they
necessary to get the job done.
threaten the project operation or cost.
7. (b) Staffing is the apt choice among the given
Chapter 9.indd 488 4/9/2015 10:06:52 AM

UNIT X: COMPUTER NETWORKS
7
Number of questions
6
5
4 Marks 1
Marks 2
3
2
1
0
2015 2014 2013 2012 2011 2010
Year Concept
2015 TCP, IP Header, File organization, Symmetric key cryptography, LAN, Stop and
wait protocol, CSMA/CD, Go-back-N sliding window protocol, IPv4 Addressing,
Data transmission, HTTP, UDP datagram, Routing table
2014 IPv4 addresses, MTU, OSI layer, Frames, IP header, Classless inter-domain routing
2013 Transport layer protocols, Data link layer concepts, IPv4 concepts
2012 IPv4 concepts, Transport layer concepts, Application layer concepts, IP address,
Propagation delay in networks
2011 Firewall concept, Distance vector, Application layer concepts, HTML
2010 IP address concepts, Application layer, Router concepts
Chapter 10.indd 489 4/9/2015 10:08:32 AM

Chapter 10.indd 490 4/9/2015 10:08:32 AM
CHAPTER 10
COMPUTER NETWORKS
Syllabus: Computer networks: ISO/OSI stack, LAN technologies (Ethernet, token ring), Flow-and error-control
techniques, Routing algorithms, Congestion control, TCP/UDP and sockets, IP(v4), Application-layer protocols
(ICMP, DNS, SMTP, POP, FTP, HTTP); Basic concepts of hubs, switches, gateways and routers; Network security:
basic concepts of public key and private key cryptography, digital signature, firewalls.
10.1 INTRODUCTION 10.2 NETWORK
Computer network is a collection of autonomous devices Two or more devices connecting to each other through
interconnected via a medium. The medium may be a any medium forms a network. To connect the devices
guided medium, a wireless medium or a satellite commu- there are two possible ways:
nication. To understand data communication, different
layered architectures have been presented. All layers exe- 1. Point-to-point connection: This provides a
cute different protocols to communicate the data success- dedicated link between the two devices.
fully from one end to the other. The intermediary devices 2. Multipoint connection: This is also known as
such as switch, hub, router or gateway help in advancing multi-drop connection. In this case, channel capac-
this communication. Although the security requirements ity is shared by more than two devices. In general,
may be different depending upon the application, network this is used in practice. All the five examples in
security cannot be left aside when studying networks. Fig. 10.1 are multipoint.
Chapter 10.indd 491 4/9/2015 10:08:33 AM

492 Chapter 10: Computer Networks
is active, it may amplify or regenerate the signals. The

following are the drawbacks of a star topology:
Mesh Ring Bus Star Tree 1. If the central hub fails, no communication is possible.
topology topology topology topology topology 2. Cabling cost is more.
Figure 10.1 | Network topology. For example, Ethernet 10 base T is a popular example.
10.2.1 Network Topology 10.2.1.4 Mesh Topology

The physical or logical view of interconnection among All the devices are connected through peer-to-peer links.
devices is known as topology. A fully connected mesh will have n(n−1)/2 physical
Ring and mesh topology are examples of peer-to-peer channels to connect n devices. The advantage of mesh
relationship because here all the devices share the link topology is security and privacy. A dedicated link will
equally. Bus, star and tree topologies are examples of eliminate traffic problems, and fault diagnose is easy
primary–secondary relationship, where one device con- here. But cabling cost and other hardware required make
trols and the other devices have to transmit through it. it difficult to implement in real practice. It is better to
use this topology in backbone network and other topolo-
gies for further network configuration.
10.2.1.1 Bus Topology
In a bus network, all the devices are connected with one 10.2.1.5 Tree Topology
cable. The major benefit here is, we require less cable
length than star topology and expansion is quite easy A tree topology maintains the devices connected to a
with the help of repeaters. central hub as well as to some secondary hubs, which are
The issues associated with bus topology are as follows: again connected to the central hub. It allows an isolated
network which prioritizes communication from different
1. Only one device can send the data at one time. All
computers. It also faces the same problem as in a star
the devices will listen at that time.
topology; failure of central hub will crash the entire net-
2. The data communication is only in one direction.
work. Also, it has high cabling cost.
3. In a bus topology, if the network shuts down then
there is problem in identifying the culprit device.
10.3 LAN TECHNOLOGIES
10.2.1.2 Ring Topology
In a ring topology, each deviceis connected exactly to two LAN (local area network) is an integral part to create
devices to form the ring. Repeaters are used to regener- a network. There are many LAN technologies such as
ate and retransmit each bit. Data travels around the net- Ethernet, token ring, token bus, FDDI (fiber distributed
work in one direction. Data travels in the form of token. data interface) and ATM LAN. Table 10.1 shows the
Additional components do not affect the performance of comparison of Ethernet, token bus and token ring.
the network. Even if the load on the network increases,
the performance of a ring topology is better than a bus
10.3.1 Ethernet
topology. The problems may be listed as follows:
1. Failure of one computer in the ring may lead to Xerox Corporation, Digital Equipment Corporation and
entire communication loss. Intel Corporation developed Ethernet LAN technology
2. Network scaling is difficult. in 1976. Ethernet is based on the IEEE 802.3 specifica-
tion. It is a linear-bus logical topology. Ethernet is the
For example, token ring is defined by IEEE 802.5
most widely used LAN technology in the world. It has
standard.
passed four generations: Standard Ethernet (10 t Mbps),
Fast Ethernet (100 Mbps), Gigabit Ethernet (1 Gbps)
10.2.1.3 Star Topology and Ten-Gigabit Ethernet (10 Gbps). Earlier Ethernet
designed were of two types: Thicknet (thick coaxial
In a star topology, a hub is placed at the central location main trunk cable of 10 mm and Thinnet (thin coaxial
and all the devices are connected to the hub. All com- cable: RG-58 of 5 mm). In 1990, unshielded twisted-pair
munication is possible through the hub only. If the hub (10Base-T) Ethernet came into existence. Ethernet uses
Chapter 10.indd 492 4/9/2015 10:08:33 AM

10.3 LAN TECHNOLOGIES 493
Table 10.1 | Comparison of Ethernet, token bus and token ring

Attribute Ethernet Token Bus Token Ring
Physical Linear Linear Star
Topology
Logical None Ring Ring
Topology
Connection Random By token By token
Node Node added anywhere and Distributed algorithms are Between two specified nodes
Addition anytime responsible for node addition
Cable Used Twisted pair, co-axial and fibre Co-axial Twisted pair and fibre optic
optic
Cable Length 50—2000 m 200—500 m 50—1000 m
Frequency 10—100 Mbps 10 Mbps 4—100 Mbps
Frame 1500 byte 8191 bytes 5000 bytes
Structure
IEEE 802.3 802.4 802.5
Standard
Maintenance No central maintenance By distributed algorithms By a designated monitor node
Performance Immediately transmitted by the Nodes must wait for the token Nodes must wait for token even
nodes, heavy traffic can reduce if no other node is transmitting. if no other node is transmitting.
the effectiveness of transmission During heavy traffic, token During heavy traffic, token
passing provides fair access to passing provides fair access to
all nodes all nodes
Maximum None Bounded, depending on distance Bounded, depending on distance
Delay before spanned and number of nodes spanned and number of nodes
Transmitting
Carrier-Sense Multiple Access with Collision Detection In early token release,

(CSMA/CD) access control scheme. The drawback of
this topology is that if one of the links between the two Throughput for single station or N stations
adjacent nodes fails, the whole network fails.
Data
=
Transmission time
10.3.2 Token Bus
+ (Ring latency Number of stations)
Token bus is a bus (physical view) ring (logical
view) topology. In token bus, nodes are connected In delayed token release,
linearly. However, they make a logical ring, as each node
knows the address of its successor.
Throughput for single station
10.3.3 Token Ring Data
=
Transmission time + Ring latency
Token ring is a star (physical view) ring (logical view) + (Ring latency/Number of stations)
topology. The hub acts as a connector. The wiring inside
the hub makes the ring; the stations are connected to Throughput for N stations
this ring through the two wire connections. It is more
efficient; if a link goes down, it will bypass the hub and Data
=
operate the other nodes. It also improves the scalability Ring latency
of the network. + (Ring latency/Number of stations)
Chapter 10.indd 493 4/9/2015 10:08:34 AM

10.4 ISO/OSI STACK The following are some issues listed which may create
problems before/during transmission:
1. Compatibility of mechanical and electrical interfaces
There are two reference models based on the network 2. Working of physical transmission media
architectures. One is the International Standards 3. Deciding on the number of bits per second to be sent
Organization/Open Systems Interconnection (Reference 4. Finding out whether transmission is simplex or duplex
Model 1984) (ISO/OSI) model and other is the 5. Establishing and terminating the initial communi-
Transmission Control Protocol/Internet Protocol (TCP/ cation when both sender and receiver are finished
IP) model. The layered architecture of both the models
Message size (bits)
has been compared in Fig. 10.2. Transmission time =
Bandwidth (bits/s)
OSI Model TCP/IP Model Distance
Propagation time =
Velocity
Application
Problem 10.1: Message size = 1 Kb
Presentation Application
Bandwidth = 1 Mbps
103 × 23 bits
= 8 × 10−3 s
1 Kb
Session Transmission time = =
1 Mbps 106 bits/s
Transport Tranport If link utilisation is 50%, what is the relation between
transmission time and propagation time?
Network Internet Solution:
Link utilization of sender
Data link
Transmission time
Network interface =
Physical Transmission time + 2× Propagation time
1 Transmission time
=
Figure 10.2 | OSI and TCP/IP model. 2 Transmission time + 2 × Propagation time
The ISO/OSI model has seven layers while TCP/IP ransmission time + 2 × Propagation time
T
has only four layers. Session and presentation layers are = 2 × Transmission time
completely missing. Characteristics of the session layer Therefore, Transmission time = 2 × Propagation time
are provided by the transport layer, and the application Transmission time = 2× Propagation time
layer bears the accountability of the presentation layer.
or Transmission time = Round-trip time
Host to network is the lowest layer in the TCP/IP model
and its duty is to send IP packets to the network. In the Let L be the message, B the bandwidth and R the
ISO/OSI model, the network layer provides connection- round-trip time (RTT),
oriented as well as connectionless services, but the TCP/ LB
Link utilization of sender =
IP model provides only connectionless services. (L B) + R
If M is a message and H is the header that is added at L
every layer and N layers are present in hierarchy, then h=
L + BR
the fraction of header that is passed in the total content
is calculated as follows: In the above expression,
M 1. if L = BR, h = 50%
Fraction of data =
NH + M 2. if L > BR, h > 50%
3. if L < BR, h < 50%
10.4.1 ISO/OSI Model 4. if L >>> BR, h ∼− 100%
5. if L <<< BR, h ∼− 0%
10.4.1.1 Layer 1: Physical Layer
10.4.1.2 Layer 2: Data Link Layer
The major responsibilities of physical layer are transmission
of raw bit stream and to form the physical interface between Data link layer provides reliable transfer of informa-
two communicating devices. The conversion of analog to tion between two adjacent nodes. Also, it provides
digital and digital to analog is performed at this layer. frame-level error control and flow control. It provides
Chapter 10.indd 494 4/9/2015 10:08:36 AM

10.4 ISO/OSI STACK 495
communication between machines on the same network. frame, the frame is resent, the copy is held and the timer
Communication between two devices can be simplex, is restarted. For the corrupted and lost ACK frame,
half-duplex, or full-duplex. In simplex, the communica- sequence numbers can be used. In the data frame, a field
tion is unidirectional. In half-duplex, each device can is added for the sequence number. The sequence num-
both transmit and receive, but not at the same time. bers are based on modulo-2 arithmetic. This protocol
In full-duplex, both devices can transmit and receive is also having acknowledgement numbers, which speci-
simultaneously. fies the sequence number of the next frame expected by
This layer is responsible for encoding and decoding i.e. the receiver. A data frame uses a sequence number and
converting bits to signals at sender site and recovering an ACK frame uses an acknowledgement number. The
bits from received signals at receiver side; frame cre- control variable of sender keeps the sequence number for
ation i.e. deciding a minimum unit for sending bits; error the next frame to be sent (0 or 1). The control variable
detection and/or correction of frames through parity or of receiver keeps the number of the next frame expected.
CRC and flow control using ARQ, sliding WINDOW When a frame is sent, the value of the control variable
etc. The functionality of data link layer is as follows: of sender is incremented. When a frame is received, the
value of the control variable of receiveris incremented.
1. Encoding: Signals propagate over a physical The stop-and-wait ARQ protocol is very inefficient if
medium — (1) modulate electromagnetic waves (var the channel is thick (large bandwidth) and long (long
ying voltage); (2) encode binary data onto signals round-trip delay). Other drawback of this protocol is
(e.g., 0 as low signal or non-return to zero, NRZ, that it does not support pipelining, as it does not sup-
and 1 as high signal or non-return to zero inverted, port multiple frames.Pipelining helps in improving the
NRZI); make a transition from current signal to efficiency of the transmission, if the number of bits in
encode a 1 or stay at the current signal to encode transition is large with respect to the bandwidth-delay
a 0;(3) Manchester (transmit XOR of the NRZ- product.
encoded data and the clock only 50% efficient). For stop-and-wait ARQ,
In Manchester encoding, a clock signal and data
signal are mixed together by XORing operation. The Message size
Transmission time =
clock makes a clock transition in every bit time, so Bandwidth
it runs at twice the bit rate. When it is XORed with
0 then it makes low-to-high transition, it acts as a Distance of the link
Propagation delay =
clock and when it is XORed with 1 then it makes Velocity
high-to-low transition, it acts as a data signal.
2. Framing: The basic data unit at the date link Link utilisation of sender or throughput is given by
layer is called a `frame’ which is a collection of bits
Transmission time
in sequence boundary. In order to mark boundaries h=
of frame starting and ending characters are used. Transmission time + 2× Propagation delay
3. Flow control: It is a mechanism which informs
the sender about the amount of data transmission • G
o-back-N automatic repeat request: This protocol
before receiving an acknowledgement from the helps in improving the efficiency of transmission by
receiver. As the receiving device has limited speed filling the pipe. It supports multiple frames during
for processing the incoming data and limited wait for acknowledgement. The sequence numbers
memory to store data, so it informs the sending are modulo 2m, where m is the size of the sequence
device by sending few frames and stop. The receiv- number field in bits. Sliding window defines the
ing device has a buffer, a block of memory, for stor- range of sequence numbers related to the sender
ing extra incoming data before processing. and receiver. When the timer expires, the sender
4. Error control: It is a mechanism which informs resends all outstanding frames. Stop-and-wait
the sender about the retransmission of the damaged ARQ is a special case of Go-Back-N ARQ in which
and lost frames during transmission. It is a method the size of the send window is 1. It supports one
of error detection and error correction. Automatic receiver window size. This protocol is very inef-
repeat request (ARQ) is a process in which when- ficient for a noisy link. In noisy link, frames are
ever an error is detected, the receiving device sends resending again and again, which uses bandwidth
the request for retransmission to sender. and slow down the transmission. For Go-Back-N
5. Techniques of flow and error control: ARQ,
•• Stop-and-wait automatic repeat request: In this
protocol, the sender starts the timer and keeps the Sender window size < 2m
copy of the sent frame. If the timer expires and Maximum sequence number = 2m − 1
there is no acknowledgement (ACK) for the sent where m is the number of segment bits.
Chapter 10.indd 495 4/9/2015 10:08:36 AM

If maximum sequence number is s, then the number •• Subtract the remainder from S(x) using
of sequence bits = log (s+1) modulo-2 subtraction.
Message size The result is the checksummed message.
Transmission time = 9. Decoding a CRC procedure:
Bandwidth
•• Let n be the length of the checksummed mes-

Distance
Propagation delay = sage in bits.
Velocity •• Divide the checksummed message by the code
Link utilisation of sender or throughput is given by polynomial using modulo-2 division. If the
Transmission time
h= remainder is zero, there is no error detected.
Transmission time + 2× Propagation delay 10. Choosing a CRC polynomial: The longer the
Total bits polynomial, the smaller the probability of unde-
Number of frames (Window size) = tected error.
Frame size
•• Selective repeat automatic repeat request: In
this protocol, the damaged frame is resent in Problem 10.3: If the frame is 1101011011 and generator
the network. It is efficient for noisy links, but it is x4 + x + 1, what would be the transmitted frame?
requires complex processing at the receiver end. Solution: The polynomial x4 + x + 1 corresponds to
Sender window size divisor 10011(k = 5 bits)
= Receiver window size ≤ 2m − 1 Data word (1101011011) of N = 10 bits is augmented
If Q is the size of the window, then the number with (k − 1) zero’s.
of sequence bits = log2Q + 1
Dividend = 11010110110000
Total bits 110000101
Number of frames (Window size) =
Frame size
10011 11010110110000
Problem 10.2: Calculate the link utilisation for stop-and- 10011
wait flow control mechanism if the frame size is 4800 010011
bits, bit rate is 9600 bps and distance between devices 10011
is 2000 km. Given propagation speed is 200000 km/s. 0000010110
10011
Solution:
0010100
4800 9600
h= = 0.96 10011
(4800 9600) + 2 × (2000 200000) 001110
After dividing the message 1101011011 by 10011 the
6. Parity bits: Append a single parity bit to a remainder is 1110, which is CRC. The transmitted
sequence of bits data is data + CRC, which is 1101011011 + 1110 =
•• If using òdd’ parity, the parity bit is calculated 11010110111110.
as making the total number of 1’s in the bit
sequence odd; Data link layer is divided into two sublayers:
•• If using èven’ parity, the parity bit makes the 1. Multiple access control sublayer: It provides
total number of 1’s in the bit sequence even controlled access to shared transmission media.
For example, if −Q is for even parity, what’s the 2. Logical link control sublayer: It is responsible
parity bit for 00010101? The problem with parity for error and flow control.
bit is that it only detects when there are an odd
number of bit errors. When multiple users share a common communication
7. Polynomial codes: It can detect errors on large link, multiple access protocols are used to coordinate
chunks of data; has low overhead; is more robust access to common link (Fig. 10.3).
than parity bit; and requires the use of a code 1. Random Access Protocols: The following
polynomial. are the different random access protocols used for
8. Cyclic Redundancy Check (CRC): Example accessing the shared transmission channel:
of a polynomial code •• Pure ALOHA: It says that whenever a station
Procedure: is having the data, they can send immediately.
•• Let r be the degree of the code polynomial. The time at which the collision occurs is called
Append r zero bits to the end of the transmitted vulnerable time.
bit string. Call the entire bit string S(x). Tp = Maximum propagation time
•• Divide S(x) by the code polynomial using Distance between two stations
modulo-2 division. =
Veelocity
Chapter 10.indd 496 4/9/2015 10:08:38 AM

10.4 ISO/OSI STACK 497
Multiple access
protocols
Random access Controlled Channelization

access
ALOHA CSMA CSMA/CD CSMA/CA Reservation FDMA
Pure ALOHA Polling TDMA
Slotted ALOHA Token passing CDMA
Figure 10.3 | Multiple access protocols.
Suppose Tfr is the average transmission time for a 10.4.1.3 Layer 3: Network Layer
frame, then
Frame size The network layer is responsible for host-to-host delivery
Tfr = =G and path selection between endsystems (routing). The
Bandwidth
fragmentation, reassembly and translation between dif-
Throughput (S) = G × e−2G ferent network types are also performed at this layer. In
Vulnerable time = 2 × Tfr other words, communication between nodes is possible in
Smax = 18.4% different networks through this layer.
•• Slotted ALOHA: It says that if stations are Packet delivery can be accomplished by using either
ready with the data, they have to wait for the a connection-oriented or a connectionless network ser-
required time slot and can transmit data exactly vice. In a connection-oriented protocol, the connection
at that timeslot. is established before sending the packets so route is
established before and all the packets have to follow that
Throughput (S) = G × e−G route. Example: Frame relay and ATM uses this service.
Vulnerable time = Tfr In connectionless protocols, the network layer protocol
Smax = 36.8% treats each packet independently. The packets in a mes-
•• CSMA (Carrier Sense Multiple Access): If a sage may or may not follow the same path to their des-
station is ready with the data, it senses the tination. For example, Internet uses this type of service.
channel, and if channel found idle, data is trans- Switching can be broadly divided into three catego-
mitted, otherwise the station has to wait for ries: circuit switching, packet switching and message
random amount of time. switching, as shown in Table 10.2.
Table 10.2 | Comparison of circuit, message and packet switching

Attribute Circuit Message Packet
Dedicated Physical Path Yes No No
Bandwidth Available Fixed Dynamic
Route Selection Static Dynamic (per message) Dynamic (per packet)
Potentially Wasted Yes No No
Bandwidth
Stored and Forward No Stored Queued not stored
Transmission
Transmission Length Unlimited Maximum length No maximum length
Same Route Follows Yes No No
(Continued)
Chapter 10.indd 497 4/9/2015 10:08:38 AM


Attribute Circuit Message Packet
Packets Arrive in Order Yes No
Call Setup Required Not Required Not required
Congestion Route Blocking At setup time, if user busy No message blocking On every packet
Possible Reordering No No Yes
Response to Link Failure Data loss Rerouting/Retransmission Rerouting/Retrans-mission
Message Delivery Guaranteed Depends on the network Depends on the network
Delivery Time Negligible Long Short
Path Establishment Switch path for entire For each message For each packet
connection time
Charging Per minute Per message Per packet
The Internet is a global system of computer net- Table 10.3 | Comparison between TCP and UDP
works that are interconnected worldwide and all use the
standard Internet protocol suite (TCP/IP) for linking. Attribute TCP UDP
1. Internet as a Datagram Network: The Connection Connection Connectionless
Internet, at the network layer, is a packet-switched Management oriented
network. Switching can be generally divided into End-to-End Dedicated No dedicated
three broad categories: circuit switching, packet Connection connection connection
switching, and message switching. Packet switch-
ing uses either the virtual circuit approach or the Reliability Reliable Unreliable
datagram approach. Transmission Byte oriented Message oriented
The Internet chooses the datagram approach Acknowledgement Yes No
to switching in the network layer, and uses
Retransmission Automatically If needed
the universal addresses defined in the network
layer to route packets from the source to the Congestion Yes No
destination. Switching at the network layer in the Control
Internet uses the datagram approach to packet Flow Control Yes No
switching.
Fault Tolerance No No
2. Why Internet Uses Connectionless Network?
Delivery of a packet can be accomplished by using Data Delivery Strictly Unordered
either a connection-oriented through TCP or a ordered
connectionless network service through UDP (see Security Yes Yes
Table 10.3). In a connection-oriented service, the Overhead Low Very low
source has to make a connection with the desti-
nation before sending a data packet. Only after Transmission High Very high
establishing connection, a sequence of packets Speed
from source to the destination can be sent on the Data Quantity Small to very Small to moderate
same path that is established before in a sequential Suitability large amount amount of data
order. The connection is terminated only when all of data
the packets of a particular message have been suc-
cessfully. But the communication at the network
layer in the Internet is connectionless. The reason 10.5 ROUTING ALGORITHMS
is that Internet is made of so many heterogeneous
networks that it is almost impossible to create a
connection between every source and destination When the router receives a packet, which route this packet
pair without knowing the nature of the networks should follow to reach to destination is an important con-
in advance. cern. This is one of the major responsibilities of network
Chapter 10.indd 498 4/9/2015 10:08:38 AM

10.5 ROUTING ALGORITHMS 499
layer. This decision is taken by router on the basis of path between a pair of routers, the shortest path
the routing table maintained by it. The algorithm which is chosen. The shortest path is chosen based on the
decides the suitable route is known as routing algorithm. number of hops or geographical distance in kilome-
The desirable properties of any routing algorithm are tres. Dijkstra’s and Bellman—Ford’s algorithm are
correctness, fairness, stability, robustness and optimality. the most famous shortest path algorithms.
These algorithms can be broadly divided into two
Example 10.1
categories:
Flooding: This is again a non-adaptive algorithm and it
1. Adaptive algorithms: The dynamic rout-
sends a copy of the packet to every outgoing line except
ing decision depends on the topology and traf-
the line on which it was received. This guarantees the
fic. Furthermore, adaptive algorithms have been
packet delivery to destination but a large number of
divided into three forms:
•• Centralised: The decision is taken on the basis
packet copies will be generated. Sometimes this count
approaches to infinity and the only solution is to discard
of global information and this is performed by a
the packet using any of the following approach:
centralised node.
•• Isolated: The routing decision is taken based (a) Using a hop counter to avoid forwarding of packets as
on local information. Generally, routers do not number of hops reaches the diameter of the network.
share information with their neighbours. (b) Keeping track of flooded packets.
•• Distributed: A combination of local and global (c) Selective forwarding by forwarding only those
information. relevant packets which approaches to the right
2. Non-adaptive algorithms: The static routing destination.
decision is taken in advance and it is downloaded
by the routers, that is, never change once initial To avoid looping, a sequence number may be added
route has been selected. to each packet’s header. This helps in discarding those
packets whose sequence number is lower than the one
The properties of non-adaptive routing algorithms are already received.
as follows:
Note: Although flooding is inefficient in most applica-
1. Optimality principal: It states that “if router J tions, an exception may be in the case of military appli-
is on the optimal path from router I to router K, cation where large number of routers are placed and
then the optimal path from J to K also falls along robustness is highly desirable.
the same route”. In other words, suppose r1 is the
route from I to J and r2 is rest of the route. Then, if
2. Flow-based routing: This is a non-adaptive
any route is better than r2, it could have improved
algorithm which uses topology and traffic condi-
the overall optimal route. Hence, r1r2 is optimal.
tion for deciding the route. If traffic on some route
2. Sink tree: A set of all optimal routes from any
is more than average, then the route should be
source to a fixed destination form a tree called sink
avoided to achieve optimal path.
tree. Sink trees may be more than one with the
If line capacity and flow is given, delay can be
same path length. The concern of sink tree here is
determined easily using the following formula:
to help routers find the best path.
1
T =
In addition to adaptive and non-adaptive categorisa- mC − l
where 1/m is mean packet size in bits, l is the mean
tion, routing algorithm can be simply categorised into
the following:
number of packets arrived per second and C is the
1. Static routing (shortest path, flooding) line capacity in Kbps.
2. Flow-based routing
3. Dynamic routing (distance vector, link state routing) Example 10.2
4. Hierarchical routing
Distance Vector/Distributed Bellman—Ford/
5. Routing for mobile hosts
Ford—Fulk‑erson Routing Algorithm: This is one
6. Broadcast routing
of the popular examples of dynamic routing algorithm.
7. Multicast routing
Each router maintains a table (called vector) which
These are explained as follows: helps in finding the best-known distance to any destina-
1. Shortest path routing: This non-adaptive tion. It also indicates preferred outgoing line to be used
approach is based on the simplest and most widely to reach the destination. The performance metric can be
used principle. Each node is treated as a router the number of hops, time delay or number of packets in
and each arc as communication link. To find a the queue.
Chapter 10.indd 499 4/9/2015 10:08:39 AM

Issues: Step 2: Each node periodically sends its own distance

(a) Slowness in converging to the right answer and this vector (DV) to neighbours. When node A receives DV
problem is well known as count to infinity (can be from neighbour B, it keeps it and updates its own DV
solved using split-horizon algorithm). as follows:
(b) Line bandwidth should be a metric when choosing root. DA (B) = min v { C(x, v) + Dv (B)}
Whenever a packet comes to a router, the neighbour- Node A updated table after receiving DV from node B
ing router will give their routing table and a new vector and node C is:
table is created at that router. DA(B) = min {2 + 0,7 + 1} = 2 and
DA(C) = min {2 + 1,7 + 0} = 3
Example 10.3
Node A table
Consider a network shown in Fig. 10.4:
Cost to
2 B
A B C
A A 0 2 3
1
From B 2 0 1
7
C 7 1 0
C
Node B table
Figure 10.4
Step 1: Initialise cost of direct links and set to ∞cost Cost to
from neighbours. A B C
Node A table
A 0 2 7
Cost to
From B 2 0 1
A C
A 0 2 7 C 7 1 0
B ∞ ∞ ∞
From
Node C table
C ∞ ∞ ∞
Cost to
Node B table A B C
Cost to A 0 2 7
A B C From B 3 0 1
A ∞ ∞ ∞ C 3 1 0
From B 2 0 1
∞ ∞ ∞
Step 3: In similar fashion, algorithm proceeds until all
C nodes have updated tables.
Node C table Node A table
Cost to Cost to
A B C A B C
A ∞ ∞ ∞ A 0 2 3
From B ∞ ∞ ∞ From B 2 0 1
C 7 1 0 C 3 1 0
Chapter 10.indd 500 4/9/2015 10:08:40 AM

10.6 NETWORK LAYER PROTOCOLS 501
Node B table 10.6 NETWORK LAYER PROTOCOLS

Cost to
A B C In the Internet model, or the TCP/IP suite, there are
five main network layer protocols: ARP, RARP, IP,
A 0 2 3 ICMP and IGMP (Fig. 10.5).
From B 2 0 1
IGMP ICMP
C 3 1 0
IP
Node C table
ARP RARP
Cost to
A B C Figure 10.5 | Network protocols.
A 0 2 3 The main protocol in this layer is IP. It is responsible for
host to host delivery of packets from a source to destina-
From B 2 0 1 tion. IP needs services of other protocols for better net-
C 3 1 0 work performance. IP needs ARP to find MAC address
of the next hop. As IP is an unreliable protocol, it needs
ICMP (Internet Control Message Protocol) to handle
unusual situations and errors. IGMP (Internet Group
3. Link state routing: This is simply a modern Message Protocol) is used for multicast delivery.
replacement of distance vector routing. The steps
of the algorithm are as follows: 10.6.1 Internet Protocol (IPv4)
•• Each router discovers the neighbours for their
network addresses. Internet Protocol is a layer-3 protocol of network layer
•• Measure delay or cost to each of these neighbours. of OSI model. It takes data segments from layer-4 trans-
•• Construct a packet including network address port layer and divides it into packets. Thus, it encapsu-
and delays of all neighbours. lates data units received from the above layer and adds
•• Send it to all routers. its own header information (Fig. 10.6).
•• Find the shortest path to all routers (Dijkstra’s Maximum size of IP header = 60 bytes
algorithm can be used). Minimum size of IP header = 20 bytes
If the size of header is 32 bytes in IP, calculate the
Example 10.4 number of option bytes.
OSPF protocol (used in Internet) uses the link state algo- Option bytes = 32 − 20 = 12 bytes
rithm: IS-IS (intermediate system—intermediate system) IP header details are as follows:
is another example used in Internet backbones and in
some digital cellular systems. 1. Version: Version number of Internet Protocol is 4
(e.g. IPv4).
2. IHL: Internet header length indicates the size of
4. Hierarchical routing: In the situation of tele- header available in the packet.
phone networks, all above routing algorithms fail 3. TOS: Type of service that is provided by the router
because the size of the routing table is too large to the packets such as minimum delay or cost.
here. In this routing, routers are divided (placed) 4. Total length: Length of the entire IP packet
into different regions. A router will have the knowl- (including IP header and IP payload).
edge of other routers in its own region but unaware 5. Identification: If IP packet size is greater than
about the internal structure of other regions. This the maximum transmission unit (MTU), it has
will reduce the size of the routing table. to be fragmented during the transmission by the
5. Broadcast and multicasting routing: Broad router, then all the fragments of the packet contain
casting means sending packets to all other hosts in same identification number to identify original IP
the network, whereas multicasting refers to send- packet they belong to.
ing packets to a specific group or a fixed number 6. Flags: If IP packet is too large, these `flags’ tell if they
of hosts. can be fragmented or not. In 3-bit flag, the MSB is
Chapter 10.indd 501 4/9/2015 10:08:40 AM

IP Header Data
20−60 bytes
65535 bytes
Version IHL TOS Total length

(4 bits) (4 (8 bits) (16 bits)
bits)
Identification Flags Fragmentation offset

(Fragment ID) (3 bits) (13 bits)
(16 bits)
Time to live Protocol Header checksum

(8 bits) (8 bits) (16 bits)
32-Bit source address
32-Bit destination address
Options
(if any, variable length, padded with 0’s, 40 bytes maximum length)
Figure 10.6 | IPv4 packet header.
always set to `0’. The next bit is DF (do not fragment). 13. Options: These options may contain values for
If DF = 0, fragmentation can be done if required and various options such as strict source routing, secu-
buffered at the router of the receiver until all frag- rity, record route, time stamp, etc.
ment comes, and if DF = 1, fragmentation should
not be done. The third bit is MF (more fragment). If Problem 10.4: If the total length bits are 0000000000
MF = 1, then datagram is not the last fragment. If 111111 and header length is 1001, calculate the size of
MF = 0, then datagram is the last fragment. the packet, header and payload.
7. Fragment offset: This offset tells the exact
position of the fragment in the original IP packet. Solution:
8. Time to live: It controls the maximum number of (a) Packet size = Decimal equivalent of 00000000
routers visited by the datagram. To avoid looping 00111111 = 63 bytes
in the network, every datagram is sent with some- (b) Header size (1001) = 9 × 4 = 36 bytes
time-to-live(TTL) value set. Each router receiv- (c) Packet size = Header + Payload
ing the datagram decreases TTL by 1 and when it
becomes 0, the datagram is discarded. Payload = Packet size − Header = 63 − 36
9. Protocol: It tells the higher-level protocol which = 27 bytes
uses the service of the IP layer. For example, pro-
tocol number for ICMP is 1, IGMP is 2, TCP is 6
and UDP is 17. Problem 10.5: Suppose a router receives an IP packet
10. Header checksum: This field stores checksum value containing 600 data bytes and has to follow a packet
of the entire header excluding data which is used to maximum transferable unit of 200 bytes. Assume
check if the packet has been received error-free. that the IP header is 20 bytes long; specify the
11. Source address: 32-Bit address of the sender of relevant values in each fragment header.
the packet. Solution: IP packet = 600 bytes
12. Destination address: 32-Bit address of the MTU = 200 bytes
receiver of the packet.
Chapter 10.indd 502 4/9/2015 10:08:40 AM

10.6 NETWORK LAYER PROTOCOLS 503
C, D and E. The first few bits reveal the class of address

IP header = 20 bytes
when the address is in binary and first byte reveals the
Maximum possible data length per fragment = 200 − class of address when the address is in dotted decimal
20 = 180 bytes notation. This architecture is called classful addressing
Data length of each fragment must be a multiple of 8 (Table 10.4).
bytes so 22 × 8 = 176 bytes
Data packet must be divided into the following four Table 10.4 | Classful addressing architecture of IPv4
frames: Leading Value Starting Ending
(176 + 20) + (176 + 20) + (176 + 20) + (72 + 20) = Bits Range Address Address
680
Class A 0 0—126 0.0.0.0 127.255.255.255
Length ID MF Fragment Class B 10 128—191 128.0.0.0 191.255.255.255
Offset
Class C 110 192—223 192.0.0.0 223.255.255.255
Original 620 bytes X 0 0 Class D 1110 224—239 224.0.0.0 239.255.255.255
Packet
Class E 1111 240+ 240.0.0.0 255.255.255.255
Fragment 1 196 bytes Z 1 0
Note: 127 reserved for loopback address.
Fragment 2 196 bytes Z 1 22
Fragment 3 196 bytes Z 1 44 The addresses of class A, B, C are unicast, class D
Fragment 4 92 bytes Z 0 66 addresses are multicast and class E addresses are
reserved. The IP address in class A, B and C is divided
into netid and hostid. In class A, one byte defines the
10.6.1.1 IPv4 Addresses netid and the other three bytes define the hostid. In class
B, two bytes define the netid, while the other two bytes
An IPv4 address is a 32-bit address that can find the define the hostid. In class C, three bytes define the netid
device on the Internet uniquely and universally. These and one byte defines the hostid.
addresses are unique, that is, these define only one connec-
Mask: The length of the netid and hostid (in bits) is
tion by the device to the Internet. The total number of
predetermined in classful addressing but we can also use
addresses used by this protocol which is called as address
a mask (also called the default mask) which is a 32-bit
space is 2n where n is the total number of bits. As IPv4 uses
number made of contiguous 1’s. The subnet mask is
32-bit addresses, the address space of this protocol is 232.
compared to the IP address from left to right, bit for bit.
In IPv4, addresses are 32-bit binary numbers. The 1’s in the subnet mask represent the network por-
However, for ease of use of people, theses binary patters tion and 0’s represent the host portion. The subnet mask
are represented as dotted decimals. Therefore, there are is created by placing a binary 1 in each bit position that
two notations available to denote IPv4 addresses: binary represents the network portion and placing a binary 0 in
and dotted decimal. each bit position that represents the host portion.
1. Binary notation: In this notation, IPv4 addresses The subnet mask assigned along with IP address sig-
are represented as 32 bits where each octet is said nifies which part of the IP address is network and which
to be a byte. Therefore, the IPv4 address is usually part is host.
said to be the 4-byte address. The example for this
There is a flaw in this type of architecture, that is,
notation is as follows:
each class is divided into fixed number of blocks, and a
01111101 10000011 00000110 00000001 lot of addresses are wasted as blocks A and B addresses
2. Dotted-decimal notation: In this notation, each are too large to consume, block C addresses are small in
byte (8 bits) of 32-bit binary address, known as number, class D addresses are reserved for multicasting
octet is separated with a dot, and then the binary and class E addresses are reserved for future, which is
number is converted into its decimal equivalent. another wastage of addresses. Therefore, these address
The example for this notation is as follows: scheme architecture is almost obsolete and leads to an
introduction of classless addressing scheme as if there are
11000000 10101000 00001010 00001010
no address classes.
is equivalent to 192.168.10.10
Classful Addressing Classless Addressing

The architecture used by IPv4 is classful addressing In classless addressing, the addresses are granted in a
where the addresses are divided into five classes: A, B, block and the number of addresses in the block depends
Chapter 10.indd 503 4/9/2015 10:08:40 AM

upon the addresses needed by the entity. The addresses

Problem 10.8: In class B, if subnet mask is
in the block must be contiguous, the first address must
255.255.240.0, find the number of subnets and host
be evenly divisible by the total number of addresses
in each subnet.
and the number of addresses in a block must be power
of 2. Mask in classless addressing can take the value in
Solution:
the range of 0 to 32. Classless Inter-Domain Routing
provides the flexibility of borrowing bits of host part of 11111111 11111111 1111 0000000
the IP address and using them as smaller sub-networks 0000
called subnet. This process is known as subnetting.
The addresses can be defined in IPv4 as a.b.c.d/n, where Netid Subnet Host bits
a.b.c.d defines one address of the block. bits
1. The first address in the block can be found by set- Number of subnets = 2 − 2 = 14 subnets
4
ting the rightmost 32 −n bits to Os in the binary Number of hosts in each subnet = 212 − 2
notation.
2. The last address in the block can be found by set- = 4094 subnets
ting the 32 −n rightmost bits in the binary nota-
tion of the address to Is.
3. The number of addresses in the block is the dif- Problem 10.9: If IP address of a system is
ference between the first and the last address,that 199.11.171.189 and subnet mask is 255.255.255.224,
is, 232−n. calculate the subnet id.
Solution:
Problem 10.6: Determine the subnet identifier, broad- IP address 199.11.171.189
cast address and number of valid host addresses Subnet mask 255.255.255.224
having a host with IP address of 196.142.4.29/24.
Subnet id 199.11.171.160
Solution:
160 is equivalent to 101 00000
Subnet identifier: 196.142.4.0
Subnet Host
Broadcast address: 196.142.4.255
Bits Bits
Number of subnets = 2 − 2 = 6
Number of valid host addresses: 254 3

Number of hosts = 25− 2 = 30
The subnet mask of /24 in CIDR notation corresponds

to 255.255.255.0. The above subnet mask has last
eight bits set to zero (32 − 24 = 8), which means that First subnet id is 199.11.171.32 (001 00000).
we have 28 IP addresses available. Total number of Second subnet id is 199.11.171.64 (010 00000).
valid hosts in a network is obtained by subtracting
Last subnet id is 199.11.171.192 (110 00000).
two addresses: subnet address and broadcast address
28− 2 = 256 − 2 = 254. First host of first subnet is 199.11.171.33 (001 00001).
Last host of last subnet is 199.11.171.222 (110 11110).
Problem 10.7: If the IP address of a system is Supernetting is a part of classless addressing. In class-
131.121.61.189, calculate the netid, first host, last less addressing, the addresses should be contiguous in a
host and directed broadcast address. block. The first address should be exactly divisible by
Solution: the number of addresses in a block. In supernet, bits are
borrowed from netid.
IP address (class B) 131.121.61.189
Rules of supernetting:
Net mask 255.255.0.0
1. The number of blocks must be power of 2.
Net id 131.121.0.0
2. The blocks must be continuous in the address
First host 131.121.0.1 space.
Directed broadcast address 131.121.255.255 3. The third octet of the first address in the super-
block must be exactly divisible by the number of
(All host bits 1)
blocks.
Chapter 10.indd 504 4/9/2015 10:08:40 AM

10.8 CONGESTION 505
Problem 10.10: A company needs 1000 addresses, Table 10.5 | Different types defined
which of the following set of class C block can be used for ICMP messages
to form a supernet? Category Type Message
(a) 198.47.32.0, 198.47.33.0, 198.47.34.0 Error 3 Destination
(b) 198.47.31.0, 198.47.32.0, 198.47.33.0, 198.47.34.0 Reporting Unreachable
(c) 198.47.32.0, 198.47.42.0, 198.47.52.0, 198.47.62.0 Message
(d) 198.47.32.0, 198.47.33.0, 198.47.34.0, 198.47.35.0
4 Source Quench
Solution: (d) 11 Time Exceeded
(a) Not in power of 2 or blocks are 3. 12 Parameter Problem
(b) Third octet of first address is not divisible by 4. 5 Redirect Message
(c) Blocks are not contiguous.
Query 0 Echo Reply
Message
8 Echo Request
Problem 10.11: Which of the following can be 9 Router solicitation
the beginning address of a block that contains 16 10 Router advertisement
addresses?
13 Timestamp Request
(a) 205.16.37.32
14 Timestamp Reply
(b) 190.16.42.44
(c) 17.17.33.82 17 Address Mask Request
(d) 123.45.24.52 18 Address Mask Reply
Solution:
(a) 32 is divisible by 16. 10.7 LAYER 4: TRANSPORT LAYER
Although the reliability of the network layer is undoubt-

10.6.2 ICMP
ful, the transport layer has many responsibilities to carry
As IP is a connectionless and unreliable protocol, it out the following:
cannot report errors, so it takes help of ICMP to com- 1. Managing connections and timers
municate updates or error information to other interme- 2. Allowing reliable, connection-oriented byte stream
diate routers, devices or hosts. from one end to other end
Each ICMP message contains three fields: Type, Code 3. Multiplexing
and Checksum (Fig. 10.7). The Type field identifies the 4. Addressing
ICMP message, the Code field provides further informa- 5. Performing segmentation
tion about the associated Type field and the Checksum 6. Packetizing
field verifies the integrity of the message. 7. Handling error control and variable sized sliding
window for flow control
8. Allocating bandwidth with congestion control
Issues:
Type Code Checksum
(a) Headers, error detection, reliable communication
Rest of header (b) Communication between processes (running on
machines on possibly different networks)
Data section
Figure 10.7 | ICMP message.

10.8 CONGESTION
Extra options are specified in the rest of the header. Congestion occurs when packets overload the subnet
ICMP places the message to be sent in the data section. (means the number of packets sent to the network is greater
Different types defined for ICMP messages are shown in than the capacity of the network), which results in perfor-
Table 10.5. mance degrade. The following are causes for congestion:
Chapter 10.indd 505 4/9/2015 10:08:41 AM

1. A sudden stream of packet from various sources that is, a fast sender does not send at a rate faster than
reaching the same destination. the rate at which the receiver receives.
2. Slow links.
3. Slow processor.
4. Suppose an intermediate router has no free buffer, 10.9 USER DATAGRAM PROTOCOL AND
it will discard the packet. As the ender will not TRANSMISSION CONTROL PROTOCOL
receive any acknowledgement, it will resend the
packet and hence congestion will be there.
Congestion is unavoidable, but it is necessary to control
10.9.1 User Datagram Protocol
it. It can be handled with the following approach:
UDP, or User Datagram Protocol, is an unreliable and
1. Congestion information may be forwarded to connectionless protocol used by applications that trans-
the concerned node so that it is possible to limit mit small amount of data at one time and do not require
senders (prevent one sender from overflowing the receipt of acknowledgement of data, for example, audio
receiver) or reroute the packets. or video broadcasting (Fig. 10.8).
2. Resource availability may reduce the congestion.
3. Prevent additional packets from entering the con-
gestion region. Source port (16 bits) Destination port (16 bits)
To control congestion in network traffic, leaky bucket Length (16 bits) Checksum (16 bits)
algorithm is used. This algorithm shapes the bursty traf-
fic into fixed rate traffic. It does so by averaging the data Data
Figure 10.8 | UDP packet.

rate and dropping the packets if bucket is full.
10.8.1 Congestion Versus Flow Control

10.9.2 Transmission Control Protocol
Congestion control ensures the ability to carry the offered
traffic by any subnet. The major entities that effect the TCP, or Transmission Control Protocol, provides a
congestion are behaviour of hosts, routers as well as the connection-oriented, reliable stream delivery through the
factors that reduce the carrying capacity. use of sequenced acknowledgement with retransmission
Flow control makes it sure that there is not much dif- of packets when necessary. The TCP header structure is
ference between the sending and receiving packet rate, shown in Fig. 10.9.
Source port (16 bits) Destination port (16 bits)
Sequence number(32 bits)
Acknowledgement number (32 bits)
HLEN Reserved U A P R S F Window (16 bits)
(4 bits) (6 bits)
Checksum (16 bits) Urgent pointer (16 bits)
Option + padding
Data
TCP - header structure
Figure 10.9 | TCP packet.
Chapter 10.indd 506 4/9/2015 10:08:42 AM

10.11 LAYER 5: SESSION LAYER 507
In TCP, sequence number is attached for every byte in 3. While the server is waiting, a client instantiates a
the segment. The initial sequence number for the first Socket object, which specifies the server name and
data byte will be a random number that is generated by port number for the connection.
a random number generator in the range of 0 to 232 − 1. 4. The constructor of the Socket class attempts to
Acknowledgement number will always be the sequence connect client to the specified server and port
number of the next expected data. The control bits and number. Once the communication is established,
their description is given in Table 10.6. the client has a Socket object for communicating
Table 10.6 | Control bits
with the server.
5. The accept() method returns a reference to a new
Control Bits Description socket on the server that is connected to the socket
of the client.
U (URG) Urgent pointer field significant
A (ACK) Acknowledgement field
TCP Server
significant
socket( )
P (PSH) Push function
R (RST) Reset the connection Well-known bind( )
port
S (SYN) Synchronize sequence numbers listen( )
F (FIN) No more data from sender TCP Client
accept( )
socket( )
Block until connection
10.10 SOCKETS connect( ) Connection establishment from client
(TCP three-way handshake)
write( ) Data (request)
A socket is an endpoint for communication between client read( )
process and server process across a network. A socket Process request
address is the combination of an IP address and a port
Data (reply) write( )
number. This address is used to send data packet to a read( )
particular process running on a machine. When two pro-
End-of-file notification read( )
grams are executed, a client process and a server process close( )
are created, and these processes communicate with each
close( )
other by reading from, and writing to, sockets.
There are few system calls, such as those given below: Figure 10.10 | Socket creation.
1. Socket() returns a socket descriptor (like file
descriptor), which is an integer value.
2. Bind() binds an address to a socket descriptor cre-
10.11 LAYER 5: SESSION LAYER
ated by socket.
3. Listen() announces willingness to accept The services provided by session layer are as follows:
connections.
4. Accept() blocks the caller until a connection 1. Establishes, manages and terminates a communi-
attempt arrives. cation session with remote systems
5. Connect() actively attempts to establish a 2. Allows two machines to enter into a dialog (com-
connection. munication may be half duplex or full duplex)
6. Send() sends some data over the connection. 3. Adds checkpoints or synchronisation points to a
7. Receive() receives some data from the connection. data stream.
8. Close() releases the connection. 4. Groups several user-level connections into a single
`session’.
The following steps (as shown in Figure 10.10) occur
when establishing a TCP connection between two com- Some protocol suites do not include the session layer.
puters using sockets: Checkpoint: If we need a file of 1000 pages, it is sug-
1. The server instantiates a ServerSocket object, gested to insert checkpoints after every 100 pages to
which denotes the port number on which the com- ensure that each 100-page unit is received and acknowl-
munication is to take place. edged independently. Meanwhile, if a crash occurs during
2. The server invokes accept() method of the the transmission of page 324, the only pages that need
ServerSocket class, which waits until a client con- to be resent after system recovery are pages 301 to 324.
nects to the server at the given port. The pages from 1 to 300 need not be resent.
Chapter 10.indd 507 4/9/2015 10:08:42 AM

10.12 LAYER 6: PRESENTATION and inverse domains. Generic domain specifies the
LAYER registered hosts according to their generic behaviour.
For example, com, org, edu, gov, net, etc. The coun-
try domain uses two character country abbrevia-
The following are the major concerns of presentation tions, for example, ìn’ for India. The inverse domain
layer: is used to map an address to a name (Fig. 10.11).
1. Syntax and semantics of the information exchanged
between two systems.
2. Support to different encoding schemes used by dif- Application
ferent machines. It converts the sender-dependent
format into a common format and the receiver con- Telnet DHCP
verts it back to the receiver-dependent format. HTTP RIP
3. Data encryption–to ensure privacy, sensitive infor- FTP OSPF
mation should be encrypted. Information encrypted
SMTP TFTP
to some other form is unreadable to others.
4. Data compression–to reduce the size of informa- SNMP
tion to carry. In the case of multimedia, for exam- BOOTP
ple, text, audio and video, compression is a very Transport
useful tool.
TCP UDP
Note: Many protocol suites do not include a presentation
layer. Internet
ARP ICMP
RARP IGMP
10.13 LAYER 7: APPLICATION LAYER
Network interface
Ethernet
The major responsibility of application layer is to imple- Token ring
ment communication between two applications of the
FDDI ISDN
same type. There is a common misconception that every
user application runs on application layer, but it runs X.25 PPP
only on those applications which interact with the com-
munication system. For example, FTP, HTTP, SMTP/ Figure 10.11 | TCP/IP protocols.
POP3/IMAP (email)are all application layer protocols,
but a designing software or text editor cannot be consid- 3. Simple Mail Transfer Protocol (SMTP):
ered as an application layer protocol.The following are SMTP is a message transfer agent (MTA), which
popular application layer protocols: is used to transfer mail. A system should have
1. Internet Control Message Protocol (ICMP): client MTA for sending a mail and server MTA
ICMP provides error reporting and query manage- for receiving a mail. SMTP is used two times,
ment mechanism for host, which lacks in IP. ICMP between the sender and the sender’s mail server
messages are of two types: error-reporting messages and between the two mail servers. The main task
and query messages. The header of ICMP is of 8 of SMTP is to push the message from the client to
bytes and data section is of variable size. First byte the server.
is ICMP type, second byte is code, the next two 4. Post Office Protocol (POP): POP3 (version 3)
bytes specify the checksum field and rest of the is a message access protocol which is used to extract
header is specific for each message type. the message for client to the server. It has two
2. Domain Name System (DNS): DNS is a sup- modes: the delete mode and the keep mode. In the
porting program, which is used by other programs delete mode, the mail is deleted from the mailbox
used by the users, such as email. DNS is a client— after each retrieval, while in the keep mode, the
server program used to find the IP address of an mail remains in the mailbox after retrieval.
email recipient. In DNS client—server application, a 5. File Transfer Protocol (FTP): It is used for
sender sends an email to the email address of the transferring files from one system to another. FTP
receiver. The DNS client sends the request to the establishes two connections between hosts, one for
DNS server to map the email address to IP address. data transfer and the other for control information.
DNS has three different domains: generic, country FTP uses TCP Port 21 for the control connection
Chapter 10.indd 508 4/9/2015 10:08:42 AM

10.14 DEVICES 509
and TCP Port 20 for the data connection. The FTP Thus, repeater connects two segments of network
client has three components: user interface, client cable. It works at the physical layer of the OSI
control process and the client data transfer process. model.
The server has two components: the server control 2. Bridge: There is a limited number of stations
process and the server data transfer process. The that can be connected with a single LAN. So, a
control connection remains open for entire FTP bridge is used to connect multiple LANs of the
session, whereasthe data connection remains open same type. It operates on a physical layer and data
for each file transfer. link layer. A bridge checks the physical address
6. Hypertext Transfer Protocol (HTTP): HTTP contained in the frame. It also uses table for fil-
is used to access data on the World Wide Web tering frames. It partitions the collision so that
(WWW). It works as a combination of FTP and performance increases.
SMTP. Unlike FTP, HTTP does not have any con- 3. Hub: It is a network device which is used to con-
trol connection and uses only one TCP connection. nect various computers together. Hub is the central
Unlike SMTP, it does not store and then forward connection for all the computers, which connect
the messages. It immediately sends the messages. through Ethernet. Hub can receive and send the
HTTP uses a TCP Port 80. information but cannot perform both tasks at the
same time. This makes it slower than a switch. It
is less expensive and less complex.
10.14 DEVICES 4. Switch: Switch is a small network device used
to connect one or more computers through LAN.
It is mostly used in home networks. Switches and
The devices used for internetworking at different layers hubs are used in the same network. Hubs increase
are specified in Fig. 10.12. the network by providing more ports, and switches
divide the whole network into smaller networks.
Switching reduces the amount of unnecessary traf-
fic when every port sends the same information.
Application Switch also reduces the possibility of collision in
network.
5. Router: A router is a networking device which
takes packets from one network and after analy-
Content Service Switch sis sends that packet to another network. When a
Layer 4-7 Switches data packet comes to the router, the router reads
the destination address of the packet and sends it
to the respective router which contains that desti-
nation address (listed in its routing table). A router
is more intelligent than a hub because a hub only
sends the information between the devices but the
Network router analyses the packet and then forwards it
Router, Layer-3 Switch to the other network. It controls the traffic on the
network.
6. Gateway: It is a networking system capable of
Data Link
interconnecting one or more networks that has dif-
Switch, Bridge
ferent base protocol. Gateway serves as an entry
Physical and exit point. Gateway, sometime called as proto-
Hub, Repeater col converter, is used in different layers. For exam-
ple, a gateway can be used to convert a TCP/IP
Figure 10.12 | Devices.

packet to a NetWare IPX packet.
1. Repeaters: It is an electronic device which can

receive the weak signal and retransmit it with higher Problem 10.12: If the capacity of router is 1 MB,
speed. For example, if the LAN is connected for a data output rate is 8 Mbps. Tokens are generated at
long distance, then to cover a distance the signal 6 Mbps. Calculate the time burst traffic is routed.
is sent to a repeater which is attached with two Solution: C + r⋅S = M ⋅S
LANs and retransmits the signal to a higher level.
Chapter 10.indd 509 4/9/2015 10:08:43 AM

10.15.1.1 Symmetric Cryptography
where M = output rate, C = capacity of router,
S = time of burst traffic and r = token rate
In symmetric key cryptography, a single shared key is
used for both encryption and decryption as shown in
10 + 6 × 10 ⋅ S = (8 × 10 ) ⋅ S
6 6 6
Fig. 10.13. Symmetric cryptographic primitives use block
106 (1 + 6S ) = (8 × 106 ) ⋅ S ciphers, stream ciphers, cryptographic hash functions,
1 + 6S = 8S and message authentication codes (MACs). Block cipher
2S = 1 Þ S = 0.5 s
uses a deterministic algorithm and operates on a block
(fixed length of bits) with unaltered transformation.
Stream cipher encrypts each bit individually to generate
cipher text. Hash functions or one-way hash functions
are used to map an arbitrary-length message string to
10.15 NETWORK SECURITY fixed-size message string. The final value is called hash
value.
Network security is an activity designed to protect the Key Key

network and data in terms of usability, reliability, integ- plaintext ciphertext plaintext
Figure 10.13 | Symmetric key cryptography.

rity and safety. It targets a variety of threats and pre-
vents them to enter into the network. It is accomplished
through hardware and software, for example, firewall,
The security of the symmetric key cryptography lies in
anti-virus and anti-spyware, cryptography, intrusion
the secrecy of the shared symmetric key. If the adversary
prevention systems (IPS) used to identify fast-spreading
captures the shared secret key, then it affects both con-
threats such as zero-day attacks,and virtualprivate net-
fidentiality and authentication of the message. Examples
works (VPNs) used to provide secure remote access.
of symmetric key cryptography are Twofish, Serpent,
Network security components are as follows: AES (Rijndael), Blowfish, RC4, RC5, 3DES, IDEA,
1. Confidentiality: It ensures the concealment of SEAL, SNOW, etc.
data to unauthorised individuals.
2. Integrity: It ensures that information is changed 10.15.1.2 Asymmetric Cryptography
in a specified and authorised manner. There is no
change in content or source by an unintended user. In asymmetric cryptography, a private key is used for
3. Availability: It ensures that systems are avail- the decryption of a message while a public key is used
able for the authorised users. for the encryption of the message as shown in Fig. 10.14.
The private key needs to be kept confidential while the
The trio form the term CIA (Confidentiality, Integrity public key can be published freely. Asymmetric cryp-
and Availability). tography is also known as public key cryptography
(PKC). PKC was introduced first by Diffie and Hellman
10.15.1 Basic Concepts in Cryptography in 1976. Public key algorithms are based on mathemati-
cal functions rather than substitution and transposi-
Cryptography is the science of providing secure commu- tion as in symmetric key cryptography. Examples of
nication over insecure channels. Cryptography consists asymmetric key cryptography are Diffie-Hellman, RSA,
of two operations: encryption and decryption. Merkle-Hellman, Rabin, McEliece, El Gamal, Ellliptic
Encryption is the process in which data is ciphering so curves, etc.
that only the intended recipient can know the message.
Decryption is the process of deciphering the message. Key 1 Key 2
plaintext ciphertext plaintext
Basically, encryption and decryption are two functions
of a cryptographic algorithm mathematically related to Figure 10.14 | Asymmetric key cryptography.
each other. A cryptographic algorithm is widely known,
but a key, which is used for the encryption/decryption, RSA Algorithm
is kept secret. The RSA algorithm, developed in 1977 by Riverst,
Cryptography is of two types: symmetric cryptogra- Shamir, Adelman at MIT, provides encryption and digi-
phy and asymmetric cryptography. Both the approaches tal signatures. RSA is based on factoring of large num-
have their own pros and cons. bers, which is not known to be NP-complete.
Chapter 10.indd 510 4/9/2015 10:08:44 AM

10.15 NETWORK SECURITY 511
Encryption and decryption are as follows for a plain- 2. Signature:

text block M and cipher textblock: •• For each message M, generates random k
C = M e mod n •• Computes r = (gk mod p) mod q
•• Computes s = k−1(H(M ) + xr) mod q
M = C d mod n = (M e )d mod n = M ed mod n •• Signature is (r, s)
where n and e are known to both the sender and receiver, 3. Verification:
butd is only known to the receiver. Thus, the public key •• Computes w = s−1 mod q , u1 = H(M )w modq
is P = {n, e} and the private key S = {d, n}. It is impos- •• Computes u2 = rw mod q, v =(g u1 y u2 mod p)mod q
sible to find d given e and n. •• Verify if v = r
The detailed RSA algorithm is as follows: 4. Correctness:
1. Key generation:
•• Choose two prime numbers p and q, keep them v = (g u1 y u2 mod p) mod q
secret.
•• Compute n = pq, n is public. = (g H (M )w mod q y rw mod q mod p) mod q
•• Calculate f(n) = (p − 1)(q − 1)
•• Choose e with gcd(f(n), e) = 1;1 < e < f(n), = (g H (M )w mod q y xrw mod q mod p) mod q
which is also public.
•• Compute d = e−1 mod f(n) and keep it private. = (g(H (M )w + xrw) mod q mod p) mod q
•• The private key consists of S = {d, n} and pub-
lic key consists of P = {n, e}. = (g(H (M )+ xr)w mod q mod p) mod q
−1
2. Encryption:
= (g(H (M )+ xr)k(H (M )+ xr) mod q
mod p) mod q
Plaintext M < n
Cipher text C = Me mod n = (gk mod p) mod q
3. Decryption: =r
Ciphertext C
Plaintext M = Cd mod n
10.15.3 Firewall
10.15.2 Digital Signature
It is a device that filters access to the protected
It provides the authenticity of the origin of information network from the outsider network. It is an inte-
to the user and verifies the information is intact. Hence, grated collection of security measures, which are
it provides authentication and data integrity. It also designed to prevent unauthorised electronic access
provides non-repudiation, which ensures that the sender to a network system. It has a predefined set of rules,
cannot deny the origin of information. It is based on the which can protect private network from unauthor-
public key cryptography concept. ised access by filtering incoming or outgoing traf-
fic. These predefined set of rules are called firewall
10.15.2.1 Digital Standard Algorithm policies.
Digital Standard Algorithm (DSA) is based on the dif-

ficulty of discrete logarithm problem (DLP). It is also
10.15.3.1 Functions of Firewalls
based on Elgamal and Schnorr system.
1. Examining the packet header and filtering
DSA involves the following four steps: 2. Verifying the IP address or the port
1. Key generation: 3. Granting and denying access
•• Global Public Components: p is a prime num- Packets can be filtered on the basis of the following
ber with 512-1024 bits, q is a prime divisor criteria:
of (p −1) with 160 bits, g is an integer
1. Source IP address
g = h(q −1)/q mod p. 2. Destination IP address
•• Users Private Key: x is random integer less than q. 3. TCP/UDP source port
•• Users Public Key: y = g x mod p 4. TCP/UDP destination port
Chapter 10.indd 511 4/9/2015 10:08:46 AM

10.15.3.2 Types of Firewalls 3. Application gateway: It works as a proxy.

It is made up of bastion hosts, which run special
1. Packet Filter (stateless): It is router-based software to act as a proxy server. It inspects the
filters. It does not use any context for filtering contents of the traffic, blocking inappropriate
the packets. Individual packets are accepted or contents, such as websites, viruses, vulnerabili-
rejected. It has following limitations: (i) filter ties, etc.
rules are hard to set up, (ii) inadequate primitives,
(iii) hard to manage access to RPC-based services. 10.15.3.3 Limitations of Firewall
2. Stateful Filter: It maintains records of all connec-
tions passing through it. It determines if a packet 1. It cannot protect against attacks that bypass the
is either the start of a new connection, a part of firewall.
an existing connection or is an invalid packet. It 2. It cannot protect fully against internal threats.
maintains tables of each active connection, includ- 3. It cannot provide protection against malicious
ing the IP addresses, ports and sequence numbers code problems such as viruses and Trojan horses,
of packets. although some are capable of scanning the code.
IMPORTANT FORMULAS
Application Protocol Message size (bits)

3. Transmission time =
Bandwidth (bits/s)
SMTP TCP
Distance
4. Propagation time =
TELNET TCP Velocity
SSH TCP 5. For stop-and-wait ARQ
HTTP TCP Message size

Transmission time =
Bandwidth
DNS TCP & UDP
PING ICMP Distance of the link

Propagation delay =
Velocity
1. In early token release, Link utilisation of sender or throughput is given by
Throughput for single station or N stations
h=
Transmission time
Data
= Transmission time + 2× Propagation delay
Transmission time
+ (Ring latency Number of stations) 6. Pure ALOHA
2. In delayed token release, Throughput (S) = G × e−2G
Throughput for single station Vulnerable time = 2 × Tfr
Data Smax= 18.4%
=
Transmission time + Ring latency
7. Slotted ALOHA
Throughput (S) = G × e−G
+ (Ring latency/Number off stations)
Throughput for N stations

Vulnerable time = Tfr
Data
=
Ring latency + (Ring latency Number of stations) Smax= 36.8%
Chapter 10.indd 512 4/9/2015 10:08:48 AM

SOLVED EXAMPLES 513
SOLVED EXAMPLES
1. Which layer is responsible for delivery from process (c) Time stamping only
to process? (d) Time stamping, loose source routing, strict
source routing
(a) Network (b) Transport
(c) Physical (d) Data link Solution: It may contain values for various
options, such as strict source routing, security,
Solution: Transport layer is responsible for end-
record route, time stamp, etc.
to-end process communication.
Ans. (d)
Ans. (b)
7. In a fully connected mesh network with d devices
and c connections, there are ________ physical
2. The minimum size of an Ethernet frame is
(a) 18 bytes (b) 64 bytes channels to link all devices.
(c) 46 bytes (d) 56 bytes
(a) d*(d − 1)/2 (b) d*(d + 1)/2*c
Solution: Ethernet header = 18 Bytes [Dest. Mac (c) (2*d + 2*c)/2 (d) 2*(d + 1) + c
(6) + Source Mac (6) + Length (2) + CRC (4)]
Solution: The total number of wired links required
Minimum Data Portion = 46 Bytes and Minimum
to establish a fully connected mesh network of d
Ethernet Frame Size = 64 Bytes
nodes can be calculated as c = d (d − 1)/2.
Ans. (b) Ans. (a)
3. Using a 7-bit sequence number, what is the maxi- 8. In IPv4 header, the ______ field is used to determine to
mum size (in bits) of the sender and receiver which datagram a newly arrived fragment belongs to.
window using stop-and-wait protocols?
(a) Identification (b) Datagram-id
(a) 1 and 1 (b) 1 and 7 (c) Fragment offset (d) Time to live
(c) 7 and 1 (d) 7 and 7
Solution: Identification field is used to identify
Solution: Stop-and-wait protocol uses 1 bit sequence. original IP packet the fragments belong to.
Ans. (a) Ans. (a)
4. Vulnerable time in CSMA is 9. Encryption and decryption is the responsibility
(a) It is double the transmission time as it includes of _______ layer.
the sensing time (a) Session (b) Data link
(b) Half of average frame transmission time as (c) Application (d) Network
collisions are less
(c) 2 × (the average frame transmission time) Solution: Application layer is responsible for the
(d) None of the above encryption and decryption processes.
Ans. (c)
Solution: Vulnerable time for CSMA is the prop-
agation time Tp needed for a signal to propagate 10. Maximum throughput of an ALOHA network is
from one end of the medium to the other. (a) 18.4% (b) 35.8% (c) 36.8% (d) 50%
Ans. (d)
Solution: When G = 1, the throughput is increased
5. Which class of IP addresses is used for multicasting? to the maximum value of 36.8%.
(a) Class E (b) Class C Ans. (c)
(c) Class A (d) Class D 11. A terminal multiplexer has six 1200 bps terminals and
Solution: Class D is used for multicasting. Refer `N’ 300 bps terminals connected to it. The outgoing
Table 10.4. line is 9600 bps. What is the maximum value of N?
Ans. (a) (a) 4 (b) 6 (c) 8 (d) 12
6. The options field of IPv4 is used for? Solution: Since, there are six 1200 bps terminals.
(a) Time stamping, strict source routing So, 6 × 1200 + n × 300 = 9600 ⇒ n = 8
(b) Loose source routing, strict source routing Ans. (c)
Chapter 10.indd 513 4/9/2015 10:08:49 AM

12. The total number of wired links required to establish 16. Consider the network with subnet mask
a fully connected mesh network of 9 nodes will be 153.224.0.0/13. Determine the last host address in
the network.
(a) 36 (b) 56 (c) 72 (d) 64
(a) 153.208.255.255
Solution: Total number of wired links required
(b) 153.224.255.254
to establish a fully connected mesh network of n
nodes can be calculated as c = n (n − 1) / 2
(c) 153.231.255.255
(d) 153.231.255.254
So, for 9 nodes, total links are 36.
Solution: Network id: 153.224.0.0
Ans. (a)
Subnet mask: 255.248.0.0
13. Considering a classful addressing, the IP address By doing XOR between network id and subnet
128.252.144.84 denotes mask, one gets network id= 153.231.0.0
We have 19 bits for host, so first host address is
(a) 0.0.0.0 as network ID and 128.252.252.84 as
153.224.0.1 and last host is 153.231.255.254
node ID
Ans. (d)
(b) 128.0.0.0 as network ID and 128.252.127.84 as
node ID 17. Consider a token ring LAN of length 12 km and
(c) 128.252.0.0 as network ID and 128.252.144.84 having 40 stations, signal propagation speed is
as node ID 8 ns/m and data rate is 100 Mbps. An average
(d) 128.252.144.0 as network ID and 128.252.144.84 frame contains 220 bytes. Delay occurred at each
as node ID station is equivalent to 10-bit delay. What is the
utilisation of token ring approximately?
Solution: The IP belongs to class B. The network
id is 128.252.0.0 and the node id is 128.252.144.84. (a) 10% (b) 12%
Ans. (c) (c) 15% (d) 21%
14. In Ethernet CSMA/CD, the special bit sequence Solution: Transmission time of a frame
transmitted by media access management for colli- 220 × 8
sion handling is called Tt = s = 17.6 µs
100 × 106
(a) Hamming code (b) CRC
Propagation time around the ring
(c) Jam (d) Preamble
Tp = 10 × 12000 ns = 120 µs
Solution: Hamming code is a set of error-correc-
tion code, which is used to detect and correct bit Delay at each station = 1/10 ms, so delay at 40 sta-
errors. CRC (cyclic redundancy check) is used to tions will be = 4 ms
detect data transmission errors. Now,
Preamble is used in network communications for Tt
synchronizing transmission time between systems. Utilisation (U) =
Tt + Tp + delay at all stations
Ans. (c)
17.6 17.6
= =
15. A Gateway operates at ______________ layers. 17.6 + 120 + 4 141.6
(a) A
ll layers except physical and application = 0.124 × 100 = 12.4% ≈ 12%
layer Ans. (b)
(b) All the seven layers
18. A CSMA/CD-based network has transmission rate
(c) Only on session, transport and network layers
120 Mbps, length 1 km and speed of signal is 109
(d) Same layers on which the switch and bridge
m/s. What should be the minimum frame size?
operates
(a) 120 B (b) 240 B
Solution: Hubs, repeaters(or active hubs) operate
(c) 400 B (d) 440 B
at physical layer (1); Bridges, Switches operates
at data link layer (2) ; Routers operate at net- Solution: We know that
work layer (3) ; content-switches (or web-switches Length of packet Distance
or application-switches) operates at layers 4 to 7. = 2×
Bandwidth Velocity
Gateway operates at all the seven layers—physical,
1000
data link, network, transport (4), session (5), pres- So, packet length = 2× ×120 ×106 ⇒ 240 bytes
entation (6), application (7). 109
Ans. (b) Ans. (b)
Chapter 10.indd 514 4/9/2015 10:08:51 AM

19. How many 8-bit characters can be transmitted per 22. In the IPv4 addressing format, the number of net-
second over a 7800 baud serial communication link works allowed in class C addresses is
using asynchronous mode of transmission with one
(a) 221 (b) 222
start bit, 8 data bits and 1 parity bit?
(c) 223 (d) 224
(a) 600 (b) 660
(c) 780 (d) 1200 Solution: In class C,
Net id = 24 bit
Solution: Total number of bits = 10 Host id = 8 bits
Modulation rate = 7800 baud Out of 24 bits 3 bits are reserved for representation
Numbers of 8-bit characters are transmitted = class `C’, that is, 110.
7800 So, C class has 221networks.
= 780 Ans. (a)
10 bits
Ans. (c)
23. Consider a 2.5 Mbps token ring LAN and frame
20. In a token ring network, the transmission speed is size of 180 bytes. If the ring latency is 210 ms, then
20 bps and the propagation speed is 180 m/ms. The what will be the effective data rate of the LAN?
1-bit delay in this network is equivalent to:
(a) 1.40 Mbps (b) 1.56 Mbps
(a) 20 m of cable (b) 18 m of cable (c) 1.76 Mbps (d) 1.84 Mbps
(c) 9 m of cable (d) 5 m of cable
Solution:
Solution: Given that R = 2×107 bps, B = 1 bps, Data size 180 × 8
V = 180 m/ms; d =? Transmission time = = = 0.00057 s
Bandwidth 2.5×106
d
B = R× Ring latency = Round-trip time = 210 ms = 0.00021 s
v
Bv 1 ´ 180 Total time required to transfer 1440 bits =
d= = = 9 m of cable
2 × 107 × 10−6
(0.00057 + 0.00021) = 0.00078 s
R
So, effective data rate = 1440/0.00078 = 1.84 Mbps
Ans. (c)
Ans. (d)
21. Determine the maximum length of the cable (in
km) for transmitting data at a rate of 400 Mbps 24. How many number of parity bit is required in ham-
in an Ethernet LAN with frames of size 8000 bits. ming code if message size is 8 bit?
Assume the signal speed in the cable to be 250000
km/s. (a) 2 (b) 4
(c) 6 (d) 8
(a) 4 b) 5 (c) 5.5 (d) 7
Solution:
Solution: Maximum length of cable can be calcu-
Given message (m) = 8-bit
lated by the given problem:
To calculate the number of parity bits, we have
8000 bits 2×L the formula:
=
400 × 10 bits/s
6
25 × 104 km/s 2r ≥ m + r + 1, by putting r = 1, 2, 3, 4…
L = 5 km r = 4 will satisfy the given equality.
Ans. (b) Ans. (b)
1. Which of the following assertions is FALSE about (c) IP ensures that a packet is discarded if it is
the Internet Protocol (IP)? unable to reach its destination within a given
number of hops.
(a) It is possible for a computer to have multiple
(d) The packet source cannot set the route of an
IP addresses.
outgoing packet; the route is determined only
(b) IP packets from the same source to the same des-
by the routing tables in the routers on the way.
tination can take different routes in the network.
(GATE 2003: 1 Mark)
Chapter 10.indd 515 4/9/2015 10:08:52 AM

Solution: The packet source can set the route of Since Tx = N/B, where N is the number of bits to
an outgoing packet; the route is determined only be transmitted.
by the routing tables in the routers on the way. Number of bits transmitted, N = Round-trip delay
Ans. (d) × B = 2 × 105× 107 = 200 bytes
Ans. (c)
2. Which of the following functionalities must be
implemented by a transport protocol over and 5. Host A is sending data to host B over a full duplex
above the network protocol? link. A and B are using the sliding window protocol
for flow control. The send and receive window sizes
(a) Recovery from packet losses are 5 packets each. Data packets (sent only from A
(b) Detection of duplicate packets to B) are all 1000 bytes long and the transmission
(c) Packet delivery in the correct order time for such a packet is 50 ms. Acknowledgement
(d) End-to-end connectivity packets (sent only from B to A) are very small
(GATE 2003: 1 Mark) and require negligible transmission time. The
propagation delay over the link is 200 ms. What
Solution: End-to-end delivery is the responsibil- is the maximum achievable throughput in this
ity of a transport layer. communication?
Ans. (d) (a) 7.69 × 106 bps (b) 11.11 × 106 bps
3. The subnet mask for a particular network is (c) 12.33 × 106 bps (d) 15.00 × 106 bps
255.255.31.0. Which of the following pairs of IP (GATE 2003: 2 Marks)
addresses could belong to this network? Solution:
(a) 172.57.88.62 and 172.56.87.233 Throughput = 1 window/RTT
(b) 10.35.28.2 and 10.35.29.4 Round-trip time (RTT) = Transmission time + 2 ×
(c) 191.203.31.87 and 191.234.31.88 Propagation time
(d) 128.8.129.43 and 128.8.161.55 = 50 ms + 2 × 200 ms = 450 ms
As the size of window is 5 packets and 1 packet
Solution: The two addresses should belong to the contains 1000 bytes.
same network, if binary AND operation of both The total size of the packet in bytes is 5 × 1000 =
addresses with net mask should be same. 50000 bytes
5000 bytes
The last octets of IP addresses of 0 is 000 0000. Therefore, throughput = −6
= 11.11 × 106 bps
The last octets of IP address of 43 and 55 and their 450 × 10 s
AND with net mask gives the same result. 5000 bytes
= 11.11 × 106 bps
450 × 10−6 s
Ans. (d)
Ans. (b)
4. A 2-km long broadcast LAN has 107 bps bandwidth
and uses CSMA/CD. The signal travels along the 6. Choose the best matching between Groups 1 and 2.
wire at 2 × 108 m/s. What is the minimum packet
size that can be used on this network? Group: 1 Group: 2
(a) 50 bytes (b) 100 bytes
(c) 200 bytes (d) None of the above P. Data link layer 1. Ensures reliable
(GATE 2003: 2 Marks) transport of data over
a physical point-to-
Solution: Given that length L = 2 km = 2000 m point link
Bandwidth, B = 107 bps Q. Network layer 2. Encodes/decodes
Signal travels S = 2 × 108 m/s data for physical
L 2 × 103
transmission
1
So, propagation delay (Tp ) = = = 5s
2 × 10
S 8 R. Transport layer 3. Allows end-to-end
10
In CSMA/CD network, communication between
two processes
Round-trip delay = 2 ×Tp = 2 × 105 s
The minimum packet size must take round-trip 4. Routes data from one
delay to transmit. network node to the
So, transmission delay (Tx) = Round-trip delay next
Chapter 10.indd 516 4/9/2015 10:08:52 AM

(a) P:1, Q:4, R:3 (b) P:2, Q:4, R:1 So, A’s collision counter will be reinitialized to 0.
(c) P:2, Q:3, R:1 (d) P:1, Q:3, R:2 At the second collision, A and B will have first and
(GATE 2004: 1 Mark) second collisions, respectively.
So, A will have to select random number from [0,1].
Solution: Data link layer provides reliable trans-
But, B will have to select from [0, 1, 2, 3].
mission of data over a physical link.
On mapping, the following are total favouring cases
Network layer provides routing of data packets in
to A’s winning.
the network.
Transport layer is responsible for end-to-end 0—1 0—2 0—3 1—2 1—3
process communication. So, the probability is 5/8 = 0.625.
Ans. (a) Ans. (b)
7. Which of the following is NOT true with respect to 10. The routing table of a router is shown below:
a transparent bridge and a router?
Destination Subnet Mask Interface
(a) Both bridge and router selectively forward data
128.75.43.0 255.255.255.0 Eth0
packets.
(b) A bridge uses IP addresses while a router uses 128.75.43.0 255.255.255.128 Eth1
MAC addresses. 192.12.17.5 255.255.255.255 Eth3
(c) A bridge builds up its routing table by inspect-
Default Eth2
ing incoming packets.
(d) A router can connect between a LAN and a WAN.
On which interface will the router forward pack-
(GATE 2004: 1 Mark)
ets addressed to destinations 128.75.43.16 and
Solution: Both router and bridge selectively for- 192.12.17.10, respectively?
ward data packets and both can connect between
(a) Eth1 and Eth2 (b) Eth0 and Eth2
a LAN and WAN. Routing and bridge builds their
(c) Eth0 and Eth3 (d) Eth1 and Eth3
routing table by inspecting incoming packets but
router and bridge both use MAC address. So,
option (b) is correct. Solution: On performing AND operation between
Ans. (b) incoming IP address and subnet-mask and compar-
ing the result with the destination.
8. How many 8-bit characters can be transmitted per
second over a 9600 baudserial communication link using If there is a match between multiple destinations,
asynchronous mode of transmission with one start bit, then select the destination with the longest length
eight data bits, two stop bits, and one parity bit? subnet mask.
128.75.43.16 matches with 128.75.43.0 and
(GATE 2004: 1 Mark) 128.75.43.0. But the packets addressed to
(a) 600 (b) 800 (c) 876 (d) 1200 128.75.43.16 will be forwarded to Eth1.
Solution: If a result is not matching with any of the given
destinations, then the packet is forwarded to the
9600 default interface (here Eth2).
= 800
1+ 8 + 2 +1 Therefore, the packets addressed to 192.12.17.10
Ans. (b) will be forwarded to Eth2.
9. A and B are the only two stations on an Ethernet. Ans. (a)
Each has a steady queue of frames to send. Both A Common Data Questions 11 and 12: Consider
and B attempt to transmit a frame, collide, and A three IP networks A, B and C. Host HA in net-
wins the first back-off race. At the end of this suc- work A sends messages each containing 180 bytes
cessful transmission by A, both A and B attempt of application data to a host HC in network C. The
to transmit and collide. The probability that A TCP layer prefixes a 20-byte header to the mes-
wins the second back-off race is sage. This passes through an intermediate network
(a) 0.5 (b) 0.625 (c) 0.75 (d) 1.0 B. The maximum packet size, including 20-byte IP
header, in each network is
A: 1000 bytes
Solution: At the first collision, both A and B will B: 100 bytes
have first collision but A wins. C: 1000 bytes
Chapter 10.indd 517 4/9/2015 10:08:53 AM

The network A and B are connected through a 1 15. The maximum window size for data transmission
Mbps link, while B and C are connected by a 512 using the selective reject protocol with n-bit frame
Kbps link (bps = bits per second). sequence numbers is
(a) 2n (b) 2n−1 (c) 2n − 1 (d) 2n−2
1 Mbps 512 Mbps (GATE 2005: 1 Mark)
Network A Network B Network C
Solution: For selective reject protocol, window
11. Assuming that the packets are correctly delivered, size = 2n/2 = 2n−1
how many bytes, including headers, are delivered Ans. (b)
to the IP layer at the destination for one applica- 16. In a network of LANs connected by bridges,
tion message, in the best case? Consider only data packets are sent from one LAN to another
packets. through intermediate bridges. Since more than
(a) 200 (b) 220 one path may exist between two LANs, packets
(c) 240 (d) 260 may have to be routed through multiple bridges.
(GATE 2004: 2 Marks) Why is the spanning tree algorithm used for
bridgerouting?
Solution: When all the 3 packets are delivered,
the bytes are (a) For shortest path routing between LANs
(b) For avoiding loops in the routing paths
(80 + 20) + (80 + 20) + (40 + 20) = 260 (c) For fault tolerance
Ans. (d) (d) For minimising collisions
12. What is the rate at which application data is trans- (GATE 2005: 1 Mark)
ferred to host HC? Ignore errors, acknowledgements
Solution: Spanning tree protocol is used for bridge
and other overheads.
routing to avoid loops in routing paths.
(a) 325.5 Kbps (b) 354.5 Kbps Ans. (b)
(c) 409.6 Kbps (d) 512.0 Kbps
17. An organisation has a class B network and wishes
to form subnets for 64 departments. The subnet
Solution: Actual data sent =180 out of 260 mask would be
So, data rate (180/260) × 512 Kbps = 354.461 Kbps
(a) 255.255.0.0 (b) 255.255.64.0
Ans. (b)
(c) 255.255.128.0 (d) 255.255.252.0
13. Packets of the same session may be routed through (GATE 2005: 1 Mark)
different paths in:
Solution: For class B, 16 bits are reserved as
(a) TCP, but not UDP (b) TCP and UDP network bits. To allocate 64 subnets, 6 bits are
(c) UDP, but not TCP (d) Neither TCP nor UDP added to network bits. 22 bits are for network.
(GATE 2005: 1 Mark) Subnet mask is created by assigning 1 to all net-
work bits.
Solution: Packets travel on network layer. TCP
So, mask is 255.255.252.0.
and UDP are transport layer protocols.
Ans. (d)
Ans. (b)
14. The address resolution protocol (ARP) is used for: 18. In a packet-switching network, packets are routed
from source to destination along a single path
(a) finding the IP address from the DNS. having two intermediate nodes. If the message size
(b) finding the IP address of the default gateway. is 24 bytes and each packet contains a header of 3
(c) finding the IP address that corresponds to a bytes, then the optimum packet size is
MAC address.
(d) finding the MAC address that corresponds to (a) 4 (b) 6 (c) 7 (d) 9
an IP address. (GATE 2005: 2 Marks)
(GATE 2005: 1 Mark) Solution: Optimal packet size is 9.
Solution: ARP works to find the physical address 9−3 = 6 B will be transferred in one packet, so the
of a machine. total message will travel in 4 packets.
Ans. (d) Ans. (d)
Chapter 10.indd 518 4/9/2015 10:08:53 AM

19. Suppose the round-trip propagation delay for a and netmask 255.255.192.0. Which one of the fol-
10 Mbps Ethernet having 48-bit jamming signal is lowing statements is true?
46.4 s. The minimum frame size is
(a) C 1 and C2 both assume they are on the same
(a) 94 (b) 416 network.
(c) 464 (d) 512 (b) C2 assumes C1 is on the same network, but C1
(GATE 2005: 2 Marks) assumes C2 is on a different network.
Solution: RTT = 46.4 × 10−6s
(c) C1 assumes C2 is on the same network, but C2
assumes C1 is on a different network.
10 Mb are sent in = 1 s (d) C1 and C2 both assume they are on different
1 will be sent = 1/107 networks.
48 bits will be in = 4.8 × 10−6 s (GATE 2006: 2 Marks)
Total delay = 46.4 + 4.8 = 51.2 Solution: Network address of C1: 203.197.2.53
Minimum frame size = 51.2 × 10 = 512 bits AND 255.255.128.0 = 203.197.0.0
Ans. (d)
Network address of C2: 203.197.75.201 AND
20. For which one of the following reasons does Internet
255.255.192.0 = 203.197.0.0
Protocol (IP) use the time- to-live (TTL) field in
C1 assumes C2 is on the same network, but C2
the IP datagram header?
assumes C1 on different network.
(a) Ensure packets reach destination within that Ans. (c)
time
23. Station A needs to send a message consisting of
(b) Discard packets that reach later than that
9 packets to station B using a sliding window
time
(window size 3) and Go-Back-N error control strat-
(c) Prevent packets from looping indefinitely
egy. All packets are ready and immediately avail-
(d) Limit the time for which a packet gets queued
able for transmission. If every 5thpacket that A
in intermediate routers
transmits gets lost (but no packets from B ever get
(GATE 2006: 1 Mark) lost), then what is the number of packets that A
Solution: Decrementing TTL will prevent packet will transmit for sending the message to B?
from looping. (a) 12 (b) 14 (c) 16 (d) 18
21. Station A uses 32 byte packets to transmit mes- Solution:
sages to Station B using a sliding window protocol. A B
The round trip delay between A and B is 80 milli- 1 → 1
seconds and the bottleneck bandwidth on the path 2 → 2
between A and B is 128 kbps. What is the optimal 3 → 3
window size that A should use? 4 → 4
(a) 20 (b) 40 5 → lost
(c) 160 (d) 320 6 → discard
7 → discard
→
8 5
Solution: Round-trip delay = 80 ms 9 → 6
10 → lost 7
11 →
Bandwidth = 128 Kbps
discard 8
12 →
In 1 s = 128 Kbits are sent
In 80 ms = 128K × 80 × 10−3 = 128 × 80 bits
discard 9
13 → 7
128 × 80 14 → 8
Window size =
32 × 8
= 40 15 → lost 9
Ans. (b) 16 → 9
Ans. (c)
22. Two computers C1 and C2 are configured as fol-
lows. C1 has IP address 203.197.2.53 and netmask Linked Answer Questions 24 and 25: Consider the
255.255.128.0. C2 has IP address 203.197.75.201 diagram shown below where a number of LANs
Chapter 10.indd 519 4/9/2015 10:08:53 AM

are connected by (transparent) bridges. To avoid (b)

packets looping through circuits in the graph, the Hosts Port
bridges organize themselves in a spanning tree.
First, the root bridge is identified as the bridge H1, H2 4
with the least serial number. Next, the root sends H3, H4 3
out (one or more) data units to enable the setting H5, H6 1
up of the spanning tree of shortest paths from the H5, H8, H9, H10, H11, H12 2
root bridge to each bridge.
Each bridge identifies a port (the root port) (c)
through which it will forward frames to the root Hosts Port
bridge. Port conflicts are always resolved in favour
of the port with the lower index value. When there H3, H4 3
is a possibility of multiple bridges forwarding to H5, H6, H9, H10 1
the same LAN (but not through the root port), ties H1, H2 4
are broken as follows: bridges closest to the root get H7, H8, H11, H12 2
preference and between such bridges the one with
the lowest serial number is preferred. (d)
B1 Hosts Port
H1 H2 1 2 H3 H4 H1, H2, H3, H4 3
H5, H7, H9, H10 1
2 3
4 H7, H8, H11, H12 4
B5 3 B3
1 4 1 2 (GATE 2006: 2 Marks)
Solution: Forwarding table for B3 is
H5 H6 H7 H8
2 2 1 Hosts Port
1 B4 B2
H1, H2, H3, H4 3
3 3
H5, H6, H9, H10 1
H9 H10 H11 H12
H7, H8, H11, H12 2
24. For the given connection of LANs by bridges, which
one of the following choices represents the depth Ans. (a)
first traversal of the spanning tree of bridges? 26. In Ethernet, when Manchester encoding is used,
the bit rate is
(a) B1, B5, B3, B4, B2 (b) B1, B3, B5, B2, B4
(c) B1, B5, B2, B3, B4 (d) B1, B3, B4, B5, B2 (a) Half the baud rate (b) Twice the baud rate
(c) Same as the baud rate (d) None of the above
Solution: DFS (depth first search) is analgo- (GATE 2007: 1 Mark)
rithm for traversing tree or graph. One starts at
the root (selecting some arbitrary node as the root Solution: For transmission of digital information,
in the case of a graph) and explores as far as pos- Manchester coding is used to convert digital infor-
sible along each branch before backtracking. So mation into electrical signals for transmission. It
depth first search traversal is B1, B5, B3, B4, B2. uses two baud for one bit.
Ans. (a) 27. Which one of the following uses UDP as the trans-
port protocol?
25. Consider the correct spanning tree for the previous
question. Let host H1 send out a broadcast ping (a) HTTP (b) Telnet
packet. Which of the following options represents (c) DNS (d) SMTP
the correct forwarding table on B3? (GATE 2007: 1 Mark)
(a) Solution: DNS uses services of UDP protocol.
Hosts Port Ans. (c)
H1, H2, H3, H4 3 28. There are n stations in a slotted LAN. Each station
H5, H6, H9, H10 1 attempts to transmit with a probability p in each
time slot. What is the probability that ONLY one
H7, H8, H11, H12 2
station transmits in a given time slot?
Chapter 10.indd 520 4/9/2015 10:08:54 AM

(a) np(p 1)n−1 (b) (1 − p)n−1 é 2LtR + 2K ù  2LtR 

(d) 1 − (p 1)n−1 (a) êlog2 úû (b) log2
(c) p(p 1)n−1 ë K  K 
é 2LtR + K ù
(c) éêlog2
2LtR + K ù
úû (d) êlog2 úû
Solution: Probability of sending by one station = ë K ë 2K
p(1 − p)n−1 (GATE 2007: 2 Marks)
For n stations, it is np(1 − p)n−1 Solution: Distance between stations M and N = L km
Ans. (a) Propagation delay = t s
29. In a token ring network, the transmission speed Total propagation delay = Lt s
is 107 bps and the propagation speed is 200 m/ Frame size = K bits
ms. The 1-bit delay in this network is equivalent Channel capacity = R bits/s
to: Transmission time = K/R
Let n be the window size.
(a) 500 m of cable (b) 200 m of cable
n Propagation time
(c) 20 m of cable (d) 50 m of cable Utilisation = , where a =
(GATE 2007: 2 Marks) 1 + 2a Transmission
n nK
Solution: Transmission delay for 1 bit t = 1/(107) = =
= 0.1 ms.
1 + (2LtR / K ) 2LtR + K
200 m can be travelled in 1 ms. Therefore, in 0.1 ms, For maximum utilisation: nK = 2LtR + K
20 m can be travelled. 2LtR + K
Therefore, n=
Ans. (c) K
Number of bits needed for n frames is log n.
30. The address of a class B host is to be split into
Ans. (c)
subnets with a 6-bit subnet number. What is the
maximum number of subnets and the maximum 33. Match the following:
number of hosts in each subnet?
Column I Column II
(a) 62 subnets and 262142 hosts
(b) 64 subnets and 262142 hosts (P) SMTP (1) Application layer
(c) 62 subnets and 1022 hosts (Q) BGP (2) Transport layer
(d) 64 subnets and 1024 hosts (R) TCP (3) Data link layer
(S) PPP (4) Network layer
Solution: Here 16 + 6 = 22 bits are reserved for
(5) Physical layer
the network
Number of hosts = 232−22 = 210 − 2 = 1022 (a) P − 2, Q − 1, R − 3, S − 5
Number of subnets = 26 − 2 = 62 (b) P − 1, Q − 4, R − 2, S − 3
Ans. (c) (c) P − 1, Q − 4, R − 2, S − 5
31. The message 11001001 is to be transmitted using (d) P − 2, Q − 4, R − 1, S − 3
the CRC polynomial x3 + 1 to protect it from (GATE 2007: 2 Marks)
errors. The message that should be transmitted is Solution:
(a) 11001001000 (b) 11001001011 SMTP: Application layer for mail transfer
(c) 11001010 (d) 110010010011 BGP: Network layer routing protocol
(GATE 2007: 2 Marks) TCP: Transport layer transmission protocol
PPP: Data link layer protocol for direct connections
Solution: The polynomial is equivalent to 1001. Ans. (b)
Divide 1001 with 11001001000 and find the remainder.
34. What is the maximum size of data that the appli-
Remainder is 011. Append 011 at the end of input
cation layer can pass on to the TCP layer below?
string, it will be 11001001011.
Ans. (b) (A) Any size
32. The distance between two stations M and N is (B) 216 bytes − size of TCP header
L km. All frames are K bits long. The propaga- (C) 216 bytes
tion delay per kilometre is t s. Let R bits/s be the (D) 1500 bytes
channel capacity. Assuming that processing delay (GATE 2008: 1 Mark)
is negligible, the minimum number of bits for the Solution:
sequence number field in a frame for maximum util- There is no limit of data passing at application layer.
isation, when the sliding window protocol is used, is Ans. (a)
Chapter 10.indd 521 4/9/2015 10:08:55 AM

35. Which of the following system calls results in the Subsequently, the client process P executes a
sending of SYN packets? socket() system call followed by connect() system
call to connect to the server process S. The server
(a) Socket (b) Bind
process has not executed any accept() system call.
(c) Listen (d) Connect
Which one of the following events could take place?
(GATE 2008: 1 Mark)
(a) Connect() system call returns successfully
Solution: Connect() is called by the client and con- (b) Connect() system call blocks
nection is established using three-way hand shake. (c) Connect() system call returns an error
Ans. (d) (d) Connect() system call results in a core dump
36. In the slow start phase of the TCP congestion con-
trol algorithm, the size of the congestion window Solution: The accept() call does not execute.
So, connect() call did not get response for a time
(a) does not increase. stamp to wait, therefore, connect() system call
(b) increases linearly. returns an error.
(c) increase quadratically. Ans. (c)
(d) increase exponentially.
(GATE 2008: 2 Marks) 40. In the RSA public key cryptosystem, the private
and public keys are (e,n) and (d,n), respectively,
Solution: In the slow start phase of the TCP con- where n = p*q and p and q are large primes. Besides,
gestion control algorithm, the size of the congestion n is public and p and q are private. Let M be an
window increases exponentially. integer such that 0 < M < n and f (n) = (p − 1)
Ans. (d) (q − 1). Now, consider the following equations:
37. If a class B network on the Internet has a subnet I. M′ = Memod n
mask of 255.255.248.0, what is the maximum M = (M′)d mod n
number of hosts per subnet? II. ed ≡ 1 mod n
III. ed ≡ 1 mod f (n)
IV. M′ = Memod f (n)
(a) 1022 (b) 1023
M = (M′)dmod f (n)
(c) 2046 (d) 2047
Which of the above equations correctly represent
Solution: Number of network bits = 21 the RSA cryptosystem?
Number of host bits = 11 (a) I and II (b) I and III

Number of hosts = 211 − 2 = 2046 (c) II and IV (d) III and IV
38. A computer on a 10 Mbps network is regulated by Solution: Encryption: M′ = Me mod n

a token bucket. The token bucket is filled at a rate Decryption: M = (M′)d mod n
of 2 Mbps. It is initially filled to capacity with 16 ed ≡ 1 mod f(n)
Mbits. Ans. (b)
What is the maximum duration for which the com- 41. While opening a TCP connection, the initial
puter can transmit at the full 10 Mbps? sequence number is to be derived using a time-of-
(a) 1.6 s (b) 2 s day (ToD) clock that keeps running even when the
(c) 5 s (d) 8 s host is down. The low order 32 bits of the coun-
(GATE 2008: 2 Marks) ter of the ToD clock is to be used for the initial
sequence numbers. The clock counter increments
Solution: Data transfer of token bucket = 10 Mbps once per millisecond. The maximum packet life-
Rate of transfer = 2 Mbps time is given to be 64s.
Initially filled capacity = 16 Mbits Which one of the choices given below is closest to the
minimum permissible rate at which sequence num-
b 16
Maximum burst time = = =2 s bers used for packets of a connection can increase?
M − r 10 − 2
Ans. (b) (a) 0.015/s (b) 0.064/s
39. A client process P needs to make a TCP connec- (c) 0.135/s (d) 0.327/s
tion to a server process S. Consider the following (GATE 2009: 2 Marks)
situation: the server process S executes a socket(), Solution: The frames from the sending station are
a bind() and a listen() system call in that order, numbered sequentially.
following which it is pre-empted. Ans. (b)
Chapter 10.indd 522 4/9/2015 10:08:56 AM

42. Let G(x) be the generator polynomial used for CRC (c) It can be used to optimise throughput.
checking. What is the condition that should be sat- (d) It can be used to prevent packet looping.
isfied by G(x) to detect odd number of bits in error? (GATE 2010: 1 Mark)
(a) G(x) contains more than two terms. Solution: Value of TTL is decremented to pre-
(b) G(x) does not divide 1+xk, for any not exceed- vent packet looping.
ing the frame length k. Ans. (d)
(c) 1+x is a factor of G(x). 46. Which one of the following is not a client—server
(d) G(x) has an odd number of terms. application?
(a) Internet chat (b) Web browsing
Solution: To detect odd number of errors, (x + 1) (c) E-mail (d)Ping
must be present. (GATE 2010: 1 Mark)
Ans. (c)
Solution: ping is a command not an application.
Linked Answer Questions 43 and 44: Frames of 1000 It is to check connectivity and there is no need to
bits are sent over a 10 bps duplex link between two communicate with server.
hosts. The 6 propagation time is 25 ms. Frames are Ans. (d)
to be transmitted into this link to maximally pack
47. Suppose computers A and B have IP addresses
them in transit (within the link).
10.105.1.113 and 10.105.1.91, respectively, and
43. What is the minimum number of bits (l) that will they both use the same net mask N. Which of the
be required to represent the sequence numbers dis- values of N given below should not be used if A and
tinctly? Assume that no time gap needs to be given B should belong to the same network?
between transmissions of two frames.
(a) 255.255.255.0 (b) 255.255.255.128
(a) l = 2 (b) l = 3 (c) 255.255.255.192 (d) 255.255.255.224
(c) l = 4 (d) l = 5 (GATE 2010: 2 Marks)
Solution: To belong to a different network, A and
Solution: Transmission delay of link = 1000/106 B should have different network address.
= 1 ms
Perform logical AND operation on all the options.
Propagation delay = 25 ms For example,option (d)
Maximum 25 frames can be sent, for 25 frames bits IP address of A: 10.105.1. 01110001
required are 5. Network mask: 255.255.255. 11100000
Ans. (d) Network address of A: 10.105.1. 01100000
44. Suppose that the sliding window protocol is used IP address of B: 10.105.1. 01011011
with the sender window size of 2l, where l is the Network mask: 255.255.255. 11100000
number of bits identified in the earlier part and l Network address of A: 10.105.1. 01000000
acknowledgements are always piggy backed. After Both have different network addresses.
sending 2l frames, what is the minimum time the Ans. (d)
sender will have to wait before starting transmis-
Linked Answer Questions 48 and 49: Consider a net-
sion of the next frame? (Identify the closest choice
work with 6 routers R1 to R6 connected with links
ignoring the frame processing time.)
having weights as shown in the following diagram
(a) 16 ms (b) 18 ms (c) 20 ms (d) 22 ms
(GATE 2009: 2 Marks) R2 7 R4
Solution: Size of window = 32 6 8
Round trip time = 2 × 25 ms = 50 ms R1 2 R6

1
To send the next frame, the sender has to wait for
50 − 32 = 18 ms
3 4
Ans. (b) R3
9 R5
45. One of the header fields in an IP datagram is the
time-to-live (TTL) field. Which of the following
48. All the routers use the distance vector based rout-
statements best explains the need for this field?
ing algorithm to update their routing tables. Each
(a) It can be used to prioritise packets. router starts with its routing table initialised to
(b) It can be used to reduce delays. contain an entry for each neighbour with the
Chapter 10.indd 523 4/9/2015 10:08:56 AM

weight of the respective connecting link. After all Following will be distance vectors of all nodes
the routing tables stabilise, how many links in the R1 (2, 3, 9, 10, 11)
network will never be used for carrying any data? Links used: R1—R2, R1—R3, R2—R4, R4—R5, R4—R6
(a) 4 (b) 3 (c) 2 (d) 1 R2 (2, 7, 8, 9)
(GATE 2010: 2 Marks) Links used: R2—R3, R2—R4, R4—R5, R4—R6
R3 (9, 9, 11)
Solution: We can check one by one all shortest Links used: R3—R2, R2—R4, R3—R5, R4—R6
distances. When we check for all shortest distances R4 (1, 2)
for Ri we don’t need to check its distances to R0 Links used: R4—R5, R4—R6
to Ri−1 because the network graph is undirected. R5 (3)
Following will be distance vectors of all nodes. Links used: R5—R4, R4—R6
Shortest distances from R1 to R2, R3, R4, R5 and R6 If we mark all the used links one by one, we can see
R1 (5, 3, 12, 12, 16) that following links are never used.
Links used: R1—R3, R3—R2, R2—R4, R3—R5, R5—R6 R5—R6
Ans. (b)
Shortest distances from R2 to R3, R4, R5 and R6
R2 (2, 7, 8, 12) 50. A layer-4 firewall (a device that can look at all pro-
Links used: R2—R3, R2—R4, R4—R5, R5—R6 tocol headers up to the transport layer) CANNOT
Shortest distances from R3 to R4, R5 and R6 (a) B lock entire HTTP traffic during 9:00PM and
R3 (9, 9, 13) 5:00 AM
Links used: R3—R2, R2—R4, R3—R5, R5—R6 (b) Block all ICMP traffic
Shortest distances from R4 to R5 and R6 (c) Stop incoming traffic from a specific IP address
R4 (1, 5) but allow outgoing traffic to the same IP address
Links used: R4—R5, R5—R6 (d) Block TCP traffic from a specific user on a
Shortest distance from R5 to R6 multi-user system during 9:00 PM and 5:00 AM
R5 (4) (GATE 2011: 1 Mark)
Links used: R5—R6 Solution: As Layer-4 firewall cannot block traf-
If we mark all the used links one by one, we can see fic of Layer-5 (application layer) and HTTP is an
that following links are never used. application layer protocol. So, option (a) is correct.
R1—R2 R4—R6 Ans. (a)
Ans. (c)
51. Consider different activities related to email.
49. Suppose the weights of all unused links in the pre- m1: Send an email from a mail client to a mail
vious question are changed to 2 and the distance server
vector algorithm is used again until all routing tables m2: Download an email from mailbox server to a
stabilise. How many links will now remain unused? mail client
(a) 0 (b) 1 (c) 2 (d) 3 m3: Checking email in a web browser
(GATE 2010: 2 Marks) Which is the application level protocol used in each
Solution: After the weights of unused links() are activity?
changed to following graph. (a) ml: HTTP m2: SMTP m3: POP
(b) ml: SMTP m2: FTP m3: HTTP
R2 7 R4 (c) ml: SMTP m2: POP m3: HTTP
(d) ml: POP m2: SMTP m3: IMAP
2 (GATE 2011: 1 Mark)
2
2 1 Solution:
R1 R6
m1: SMTP is responsible for mail transfer.
m2: POP is responsible for downloading mail from
3 3 mailbox server to mal client.
m3: HTTP is responsible for viewing application on
R3
9 web browser.
R5
Ans. (c)
Chapter 10.indd 524 4/9/2015 10:08:57 AM

Linked Answer Questions 52 and 53: Consider a 54. Which of the following transport layer protocols is
network with five nodes, N1 to N5, as shown below. used to support electronic mail?
(a) SMTP (b) IP (c) TCP (d) UDP
N1
(GATE 2012: 1 Mark)
1 Solution: Electronic mail does not require TCP

connection between sender and receiver of email.
Ans. (c)
3
N5 N2 55. The protocol data unit (PDU) for the application-
layer in the Internet stack is
(a) segment. (b) datagram.
4 6 (c) message. (d) frame.
(GATE 2012: 1 Mark)
Solution:
2
N4 N3 The protocol data unit (PDU) for Data link layer
= Frame; Network layer = Datagram; Transport
The network uses a distance vector routing proto- layer = Segment; Application layer = Message.
col. Once the routes have stabilised, the distance Ans. (c)
vectors at different nodes are as follows: 56. In the IPv4 addressing format, the number of
N1:(0,1, 7, 8, 4) N2: (1, 0, 6, 7, 3) networks allowed under class C addresses is
N3: (7, 6, 0, 2, 6) N4: (8,7, 2,0,4) (a) 214 (b) 27 (c) 221 (d) 224
N5: (4, 3, 6, 4, 0) (GATE 2012: 1 Mark)
Each distance vector is the distance of the best-
known path at that instance to nodes, N1 to N5, Solution:
where the distance to itself is O. Also, all links Network bits Host bits
are symmetric and the cost is identical in both
24 bits 8 bits
directions. In each round, all nodes exchange their
distance vectors with their respective neighbours. Starting with 3 bits (110) reserved to recognise the
Then all nodes update their distance vectors. In class. So, the number of networks are 221.
between two rounds, any change in cost of a link Ans. (c)
will cause the two incident nodes to change only
57. Consider a source computer (S) transmitting a file
that entry in their distance vectors.
of size 106 bits to a destination computer (D) over a
52. The cost of link N2—N3 reduces to 2 (in both direc- network of two routers (R1 and R2) and three links
tions). After the next round of updates, what will (L1, L2, and L3). L1 connects S to R1; L2 connects
be the new distance vector at node N3? R1 to R2; and L3 connects R2 to D. Let each link be
(a) (3, 2, 0, 2, 5) (b) (3, 2, 0, 2, 6) of length 100 km. Assume signals travel over each
(c) (7, 2, 0, 2, 5) (d) (7, 2, 0, 2, 6) link at a speed of 108 meters per second. Assume
(GATE 2011: 2 Marks) that the link bandwidth on each link is 1Mbps.
Let the file be broken down into 1000 packets each
Solution: N3: (3, 2, 0, 2, 5) of size 1000 bits. Find the total sum of transmis-
Ans. (a) sion and propagation delays in transmitting the file
53. After the update in the previous question, the link from S to D?
N1—N2 goes down. N2 will reflect this change imme-
(a) 1005 ms (b) 1010 ms
diately in its distance vector as cost, ∞. After the
(c) 3000 ms (d) 3003 ms
NEXT ROUND of update, what will be the cost to
N1 in the distance vector of N3?
Solution
(a) 3 (b) 9 (c) 10 (d) ∞
S R R D
Solution: N3 to N1 either by N2 or by N4
N2—N1 is ∞, so N3 will choose to go via N4 to N1 (2 + 8). Transmission delay for 1 packet from each of S, R1
Ans. (c) and R2 will take 1ms.
Chapter 10.indd 525 4/9/2015 10:08:57 AM

Propagation delay on L1, L2 and L3 for one packet Organisation A : 2048

is 1 ms. Organisation B : 1024
Transmission delay + Propagation delay = 2 ms Remaining : 1024
First packet will reach at 6th ms
Allocation of addresses to A: 11 bits are required
Second packet will reach at 7th ms
for organisation A.
1000 packets will reach at 1005th ms
Ans. (a) 245.248. 10000 100.00000000
58. Consider an instance of TCP’s Additive Increase Address allocated is 245.248.136.0/21
Multiplicative Decrease (AIMD) algorithm where Allocation of addresses to B: 10 bits are required
the window size at the start of the slow start for organisation B.
phase is 2 MSS and the threshold at the start of
245.248. 100000 00.00000000
the first transmission is 8 MSS. Assume that a
timeout occurs during the fifth transmission. Find Address allocated is 245.248.128.0/22.
the congestion window size at the end of the tenth Ans. (a)
transmission. 60. The transport layer protocols used for real time
(a) 8 MSS (b) 14 MSS multimedia, file transfer, DNS and email, respec-
(c) 7 MSS (d) 12 MSS tively, are
(GATE 2012: 2 Marks) (a) TCP, UDP, UDP and TCP
Solution: (b) UDP, TCP, TCP and UDP
In slow start, with each iteration, size of congestion (c) UDP, TCP, UDP and TCP
window doubles. So, (d) TCP, UDP, TCP and UDP
(GATE 2013: 1 Mark)
T = 1 ws = 2
T = 2 ws = 4 Solution: Real-time multimedia: UDP (session less
T = 3 ws = 8 protocol, used where fast data transfer is required)
Threshold is 8. So now window size will increase by File transfer: TCP
one using additive increase method. DNS: UDP
E-mail: TCP
T = 4 ws = 9
Ans. (c)
T = 5 ws = 10, here timeout occurs.
Hence, threshold = 10/2 = 5, ws is reset 61. Using public key cryptography, X adds a digital
signature s to message M, encrypts < M, s > , and
T = 6 ws = 2 sends it to Y, where it is decrypted. Which one
T = 7 ws = 4 of the following sequences of keys is used for the
T = 8 ws = 5 [threshold was 5, so apply addi- operations?
tive increase]
T = 9 ws = 6 (a) E ncryption: X’s private key followed by Y’s pri-
T = 10 ws = 7 vate key; Decryption: X’s public key followed
Ans. (c) by Y’s public key
(b) Encryption: X’s private key followed by Y’s
59. An Internet Service Provider (ISP) has the follow- public key; Decryption: X’s public key followed
ing chunk of CIDR-based IP addresses available by Y’s private key
with it: 245.248.128.0/20. The ISP wants to give (c) Encryption: X’s public key followed by Y’s pri-
half of this chunk of addresses to organisation A, vate key; Decryption: Y’s public key followed
and a quarter to organisation B, while retaining by X’s private key
the remaining with itself. Which of the following is (d) Encryption: X’s private key followed by Y’s
a valid allocation of addresses to A and B? public key; Decryption: Y’s private key fol-
(a) 245.248.136.0/21 and 245.248.128.0/22 lowed by X’s public key
(b) 245.248.128.0/21 and 245.248.128.0/22 (GATE 2013: 1 Mark)
(c) 245.248.132.0/22 and 245.248.132.0/21 Solution: X uses private key to add digital signa-
(d) 245.248.136.0/24 and 245.248.132.0/21 ture, then uses Y’s public key to encrypt. On the
(GATE 2012: 2 Marks) other end, Y first uses its private key followed by
Solution: Total numbers of addresses available X’s public key.
are 4096. Ans. (d)
Chapter 10.indd 526 4/9/2015 10:08:58 AM

62. Assume that source S and destination D are con- Solution: Data rate: 500 Mbps = 5 × 108 bps
nected through two intermediate routers labelled R. Frame size: 10000 bits
Speed: 200000 km/s
S R R D 5 × 108 bits can travel = 1s
Determine how many times each packet has to visit 104 1

104 bits will travel = =
the network layer and the data link layer during a 5 × 10 8
5 × 104
transmission from S to D.
In 1 s, distance travelled = 2 × 105 km
(a) Network layer — 4 times and Data link layer — 4 2 × 105
times In (1/5 × 104) s, distance travelled is = = 4 km
5 × 104
(b) Network layer — 4 times and Data link layer — 3 2 × 10 5
= 4 km
times
Hence, the maximum distance5is × 10 4
4/2 = 2 km
(c) Network layer — 4 times and Data link layer — 6
times Ans. (b)
(d) Network layer — 2 times and Data link layer — 6 64. In an IPv4 datagram, the M bit is 0, the value of
times HLEN is 10, the value of total length is 400 and
(GATE 2013: 1 Mark) the fragment offset value is 300. The position of the
Solution: Transmission from S to D will follow datagram, the sequence numbers of the first and
these steps: the last bytes of the payload, respectively, are
At S: Network layer → Data link layer (a) Last fragment, 2400 and 2789
At R1: Data link layer → Network layer → Data (b) First fragment, 2400 and 2759
link layer (c) Last fragment, 2400 and 2759
At R2: Data link layer → Network layer → Data (d) Middle fragment, 300 and 689
link layer (GATE 2013: 2 Marks)
At D: Data link layer → Network layer
Solution: M = 0 indicates no more fragment
Therefore, the packet has to visit 6 times data link
means last fragment
layer and 4 times network layer.
Ans.(c) Actual header length is 4 × 10 = 40
Total length = 400 bytes (40-byte header +
63. Determine the maximum length of the cable
360-byte payload)
(in km) for transmitting data at a rate of 500 Mbps
Fragment offset = 300 × 8 = 2400 bytes (measured
in an Ethernet LAN with frames of size 10000 bits.
in terms of 8 bytes)
Assume the signal speed in the cable to be 200000
First byte sequence number of last fragment is
km/s.
2400.
(a) 1 (b) 2 Sequence number of last byte payload is 2400 +
(c) 2.5 (d) 5 360 − 1 = 2759.
PRACTICE EXERCISES
3. The size of MAC address and IPv4 address is

Set 1
(a) 8 bits and 24 bits (b) 32 bits and 48 bits
1. Which layer is associated with log in/log out from (c) 48 bits and 32 bits (d) 24 bits and 32 bits
the network?
4. The bit size of cyclic redundancy code in an
(a) Transport (b) Presentation Ethernet is
(c) Data link (d) Session
(a) CRC is not used in Ethernet
2. Which layer is associated with IP addresses? (b) 8
(a) Session (b) Network (c) 24
(c) Transport (d) Data link (d) 32
Chapter 10.indd 527 4/9/2015 10:08:58 AM

5. MTU stands for 15. How many hops can be travelled by the IP packet
having TTL value 5?
(a) Minimum Transfer Unit
(b) Minimum Telephony Unit (a) It cannot travel any hop.
(c) Maximum Transfer Unit (b) It can travel only single hop.
(d) Memory Transfer Unit (c) It can travel five hops.
(d) It can travel four hops.
6. Using 7-bit sequence number, what is the maxi-
mum size (in bits) of the sender and receiver 16. What does the acronym ISDN for?
window using Go-Back-N ARQ?
(a) Indian Standard Data Network
(a) 127 and 1 (b) 1 and 127 (b) Integrated Services Digital Networks
(c) 127 and 127 (d) 1 and 1 (c) Intelligent Secure Digital Network
(d) Intelligent Services Digital Network
7. Using a 5-bit sequence number, what is the maxi-
mum size of the sender and receiver window using 17. If N is the maximum sequence number, then the
selective repeat ARQ? window size in selective repeat and Go-Back-N
(a) 16 and 16 (b) 1 and 16 protocols are, respectively,
(c) 16 and 1 (d) Depend upon the bits (a) N/2, N − 1 (b) N/2, N + 1
selected to be repeated (c) N + 1/2, N − 1/2 (d) N − 1, N
8. Which protocol is being used in wireless to over- 18. CRC can detect all burst error of up to M errors, if
come collision? generator polynomial G(x) is of degree
(a) CSMA/CD (a) It does not matter (b) M − 1
(b) CSMA/CA (c) M/2 (d) M + 1
(c) Both
(d) WEP 19. Which layers of the OSI reference model are host-
to-host layers?
9. The host bits present in class B IP address are
(a) Transport, session, presentation, application
_______.
(b) Session, presentation, application
10. Which protocol is used to provide error reporting (c) Datalink, transport, presentation, application
services to IP address? (d) Physical, datalink, network, transport
(a) ICMP (b) ARP 20. A nibble is used for packed sequence number-
(c) BGP (d) OSPF ing in a sliding window protocol used in a
computer network. What is the maximum window
11. Which protocol is used to find the physical address size?
for a logical address?
(a) D epends upon the underlying network
(a) ARP (b) RARP protocol
(c) IGMP (d) ICMP (b) Depends upon the number of nodes involved in
the network
12. Which protocol cannot be used to find the logi- (c) 15
cal address if the server is residing out of your (d) 16
network?
21. Infrared signals can be used in a closed area using
_______ propagation.
(a) ARP (b) RARP
(c) BOOTP (d) IGMP
(a) Sky (b) Ground
13. Which topology requires a central controller or hub? (c) Line of sight (d) Small signal
(a) Bus (b) Star 22. A bridge has access to _______ address in the same
(c) Mesh (d) Ring network.
14. The number of nibbles are reserved in the IP (a) Physical (b) Logical
header for header length is _______. (c) Supernet (d) Link state
Chapter 10.indd 528 4/9/2015 10:08:59 AM

23. Match the following: (a) DHCP (b) SNMP

(c) SOAP (d) SMTP
List — I List — II
30. Which layer of OSI reference model is responsible
A. Physical layer i. R
esources to network for decomposition of messages and generation of
access are used sequence numbers to ensure correct re-composition
B. Data link layer ii. P
ackets are moved from end-to-end communication in a network?
from one destination (a) Physical (b) Session
to other (c) Transport (d) Data link
C. Network layer iii. P
rocess to process
31. The VLF and LF bauds use _______ propagation
message delivery
for communication.
D. Transport layer iv. B
it stream is
transmitted (a) Ground (b) Sky
(c) Line of sight (d) Space
E. Application v. Frames are formed
Layer 32. Using the 8-bit sequence numbers, what is the
maximum size of the sender and receiver window
Codes: in selective repeat ARQ?
ABCDE (a) 128 and 128
(a) iv v ii iii i (b) 1 and 127
(b) i iii iv ii v (c) 128 and 127
(c) iv ii i v iii (d) 8 and 8
(d) i ii iii iv v
24. The ________ is a unit to measure the signal
33. Vulnerable Slotted ALOHA is
strength in wireless networks (a) equal to average frame transmission time
(b) half of average frame transmission time
(a) Frequency (b) Bandwidth delay product
(c) two times the average frame transmission time
(c) Attenuation (d) Decibel
(d) none
25. Which one of the following media is multi-drop?
34. A CSMA/CD network supports data rate of
(a) UTP cable (b) STP cable 10Mbps, the maximum distance between any
(c) Thick coaxial cable (d) Multi-mode fibre two stations is found to be 2500 m for the correct
optic cable operation of the collision detection process. What
should be the maximum distance if the data rate is
26. What is the baud rate of the standard 20 Mbps
increased to 100 Mbps?
Ethernet?
(a) 25000 m (b) 2500 m
(a) 10 megabaud (b) 40 megabaud
(c) 250 m (d) 25 m
(c) 30 megabaud (d) 20 megabaud
35. Consider an IP in CIDR notation as 220.19.18.87/
27. Binary symmetric channel uses
24. The first and the last address of this network
(a) half-duplex protocol will be?
(b) full-duplex protocol
(a) 220.19.18.24 and 220.19.18.87
(c) simplex protocol
(b) 220.19.18.0 and 220.19.18.87
(d) simplex protocol with separate data and
(c) 220.19.18.0 and 220.19.18.123
control bits
(d) 220.19.18.0 and 220.19.18.255
28. Which of the following addresses is used to deliver
36. In ICMP, 8 byte of the IP datagram is included in
a message to the correct application program run-
ICMP data part because:
ning on a host?
(a) It contains the IP address.
(a) Port (b) DNS address
(b) It contains the port numbers.
(c) Logical address (d) MAC address
(c) The statement is false, IP datagram is not
29. Which of the following protocol is used for per- included in ICMP data part.
forming RPCs between applications in a language (d) It is helpful in checking the errors with CRC
and system in an independent way? used.
Chapter 10.indd 529 4/9/2015 10:08:59 AM

37. A receiver receives the IP packet with the first 8 (a) 3000 (b) 1544
bits as 10100011. The length of the IP packet is (c) 110 (d) 64
(a) 163 bytes (b) 1 byte 45. In a pure ALOHA network, 100 stations share a
(c) 127 bytes (d) 12 bytes link of 1 Mbps. The throughput of such a network
38. The router performs at ________ layer(s). having 1000 bits size of frames and each station is
transmitting at the rate of 10 frames/s is
(a) Only physical
(b) Physical, datalink and network (a) 10% (b) 13.53%
(c) Physical, datalink and transport (c) 12.8% (d) 18.16%
(d) Datalink and network 46. An HDLC does not support
39. A packet has arrived with the offset value 100 and (a) simplex
HLEN value 5. The total length value is 100. What (b) multipoint link
is the number of first byte and the last byte? (c) balanced multipoint
(a) 879 (b) 880 (d) full duplex
(c) 881 (d) 882
47. A hub in the network is
(a) a passive device
(a) In TCP/IP-based services, the destination (b) an active device
address is to be specified only during the initial (c) a server that serves every node
stage of setup. (d) a power supply concentrator
(b) Initial setup is required for UDP-based service.
(c) Packet sequencing is not guaranteed in TCP/ 48. Consider the following message: M = 1010001101.
IP-based services. The CRC of this message using the divisor polyno-
(d) Initial setup is not possible in UDP-based service. mial x5 + x4 + x2 + 1 is
41. How many characters (7 bits + 1 parity) can be (a) 1110 (b) 0101
transmitted over 2400 bps line, if the transfer is (c) 1001 (d) 0010
asynchronous (1 start and 1 stop)? 49. The broadcast address for the subnet that IP address
(a) 2400/8 (b) 2400/10 192.168.26.8, 255.255.255.248 is a member of
(c) 2400 × 8/2 (d) 256 (a) 192.168.26.225 (b) 192.168.26.255
42. In CRC, if the data unit is 110111001 and the divi- (c) 192.168.127.255 (d) 192.168.26.0
sor is 1011 then what is the dividend at the receiver? 50. A sender uses public key cryptography to send a
(a) 110111001101 secret message. Which of the following is true?
(b) 110111001000 (a) Sender encrypts using receiver’s public key
(c) 110111001011 (b) Sender encrypts using his own public key
(d) 101110111001 (c) Receiver decrypts using sender's private key
43. With respect to ICMP, which of the following (d) Receiver decrypts using own private key
statements is false?
51. Which of the following functionality must be
(a) ICMP reports about the routers. implemented by transport layer over and above the
(b) ICMP is also used to test connectivity. network protocol?
(c) ICMP and BOOTP are equivalent in their
(a) Recovery from packet loss
usage.
(b) End to end connectivity
(d) An ICMP message type is encapsulated for
(c) Packet delivery in encapsulated abstract order
transmission.
(d) Secure transmission from end to end
44. A 3000-lm long trunk is used to transmit frames
52. During the process of packet forwarding by the
routers, _____ field is not updated.
using Go-Back-N protocol. The propagation speed
is 6 ms/km and the trunk data rate is 1.544 Mbps.
We ignore the time it takes to receive the acknowl- (a) IP header source address
edgement bits. Frame size is 64 bytes. What should (b) IP header target address
be the maximum window size at the sender’s side (c) IP header TTL
in order to achieve 100% efficiency? (d) IP header checksum
Chapter 10.indd 530 4/9/2015 10:08:59 AM

53. Consider the network, using CIDR addressing, a PC 7. Station A uses 32 bytes packets to transmit mes-
having the IP address as 180.10.10.10/20. The num- sages to station B using a sliding window protocol.
bers of addresses in such a network will be _______ ? The round-trip delay between A and B is 60 ms
and the bottleneck bandwidth on the path between
(a) 255 (b) 4094 (c) 4096 (d) 1024
A and B is between 128 Kbps. What is the optimal
window size that A should use?
Set 2
(a) 20 (b) 30
1. Consider the following two cases given below, (c) 60 (d) 40
where G denotes the generator polynomial and M
denotes the received message: 8. Frames of 1000 bits are sent over a 106 bps duplex
link between the two hosts. The propagation time
Case 1: G = 1100, M = 1010101 is 34 ms. Frames are to be transmitted into this
Case 2: G = 11001, M = 111000111011 link to maximally pack them in transit (within the
In which transmissions errors occur? link). What is the minimum number of bits (l) that
(a) 1 only (b) 2 only will be required to represent the sequence of num-
(c) Both 1 and 2 (d) Neither 1 nor 2 bers distinctly? Assume that no time gap needs to
be given between transmissions of two frames.
2. A block of addresses is granted to a small
organisation. We have given a random address (a) l = 4 (b) l = 5
216.18.9.21/28. (c) l = 6 (d) l = 7
What is the first address in the block?
9. A 3-km long broadcast LAN has 107 bps bandwidth
(a) 216.18.9.0 (b) 216.18.9.16 and uses CSMA/CD. The signal travels along the
(c) 216.18.9.18 (d) 216.18.9.21 wire at 4 × 108 m/s. What is the minimum packet
size that can be used on this network?
3. How many total numbers of addresses are there in the
block using the information in the above question? (a) 50 bytes (b) 100 bytes
(c) 200 bytes (d) None of these
(a) 16 (b) 32 (c) 64 (d) 128
10. Match the following:
4. A channel has 10 Kbps bit rate using stop-and-wait
protocol and has propagation delay of 5 ms. For a (A) 204.26.37.47 i. Class A
frame with 360 bit, what will be the efficiency? (B) 145.12.13.0 ii. Class B
(C) 230.54.256.56 iii. Class C
(a) 92% (b) 88%
(D) 255.255.255.255 iv. Class D
(c) 84% (d) 78%
(E) 126.126.126.126 v. Class E
(F) 85.18.257.24 vi. Invalid IP
5. Consider a 3 Mbps token ring LAN and frame size
is of 200 byte. If the ring latency is 150 ms, then (a) A-iv B-ii C-iv D-v E-i F-i
what will be the effective data rate of the LAN? (b) A-iii B-ii C-vi D-v E-i F-vi
(c) A-iii B-ii C-iv D-iv E-vi F-i
(a) 2.10 Mbps (b) 2.25 Mbps
(d) A-iii B-ii C-iv D-vi E-i F-vi
(c) 2.35 Mbps (d) 2.90 Mbps
11. In a class C network if subnet mask is given as
6. Host A is sending data to host B over a full duplex
255.255.255.224. Calculate the number of subnet
link. A and B are using the sliding window protocol
and number of host in each subnet (practically
for flow control. The send and receive window sizes
possible).
are 6 packets each. Data packets (sent only from A
to B) are all 1000 bytes long and the transmission (a) 4 and 32 (b) 6 and 64
time for such a packet is 40 ms. Acknowledgement (c) 8 and 32 (d) 6 and 30
packets (sent only from B to A) are very small
and require negligible transmission time. The 12. We need to make a super network out of 16 class
propagation delay over the link is 180 ms. What C block, what is the supernet mask?
is the maximum achievable throughput in this (a) 255.255.255.240
communication? (b) 255.240.0.0
(a) 14.28 × 106 bps (b) 14.61 × 106 bps (c) 255.255.240.0
(c) 19.33 × 106 bps (d) 23.40 × 106 bps (d) 255.255.224.0
Chapter 10.indd 531 4/9/2015 10:08:59 AM

13. Given that bandwidth = 10 Mbps, distance = 4 km 20. If 6 bits are used for sequence number, then what
and velocity = 2 × 108 m/s. The number of bits is the sender window size and receiver window size
are transmitted in RTT is _______. in selective repeat scheme?
14. Suppose the round-trip propagation delay for a (a) 64 (b) 63

8 Mbps Ethernet having 48-bit jamming signal is (c) 32 (d) 31
50 ms. The minimum frame size is _______.
21. If 10 Base 5 cable is used and velocity is 2 × 108
15. The address of a class B host is to be split into m/s. What will be the frame size in CSMA/CD?
subnets with a 7-bit subnet number. What is the (a) 40 (b) 50
maximum number of subnets and the maximum (c) 60 (d) 64
number of hosts in each subnet?
(a) 128 subnets and 1024 hosts 22. If the capacity of a router is 1.2 mb, output rate
(b) 128 subnets and 1022 hosts of data is .8 mb/s token that are generated is 6
(c) 126 subnets and 512 hosts mb/s. Calculate the time matrix busted traffic is
(d) 126 subnets and 510 hosts routed.
(a) 0.6 s (b) 1.2 s
16. If a class B network on the Internet has a subnet (c) 1.5 s (d) 2.0 s
mask of 255.255.248.0, what is the maximum
number of hosts per subnet? 23. If the total length bits are 0000000000111111 and
(a) 1022 (b) 1023 HLEN = 1010, then which one is true about the
(c) 2046 (d) 2047 size of header and payload?
(a) 40 and 63
17. In a network,propagation delay is 100 ms and band- (b) 40 and 23
width is given as 100 Mbps. Calculate how many (c) 24 and 63
RTTs are required for transmitting 40000 bits in (d) 24 and 23
stop-and-wait ARQ.
(a) 2 (b) 4 (c) 7 (d) 9 24. In ICMP protocol, error reporting message type 11
denotes which of the message type?
18. If 5-bit sequence number is used, what is the
(a) Destination unreachable
sequence number after sending 100 frames in
(b) Source quench
Go-Back-N ARQ?
(c) Time exceeded
(a) 2 (b) 3 (c) 4 (d) 5 (d) Parameter problem
19. If N is a maximum sequence number in sliding 25. In a token ring network, the transmission speed is
window of Go-Back-N ARQ. How many sequence 32 bps and the propagation speed is 200 m/ms. The
bit will be there for use? 4-bit delay in this network is equivalent to:
(a) log2N (b) 1/log2N (a) 12 m of cable (b) 18 m of cable
(c) log2N2 (d) log2N + 1 (c) 22 m of cable (d) 25 m of cable
Set 1
1. (d) 8. (b) 15. (c) 22. (a) 29. (c) 36. (b) 43. (c) 50. (a)
2. (b) 9. (16) 16. (b) 23. (a) 30. (c) 37. (d) 44. (c) 51. (b)
3. (c) 10. (a) 17. (a) 24. (d) 31. (a) 38. (b) 45. (b) 52. (b)
4. (d) 11. (a) 18. (d) 25. (c) 32. (a) 39. (a) 46. (c) 53. (b)
5. (c) 12. (b) 19. (a) 26. (b) 33. (a) 40. (d) 47. (a)
6. (a) 13. (b) 20. (c) 27. (a) 34. (c) 41. (b) 48. (d)
7. (a) 14. (4) 21. (c) 28. (a) 35. (d) 42. (c) 49. (b)
Chapter 10.indd 532 4/9/2015 10:08:59 AM

Set 2 8. (c) Because of duplex link we need not wait for

twice the propagation time for sending the frame.
1. (a) In case 1: when M (1010101) by G (1100), we If the sender window size is N, then
will get remainder which do not have all zeros. Transmitting 106 bits require = 1 s
In case 2: when M (111000111011) by G Therefore, N × 10−3 s = N ms
(11001), we will get all zeros in remainder. Therefore, N ms = 34 ms, N = 34 < 26
So, minimum number of bits required is 6.
2. (b) One network address: 216.18.9.21
9. (d) In CSMA/CD, the minimum frame size = 2 ×
Subnet mask: 255.255.255.240 z × bandwidth
By doing XOR between network id and
subnet mask, we get first address of block = Given that distance (d) = 3 km = 3 × 103 m, veloc-
216.18.9.16. ity (v) = 4 × 108 m/s and bandwidth = 107 bps.
Therefore, minimum pocket size
3. (a) Total number of addresses = 232 −28 = 24 = 16
= 2 × (7.5 × 10−6) × 107 = 2 × 7.5 × 10
4. (d) Transmission time =
Data size
=
360
= 36 ms
= 150 bits = 150/8 ≈ 19 bytes
Bandwidth 10000
10. (b) Number greater than 255 is not allowed in
Utilization in percentage
valid IP. In option (f), 3rd octave has 257 value
Transmission time
= × 100 which is not valid. So, option (f) should match (vi).
Transmission time + (2 × Propagation time) Similarly, option (c) should match (vi).
36
= × 100 = 78.2 ≈ 78% 11. (d) Subnetmask 255.255.255.224 − 11111111.
36 + 10 11111111. 11111111. 11100000
Data size 200 × 8 Number of subnet = 23 − 2 = 6
5. (c) Transmission time = =
Bandwidth 3 × 106 Number of host in each subnet = 25 − 2 = 30
= 0.00053 s Two addresses are reduced because one is for
Ring latency = Round-trip time = 150 ms = network address and another is for broadcasting.
0.00015 s 12. (c) Default mask of
Total time required to transfer 1600 bits = (0.00053 class C = 255. 255. 255. 0
+ 0.00015) = 0.00068 s 11111111 11111111 11111111 00000000
So, effective data rate = 1600/0.00068 = 2.35 Mbps
Required 16 class C blocks, so
6. (a) We have 11111111 11111111 11110000 00000000
Window size (n) = 6 packets We have used last 4 bit for combining 16 class C
Packet size = 1000 bytes blocks. So supernet mask is {255.255.240.0}
So, total packet size = 6 × 1000 = 6000 bytes 13. (400) RTT (round-trip time) = 2× PT (propaga-
Total time = Transmission time + Propagation time tion time)
= 6 × 40 + 180 ms = 420 ms = 420 × 10−6s
Distance 4 × 103
Total size Propagation time (PT) = =
Maximum achievable throughput = Velocity 2 × 108
Total time
6000 6000 × 106 = 2 × 10 -5 s = 20 µs
= =
420 × 10−6 420 RTT = 2 × PT = 2 × 20 = 40 ms
= 14.28 × 106 bps For 1 s, bits transmitted = 107 bits
So, for 40 ms, the buts transmitted = 40 × 10−6 ×
7. (b) Given round-trip delay (T) = 60 ms = 60 × 10−3s 107 = 400 bits
R = 128 Kbps = 128 × 103 bps
14. (400) Frame size = (2TP) × (Bandwidth) = 50 ms ×
L = R × T = 128 × 103 × 60 × 10−3 = 128 × 60
8 Mbps = 400 Kbits
128×60
So, optional window size (n) =
32× 8 15. (d) The Class B is defined as follows
7680 Network id = 16 bit
= = 30
256 Host id = 16 bit
Chapter 10.indd 533 4/9/2015 10:09:01 AM

Maximum number of subnets 19. (d) If sequence bits are N, then maximum sequence
= 27 − 2 = 128 − 2 = 126 number is = 2N− 1, so
Maximum number of hosts in each subnet Number of sequence bits = log2N + 1
= 216 − 7 − 2 = 29 − 2 = 510
20. (c) If sequence bits are n, then window size = 2n−1
16. (c) Generally, the number of addresses usable for
So, for 6-bit sequence number, window size = 26−1 =
addressing specific hosts in each network is always
25 = 32
2N − 2, where N is the number of bits for host id.
21. (b) Given that distance = 500 m and bandwidth =
The binary representation of subnet mask is 11111
10 Mbps
111.11111111.11111000.00000000.
Frame transmission time = 2 × Propagation delay
There are 21 bits set in subnet. So, 11 (i.e., 32 − 21)
Therefore,
bits are left for host ids.
Frame size Distance
Total possible values of host ids is 211 = 2048 . = 2×
Bandwidth Velocity
Out of 2048 values, 2 addresses are reserved.
Frame size 500 m
The address with all bits as 1 is reserved as broad- = 2×
cast address and address with all host id bits as 0 10 × 10 b/s
6
2 × 108 m/s
is used as network address of subnet. Frame size = 50 bits
17. (a) Given that bandwidth = 100 Mbps = 100 × 22. (a) Formula to calculate time = C + rS = MS
106 = 108 bits and propagation time = 100 ms where capacity (C) = 1.2 × 106, token generation
We know that rate (r) = 6 mb/s and output rate (M) = 8 × 106.
Putting values into formula, we get
RTT = 2 × PT = 200 ms
Data transfer in 1 s = 108 (1.2 × 106) + 6 × 106 × S = (8 × 106) S
106(1.2 + 6S) = (8 × 106) S
So, data transfer in 200 ms (in 1 RTT) = 200 × 10−6 ×
108 = 20000 bits 1.2
S= = 0. 6 s
Required RTT for transmitting 40000 bits = 2
40000/20000 = 2 23. (b) Packet size = (0000000000111111)2 = 63
18. (b) Sequence bits (n) = 5 Header size = HLEN × 4 = 1010 × 4 = (1010)2 ×
Maximum value of sequence number = 2n− 1 4 = 40
Packet size = Header + Payload
0 1 2 3 - 30 31 0 1 2 - 30 31 0 1 2 - 30 31 0 1 2 3 63 = 40 + Payload ⇒ Payload = 23
- - -
24. (c) Error message types
- - -
3→ destinations unreachable
- - -
4→ source quench
(0−30) = 31 frames 11→ time exceeded
12→ parameter problem
(31−29) = 31 frames
(30−28) = 31 frames 25. (d)
Total = 93 frames Given that R = 32 × 106 bps, B = 4, V = 200 m/ms;
d =?
Total frames left 6 and they will be counted as 29,
30, 31, 0, 1, 2. Bv 4 × 200
d= = = 25 m of cable
So, after 99th frame, sequence number will be 3. R 32 × 106 × 10−6
Chapter 10.indd 534 4/9/2015 10:09:03 AM

UNIT XI: WEB TECHNOLOGIES
2
Number of questions
Marks 1
1 Marks 2
0
2015 2014 2013 2012 2011 2010
Year Concept
2015 Unix Sockets, XML and HTML
2014 HTML Web page and http server
2013 Communication technology
2012 Nil
2011 HTML elements
2010 Client—server application
Chapter 11.indd 535 4/9/2015 10:12:01 AM

Chapter 11.indd 536 4/9/2015 10:12:01 AM
CHAPTER 11
WEB TECHNOLOGIES
Syllabus: Web technologies: HTML, XML, basic concepts of client—server computing.
11.1 INTRODUCTION Table 11.1 | Major milestones in development

of Web
To communicate over the Internet, Web servers and 1980 Tim Berners-Lee invents the WWW, at
Web browsers act as client-server programs. Web data CERN (the world famous nuclear research
is distributed over Internet using these programs. HTML lab at Switzerland).
is a markup language that is used for presentation 1990 Concepts like - HTTP, Web browser,
and interaction with people in Web browsers. Web Uniform Resource Identifier (URI) and HTML
servers use IP address and port numbers for identifi- 1993 Launch of the first graphical web browser,
cation. Using Web browser, data request is sent using named - MOSAIC at USA.
HTTP protocol. Apache, Tomcat and IIS are popular
Web-server programs, whereas Internet Explorer (IE), 1994 Hosting of the first International WWW
Chrome, Safari, Opera and Firefox are popular Web Conference, formation of W3C
browsers. 1996 Commercialization of the Web
The World Wide Web (WWW), also termed as `web’, 1998 Google was founded by Larry Page and
is an information medium which is accessed through the Sergey Brin
Internet. The following table describes the major mile- 1999 Concept of Dot-com, its boom and bust.
stones in the development of the WWW (Table 11.1).
(Continued )
Chapter 11.indd 537 4/9/2015 10:12:01 AM

538 Chapter 11: Web Technologies
Table 11.1 | Continued Table 11.2 | HTML tags
2002 Launch of Web 2.0 Tag Description

2004 Launch of FaceBook and use of Internet for <html> Defines an HTML document
social networking
<head> Defines the document’s
Future Semantic Web (Web 3.0) header, which can keep other
HTML tags such as <title>,
11.2 HTML <link>, etc.
<title> Defines the document title
HTML stands for hypertext markup language, widely <body> Defines the document’s body
used for the development of Web pages. It is used to <h1> - <h6> Defines headings 1 to 6
display text, images and other objects in a specified <p> Defines a paragraph
way on the browser. Hypertext signifies the technique
of linking HTML documents together. The link available <br> Inserts a single line break
on a Web page is called hypertext. Markup language <center> Defines a centering content
signifies the marking of a text document with tags, which <hr> Defines a horizontal rule
are used to display on Web browsers. It was designed
<pre> Preserves formatting
in 1991 by Berners-Lee, and later, it was published in
1995 with standard specifications. The latest version  Defines a comment
<p> Defines a paragraph <table> Defines table in html
<br> Inserts a single line break <tr> Define rows in table
<center> Defines a centering content <td> Defines columns in table
<hr> Defines a horizontal rule
Important Abbreviations
API Application Program Interface OLTP OnLine Transaction Processing

ASP Active Server Pages RIA Rich Internet Application
CORBA COmmon Resource Broker RSS Rich Site Summary or Really
Achitecture Simple Syndication
CSS Cascading Style Sheets SGML Standard Generalized Markup
DOM Document Object Model Language
DTD Document Type Definition SVG Scalable Vector Graphics
EJB Enterprise Java Beans UIML User Interface Markup Language
HTML HyperText Markup Language URI Uniform Resource Identifier
HTTP/HTTPS HyperText Transfer URL Uniform Resource Locator
Protocol / Secure WML Wireless Markup Language
J2EE Java 2 Enterprise Edition XAL eXtensible Application Language
JAF Javabeans Activation Framework XAML eXtensible Application Markup
JMS Java Message Service Language
JNDI Java Naming and Directory XHTML Extensible HyperText Markup
Interface Language
JSP Java Server Pages XML Extensible Markup Language
JTA Java Transaction API XSL eXtensible Stylesheet Language
MathML Mathematical Markup Language XUL XML User Interface Language
MIME Multi-purpose Internet Mail
Extensions
SOLVED EXAMPLES
1. The firewall is used for 2. LDAP stands for

(a) Sensing the size of the packet. (a) lightweight digital audio protocol.
(b) Isolating intranet from extranet. (b) lightweight directory access protocol.
(c) Screening packets to/from the network and (c) large domain access protocol.
filtering of network traffic. (d) large data audio protocol.
(d) Isolating Internet from virtual LAN.
Solution: The LDAP stands for Lightweight
Solution: Firewall is used to prevent unauthor- Directory Access Protocol which is a directory ser-
ized access of private networks through packet vice protocol based on client-server model.
screening from outside user, especially Internet. Ans. (b)
Ans. (c)
Chapter 11.indd 548 4/9/2015 10:12:08 AM

SOLVED EXAMPLES 549
3. A digital signature is used to provide security Solution: <h1> will provide largest heading,
makes use of then <h2>, then <h3> and so on. <h6> gives the
(a) Digitally scanned signatures. smallest heading.
(b) A unique ASCII code number of the sender. Ans. (b)
(c) Private key encryption. 8. The following HTML tag is used for inserting a line
(d) Public key encryption. break?
Solution: Digital signature is an a symmetric (a) <br/> (b) <startbreak/>
cryptography technique which uses public key (c) <lb/> (d) <LINEBRK/>
encryption mechanism to provide integerity,
authentication and non-repudiation. Solution: <br/> tag is used for insert-
Ans. (d) ing a line break. There is no tag in html with
names<startbreak>, <lb/> and <LINEBRK>.
4. The <b> tag makes the enclosed text bold. The Ans. (a)
other alternative is to use
9. The correct syntax for adding yellow as a back-
(a) <strong> (b) <dark> ground colour in HTML is
(c) <big> (d) <highblack>
(a) <body style="background-color:yellow">
Solution: (b) <backgroundcolor>yellow</background
<strong> tag is used to define important text by color>
bold the text. (c) <color.background="yellow">
<dark> – there is no tag in html named dark. (d) <backgrndcolor="yellow">
<big> tag is used to make text bigger than the
normal text. Solution: Syntax for background color is:
<highblack> – there is no tag in html named 1. <body bgcolor="color_name|hex_number|
highblack. rgb_number">
Ans. (a) 2. <body style="background-color:color_
5. What does HTML stand for? name">
(a) High tool and text markup language Any of these two can be used.
(b) Heavy tool markup language Ans. (a)
(c) Hypertext markup language 10. The HTML tag used to make a text bold is
(d) Hypertext my language
(a) <b> (b) <heavybold>
Solution: HTML is a markup language which is (c) <blackbold> (d) <bld>
used to design web pages. It stands for Hypertext
Markup Language. Solution: <b> tag is used in HTML to make text
Ans. (a) bold.
Ans. (a)
6. Who describes and controls the making of various
Web standards? 11. The HTML tag used to make a text italic is
(a) The World Wide Web Consortium (a) <italic> (b) <i>
(b) IEEE and IETE (c) <textitalic> (d) <slantingtext>
(c) Administrator and Mozilla
(d) IEEE Solution: <i> tag is used in HTML to make text
italic. There is no tag in html with names<italic>,
Solution: The World Wide Web Consortium <textitalic> and <slantingtext>.
(W3C) is the international standard organisation Ans. (b)
for managing and controlling web standard. IEEE
12. The HTML tag used to create a hyperlink is
and IETE are professional association societies for
advancement of technology. (a) <ahref="http://www.myexams.com">
Ans. (a) MYEXAMS</a>
7. Which of the following HTML tag will provide the (b) <aurl="https://www.myexams.com">
largest heading myexams</a>
(c) <anamelink="http://www.myexams.com">
(a) <h6> (b) <h1> myexams.com</a>
(c) <head8> (d) <heading6> (d) <a>http://www.MYEXAMS.COM</a>
Chapter 11.indd 549 4/9/2015 10:12:08 AM

Solution: <a href =” URL addrss” hyperlink office systems (SGML) for text information
name </a> tag is used to create hyperlink in processing.
HTML. (a) ISO - 8878 (b) ISO - 8879
Ans. (a) (c) ISO - 8880 (d) ISO - 8881
13. Linking to another place in the same or another
Web page require two A (Anchor) tags, the first Solution: SGML is an ISO standard (ISO –
with the attribute and the second with 8879) for defining markup languages.
the attribute. Ans. (b)
(a) NAME, LINK 15. The facilities available in the Internet are
(b) LINK, HREF A. Electronic mail.
(c) HREF, NAME B. Remote login.
(d) TARGET, VALUE C. File transfer.
D. Encryption technique.
Solution: <a href =” addrss”> hyperlink
name </a> tag is used to create hyperlink in (a) A and B (b) A, B and D
HTML. (c) A, B and C (d) B, C and D
Ans. (c)
Solution: Internet provides services like e-mails,
14. HTML is defined using SGML - an file transfer and remote login; but it does not pro-
standard, information processing-text and vide encryption and decryption techniques.
Ans. (c)
1. Consider the HTML table definition given below: 3. HTML (HyperText Markup Language) has
<table border=1> language elements, which permit certain actions
<tr><td rowsp an=2>ab</td> other than describing the structure of the Web
document. Which one of the following actions
<tdcolspan=2> cd </td></tr>
is NOT supported by pure HTML (without any
<tr><td>ef</td><td rowspan=2>gh</td>
server or client-side scripting) pages?
</tr><tr><td colspan=2>ik</td></tr>
</table>
(a) E mbed Web objects from different sites into
The number of rows in each column and the the same page.
number of columns in each row are (b) Refresh the page automatically after a specified
(a) 2, 2, 3 and 2, 3, 2 (b) 2, 2, 3 and 2, 2, 3 interval.
(c) 2, 3, 2 and 2, 3, 2 (d) 2, 3, 2 and 2, 2, 3 (c) Automatically redirect to another page upon
(GATE 2009: 1 Mark) download.
(d) Display the client time as part of the page.
Solution: Two td commands are used in the first (GATE 2011: 1 Mark)
tr command and three td commands are used in
second tr command. So, it requires (2, 3, 2) rows
Solution: <OBJECT> … </OBJECT> tag is used to
in each column.
embed web objects.
Ans. (c)
<META HTTP-EQUIV="Refresh" CONTENT="5">
2. Which one of the following is not a client-server is used to refresh page after every 5 seconds.
application?
<META HTTP-EQUIV="Refresh" CONTENT="0;
(a) Internet chat (b) Web browsing URL=another-page.html"> is used to redirect.
(c) E-mail (d) Ping But for displaying the client time, there is no tag
(GATE 2010: 1 Mark) available.
Solution: Internet chat, Web browsing and email Ans. (d)
are client-server applications. Ping is a connectiv-
ity checking of the network. 4. Match the problem domains in Group I with the
Ans. (d) solution technologies in Group II.
Chapter 11.indd 550 4/9/2015 10:12:08 AM

5. A graphical HTML browser resident at a network

Group I Group II
client machine Q accesses a static HTML Web
(P) Services oriented (1) Interoperability page from a HTTP server S. The static HTML
computing page has exactly one static embedded image, which
(Q) Heterogeneous (2) BPMN is also at S. Assuming no caching, which one of the
communicating following is correct about the HTML Web page
systems loading (including the embedded image)?
(R) Information (3) Publish-find bind
(a) Q needs to send at least 2 HTTP requests to S,
representation
each necessarily in a separate TCP connection
(S) Process description (4) XML to server S.
(b) Q needs to send at least 2 HTTP requests to
(a) P-1, Q-2, R-3, S-4 S, but a single TCP connection to server S is
(b) P-3, Q-4, R-2, S-1 sufficient.
(c) P-3, Q-1, R-4, S-2
(c) A single HTTP request from Q to S is sufficient,
(d) P-4, Q-3, R-2, S-1
and a single TCP connection between Q and S
(GATE 2013: 1 Mark) is necessary for this.
Solution: Service oriented computing involves the (d) A single HTTP request from Q to S is suf-
operations such as find, publish and bind between ficient, and this is possible without any TCP
the entities (service consumer, service registry and connection between Q and S.
service provider). Solution: Client has to make atleast two http
Through interoperability, heterogeneous systems requests–one for the web page and another for
can communicate and work together. embedded image. One TCP connection is sufficient
XML is used to represent information in a specific for both requests because TCP connection is alive
format. for the request.
BPMN is used to represent business process Ans. (b)
description in a business process model.
Ans. (c)
PRACTICE EXERCISES
Set 1 5. DTD definition is used along with XML to specify

(a) the data types of the contents of XML document.
1. The importance of the markup tag helps the
(b) the front page of XML document.
browser
(c) the links of XML to other documents.
(a) how to orient the Web page. (d) the links of XML document from other documents.
(b) how to display the apps message.
6. XSL definition is used along with XML definition
(c) how to display the page.
to specify
(d) how to fetch a Web page.
(a) t he data structures and the links of XML
2. What is the default filename of a homepage of a
document.
website?
(b) the presentation of XML document.
(a) catch.html (b) index.html (c) the layered display of XML with other
(c) try.html (d) home.html documents.
3. The <Table> tag has which of the following
(d) the structure of XML document.
attributes? 7. Which of the following is a platform-free language?
(a) row (b) cellpadding (a) FORTRAN77 (b) PASCAL
(c) font (d) link (c) EIFFEL (d) JAVA
4. Which of the following is not a pair tag? 8. The importance of an HTML tag is
(a) <i> (b) <u> (a) It specifies the structure of your page.
(c) <b> (d) <img> (b) It specifies the connection to a particular page.
Chapter 11.indd 551 4/9/2015 10:12:08 AM

(c) It specifies the marking of the Web page. 17. How can you create an email link?
(d) It specifies formatting and layout instructions
(a) <ahref="mailto:xxx@yyy">
of the Web page.
(b) <ahref="xxx@yyy.zzz">
9. HTML is a subset of (c) <mailhref="xxx@yyy.abc">
(d) <mail>xxx@yyy.pqr</mail>
(a) SGML (b) SGHT
(c) STML (d) SHML 18. How can you open a link in a new browser window?
10. The format of an HTML document is (a) <ahref="url" target="newblank">
(b) <ahref="url" target="_blank">
(a) ASCII (b) Special JPEG
(c) <ahref="url" target="subwindow">
(c) Special TIFF (d) Special Web
(d) <
ahref="url" target="nextwindow">
11. The advantages of XML over HTML are:
19. What does CSS stand for?
A. It allows processing of data stored in Web
(a) Computer secure scheme.
pages.
(b) Computer securesheets.
B. It uses meaningful tags which aids in under-
(c) Cascading style sheets.
standing the nature of a document.
(d) Cascading secure sheets.
C. Is simpler than HTML.
D. It separates presentation and structure of 20. What is the correct HTML for referring to an
document. external style sheet?
(a) A, B and C only (b) A, C and D only (a) <link rel="stylesheet" type="text/
(c) A, B and D only (d) B and D only css" href="mystyle.css">
12. XML uses (b) <style src="mystyle.css">
(c) <stylesheet>mystyle.css</stylesheet>
(a) user-defined tags. (b) pre-defined tags. (d) None of these
(c) extensible tags. (d) pairing tags.
21. An external style sheet in an HTML document is
13. A URL specifies the following: placed
A. Protocol used (a) in the <body> section.
B. Domain name of server hosting Web page (b) in the <linker> section.
C. Name of folder with required information (c) in the <head> section.
D. Name of document formatted using HTML (d) in the <top> section.
(a) A, B, C and D (b) B, C and D 22. The HTML tag used to define an internal style
(c) B, C and D (d) A, C and sheet is
14. The World Wide Web supports (a) <style> (b) <stylesheet>
A. Encryption. (c) <css> (d) <internal link>
B. HTML. 23. Which of the following HTML attribute defines
C. HTTP. inline styles?
D. Firewall.
(a) style (b) inlinefont
(a) A, B and C (b) B and C (c) objectdefine (d) inlinestyles
(c) A, C and D (d) A, B, C and D
24. The correct CSS statement is
15. The time taken by Internet packets
(a) {color=black(body}
(a) can be predetermined before transmission. (b) body {color:black}
(b) depends upon the size of the packet. (c) body:color:=black
(c) is irrelevant for audio packets. (d) {body(color:black)}
(d) is irrelevant for video packets.
25. Which of the following inserts a comment in a CSS
16. The DNS maps the IP addresses to file?
(a) a binary address as strings. (a) // this is a comment
(b) an alphanumeric address. (b) /* this is a comment */
(c) a hierarchy of domain names. (c) * this is a comment*
(d) a hexadecimal address. (d) “this is a comment”
Chapter 11.indd 552 4/9/2015 10:12:09 AM

26. How can we change the background colour? 34. Which of the following statements is used for creat-
ing a function?
(a) bgcolor:"newcolor"
(b) color=background (colorname) (a) function=Functionname()
(c) background-color: (b) function Functionname()
(d) BGCOL="colorname" (c) function:Functionname()
(d) function "Functionname"
27. To add a background color for all <h1> elements,
which of the following HTML syntax is used? 35. How do you call a function named “myFunction”?
(a) h1 {background-color:#FFFFFF} (a) myFunction()
(b) {background-color:#FFFFFF}.h1 (b) call function myFunction
(c) {background-color:#FFFFFF}.h1(all) (c) callmyFunction()
(d) h1.all{bgcolor= #FFFFFF} (d) callfunction.myFunction
28. To change the text colour in HTML we use 36. The correct syntax to define “if” statement for
executing some code if “x” is equals to 2?
(a) color: (b) latest—text—color=
(c) modifytextcolor: (d) newcolor: (a) if(x=2) then (b) ifx==2.0 then
(c) if(x==2) (d) if(x:=2)
29. JavaScript is defined under which HTML
element? 37. The correct syntax for a conditional statement to
execute following code if “x” is NOT equal to 8?
(a) <jscript> (b) <script>
(c) <scriptjava> (d) <define.js> (a) if x =! 8 then (b) if (x<> 8)
(c) if x<>8 then (d) if (x != 8)
30. The correct syntax to write “Hi There” in
Javascript is 38. What does XML stand for?
(a) jscript.write("Hi There") (a) Exclusive modern links

(b) response.write("Hi There") (b) Extensible markup language
(c) print("Hi There") (c) Extra memory links
(d) print.jscript("Hi There") (d) Exit markup language
31. To define the use of JavaScript in an HTML docu-

ment, it is described in?
Set 2
(a) The <insertscript> section. 1. Which is required by the Internet?

(b) The <head> section.
(a) An international agreement to connect
(c) Both the <head> section and the <body>
computers
section.
(b) A local area network
(d) The <linkjscript> section.
(c) A commonly agreed set of rules to communi-
32. What is the correct syntax for referring to an exter- cate between computers
nal JavaScript called “ps.js”? (d) A wide area network
(a) <script type="text.javascript" 2. Which services are available on the World Wide
name="ps.js"> Web
(b) <script type="text/javascript"
A. Decryption
link="ps.js">
(c) <script
B. HTTP
type="text/javascript"
C. HTML
src="ps.js">
(d) <script
D. Firewalls
type="text/javascript"
external="ps.js"> (a) A and B (b) B and C
(c) B and D (d) A and D
33. Which of the following syntax is used to write “Hi”
in an alert box? 3. The HTML link properties are:
(a) msgBox("Hi") (a) :link, :visit, :hover, :active
(b) alert("Hi") (b) :link, :visited, :hover, :active
(c) alertmsgBox("Hi") (c) :linked, :visit, :over, :active
(d) msgalertBox="Hi"

(d) :link, :hover, :active, :inactive
Chapter 11.indd 553 4/9/2015 10:12:09 AM

4. Which of the following is used to specify border 6. Which of the following is TRUE in context of an
type (solid or dotted or double line etc.) XML language
(a) Every XML document must have a DTD.
(a) Border-style
(b) There is no difference in lower case and upper
(b) Border-layout
case tags.
(c) Border-attrib
(c) All documents must have a root element.
(d) Border-width
(d) Closing tag is not mandatory.
7. In an HTML document, the correct place to refer
5. To declare the version of XML, the correct syntax is.
to an external style sheet is
(a) <?xml version=’1.0’/> (a) In the <body> section.
(b) <*xml version=’1.0’/> (b) In <external> tag.
(c) <?xml version="1.0"/> (c) At the top of the document.
(d) </xml version=’1.0’/> (d) In the <head> section.
Set 1 8. (d) HTML is a markup language used by Web

browsers to represent text and graphics; it specifies
1. (c) A markup language, such as HTML, is used by formatting and layout instructions of the Web page.
Web browsers to display text and graphics. The
9. (a) HTML is a subset of SGML (Standard General
text includes markup tags such as <p> to indicate
Markup Language). XML is a highly functional
the start of a paragraph, and </p> to indicate the
subset of SGML. XHTML extends and subsets
end of a paragraph. HTML documents are often
HTML.
referred to as Web pages.
10. (a) The format of an HTML document is ASCII.
2. (b) A home page is the Web page that is chosen
11. (b)
by the Web server as a default page which appears
on accessing the top level of a Web site, without 12. (a)
specifying a particular file name. For example, if 13. (a) The URL consists of two parts separated by
someone visits http://www.home.com/, the Web a colon and two forward slashes–protocol identi-
server refers a list of default file names (such as fier which indicates the protocol to be used and
“index.html”, “index.htm”, etc.) and displays the resource name which specifies the IP address or
first matching file it finds. Hence, the following two the domain name of the resource.
addresses will work identically, and display the file
named “index.html”. 14. (b) WWW supports HTML, HTTP, XML,
XHTML.
http://www.home.com/
15. (c) 23. (a) 31. (c)
http://www.home.com/index.html
16. (c) 24. (b) 32. (c)
3. (b) Attributes of <Table> tag are align, bgcolor,
border, cellpadding, cellspacing, frame, rules, sort- 17. (a) 25. (b) 33. (b)
able, summary, width, etc. 18. (b) 26. (c) 34. (b)
4. (d) All tags that need to be closed are pair tags. 19. (c) 27. (a) 35. (a)
<img> need not to be closed. 20. (a) 28. (a) 36. (c)
5. (a) DTD is document type definition that gives 21. (c) 29. (b) 37. (d)
information about data types of content of docu- 22. (a) 30. (a) 38. (b)
ments it is used with.
6. (b) The Extensible Stylesheet Language Family Set 2

(XSL) represents the XML document and XLL
definition is used along with XML to specify the 1. (c) 4. (a) 7. (d)
links with other documents. 2. (b) 5. (c)
7. (d) JAVA is platform independent. 3. (b) 6. (c)
Chapter 11.indd 554 4/9/2015 10:12:09 AM

SOLVED GATE (CS) 2014
SET 1
(Engineering Mathematics and Technical Section)
Q. No. 1 — 25 Carry One Mark Each
1. Consider the statement: “Not all that glitters is Solution: Using topological sort, strongly con-
gold” Predicate glitters (x) is true if x glitters and nected components can be found. According to
predicate gold (x) is true if x is gold. Which one of the property, if (u,v) belongs to E2, (v,u) will not
the following logical formulae represents the above belong to E.
statement? Ans. (b)
(a) ∀x; glitters(x) ⇒ ¬gold(x) 4. Consider the following system of equations:
(b) ∀x; gold(x) ⇒ glitters(x) 3x + 2y = 1
(c) ∃x; gold(x) ∧ ¬glitters(x) 4x + 7z = 1
(d) ∃x; glitters(x) ∧¬ gold(x) x + y + z = 3
Solution: S: “Not all that glitters is gold” means x − 2y + 7z = 0
there exist some glitters that are not gold. So, ∃x is The number of solutions for this system is .
used along with AND operator.
∃x; glitters(x) ∧ ∼gold(x) Solution: The given equations are
3x + 2y = 1
Ans. (d)
4x + 7z = 1
2. Suppose you break a stick of unit length at a point x + y + z = 3
chosen uniformly at random. Then the expected x − 2y + 7z = 0
length of the shorter stick is . 3 2 0 1
4 0 7 1
Solution: The length of shorter stick ranges Augmented matrix r(A : B ) = 1 1 3 
between 0 and 0.5 meters. So, the expected length  1
is 0.25. 1 −2 7 0 
Ans. (0.25)
Applying row operations, the resultant matrix
3. Let G = (V,E) be a directed graph where V is the will be:
set of vertices and E the set of edges. Then which
one of the following graphs has the same strongly 1 1 1 3
 0 −4 3 −11
connected components as G? 0 0 −15 −21
(a) G1 = (V,E1) where E1 = {(u,v)  (u,v) ∉ E)} 
0 0 0 0 
(b) G2 = (V,E2) where E2 = {(u,v)  (v,u) ∉ E)}
(c) G3 = (V,E3)[ where E3 = {(u,v)  there is a ρ(A) = 3 = number of variables. So equations
path of length ≤ 2 from u to v in E} have a unique solution.
(d) G4 = (V,E4) where V4 is the set of vertices in Ans. (1)
G which are not isolated
GATE(CS) 2014 Set 1 solutions.indd 555 4/9/2015 10:25:02 AM

556 SOLVED GATE (CS) 2014
5. The value of dot product of the eigenvectors cor- Applying property [A + BC = (A + B)(A + C)],
responding to any pair of different eigenvalues we have
of a 4-by-4 symmetric positive definite matrix is
PQ + P Q (R+ R )(R + S)
.
⇒ PQ + P Q(R + S)
Solution: The Eigen for symmetric positive matrix ⇒ PQ + P Q R + P QS
⇒ Q(P + P R) + P QS
corresponding to different Eigen values are orthog-
onal. Hence, their dot product is zero always.
Ans. (0) ⇒ Q(P + R) + P QS
6. Let the function ⇒ QP + QR + P QS
sin q cos q tan q ⇒ Q(P + P )(P + S) + QR
f(q ) = sin(p /6) cos(p /6) tan(p /6) ⇒ PQ + QS + QR
sin(p /3) cos(p /3) tan(p /3)
Ans. (a)
p p 
where q ∈  ,  and f ′(q ) denote the derivative 8. The base (or radix) of the number system such
6 3 that the following equation holds is .
of f with respect to θ. Which of the following state-
ments is/are TRUE? Solution: (312/20)x = (13.1)x
p p 
(I) There exists q ∈  ,  such that f ′(q ) = 0
6 3 3x2 + x + 2
=
x2 + 3x + 1
⇒ x2 − 5x = 0 ⇒ x = 0, 5
p p  2x x
(II) There exists q ∈  3,x2 +such ′(q3)x≠+01
x + 2thatx2f+
6 3 = ⇒ x2 − 5x = 0 ⇒ x = 0, 5
2x x
(a) I only (b) II only
Base of number system cannot be 0. So, answer
(c) Both I and II (d) Neither I nor II is 5.
Ans. (5)
Solution: By applying mean value theorem, both
I and II found to be true. 9. A machine has a 32-bit architecture, with 1-word
Ans. (c) long instructions. It has 64 registers, each of which
is 32 bits long. It needs to support 45 instructions,
7. Consider the following Boolean expression for F:
which have an immediate operand in addition to
F(P, Q, R, S) = PQ + PQR + PQRS two register operands. Assuming that the immedi-
ate operand is an unsigned integer, the maximum
value of the immediate operand is .
The minimal sum-of products form of F is
(a) PQ + QR + QS (b) P + Q + R + S
Solution: Given that instruction size = 32 bits;
(c) P + Q + R + S (d) PR + PRS + P total instructions = 45, so opcode requires 6 bits;
Total registers = 64, so 6 bits for register operands
Solution: Immediate operand = 32 − 6 − 12 = 14
PQ + P QR + P Q R S So, the maximum value of the immediate operand
will be 16383
PQ + P Q(R + R S)
6 bits 6 bits 6 bits 14 bits
Opcode Reg operand 1 Reg operand 2 Immediate operand
32 bits
Ans. (16383)

SET 1 557
10. Consider the following program in C language: (a) T he graph does not have any topological
ordering
#include <stdio.h>
(b) Both PQRS and SRQP are topological orderings
main() (c) Both PSRQ and SPRQ are topological orderings.
{ (d) PSRQ is the only topological ordering.
int i;
Solution: Topological order of a directed graph is
int *pi=&i; the linear arrangement of vertices in which if there
scanf(“%d”, pi); is an edge between u to v, then u comes before v
printf(“%d \n”, i+5); in order. Option (c) satisfies this property, so both
} PSRQ and SPRQ are topological orderings.
Ans. (c)
Which one of the following statements is TRUE?
14. Let P be a quick sort program to sort numbers
(a) Compilation fails. in ascending order using the first element as the
(b) Execution results in a run-time error. pivot. Let t1 and t2 be the number of compari-
(c) On execution, the value printed is 5 more than sons made by P for the inputs [1 2 3 4 5] and [4
the address of variable i. 1 5 3 2], respectively. Which one of the following
(d) On execution, the value printed is 5 more than holds?
the integer value entered.
(a) t1 = 5 (b) t1 < t2
Solution: In the program, pi is a pointer variable
(c) t1 > t2 (d) t1 = t2
that contains the address of i. scanf gets the value
at address of i and printf prints value 5 more than i. Solution: Worst case of quick sort is when the
Ans. (d) list is already sorted. It requires n2 comparisons.
11. Let G be a graph with n vertices and m edges. Whereas an unsorted list requires nlogn compari-
What is the tightest upper bound on the running sons. List 1 is sorted, so the number of comparisons
time of Depth First Search on G, when G is repre- are more for list 1. Therefore, (c) is the correct
sented as an adjacency matrix? option (t1 > t2).
Ans. (c)
(a) Θ(n) (b) Θ(n + m)
15. Which one of the following is TRUE?
2
(c) Θ(n ) (d) Θ(m2)
(a) The language L = {an b n | n ≥ 0} is regular.
Solution: Tightest upper bound for running
(b) The language L = {an | n is prime} is regular.
depth first search using adjacency matrix is O(V2).
Number of vertices here are n. So, O(n2) is the (c) The language L = {w | w has 3k + 1b’s for some k Î N with å =
correct option. L = {w | w has 3k + 1b’s for some k Î N with å = {a,b}} is regular.
Ans. (c) (d) The language L = {ww | w ∈ ∑* with ∑ = {0,1}}
12. Consider rooted n node binary tree represented is regular.
using pointers. The best upper bound on the time Solution:
required to determine the number of sub trees (a) False. The language is not regular, because it
having exactly 4 nodes is Θ(nalogbn). Then the requires memory for remembering number of
value of a + 10b is a’s and b’s.
Solution: On taking input, the root of the binary (b) False. This is a recursive language.
tree prints all the subtrees of size 4 in O(n) time, (c) True. The language is regular because DFA
so a = 1, b = 0, and thus a + 10b = 1. construction is possible.
Ans. (1) (d) False. It is a context sensitive language.
Ans. (c)
13. Consider the directed graph given below.
16. Consider the finite automaton in the following
P Q figure.
0,1 1
R S 1 0,1 0,1
q0 q1 q2 q3
Which one of the following is TRUE?

What is the set of reachable states for the input scheduling the disk access, the request for cylinder
string 0011? 90 is serviced after servicing number
(a) {q1, q2, q3} (b) {q0, q1} of requests.
(c) {q0, q1, q2, q3} (d) {q3}
Solution: 30, 85, 90, 100, 105, 110, 135 and 145
Solution: With string `0011’, we can reach {q0, will be sort according to SSTF:
q1, q2} states. 100 →105 → 110 → 90 → 85 → 135 → 145 → 30
Ans. (a) 90 will be served after 3 requests.
Ans. (3)
17. Which one of the following is FALSE?
(a) A basic block is a sequence of instructions 20. Which one of the following is FALSE?
where control enters the sequence at the begin- (a) U ser level threads are not scheduled by the
ning and exits at the end. kernel.
(b) Available expression analysis can be used for (b) When a user level thread is blocked, all other
common subexpression elimination. threads of its process are blocked.
(c) Live variable analysis can be used for dead code (c) Context switching between user level threads
elimination. is faster than context switching between kernel
(d) x = 4 × 5 ⇒ x = 20 is an example of common level threads.
subexpression elimination. (d) Kernel level threads cannot share the code
Solution: x = 4 × 5 = 20 is not a common subex- segment.
pression elimination.
Solution: Code, text and data are sharable in kernel
Ans. (d)
threads. Only stack and registers are non-sharable.
18. Match the following Ans. (d)
(1) Waterfall model (a) S

pecifications can 21. Consider the relation scheme R = (E, F, G, H, I,
be developed J, K, L, M, N) and the set of functional dependen-
(2) Evolutionary model (b) Requirements cies {{E,F}→ {G},{F}→ {I,J}, {E,H} → {K,L} →
compromises are {M}, {K} → {M}, {L} → {N} on R. What is the
inevitable key for R?
(3) Component based (c) E
xplicit recognition (a) {E, F} (b) {E, F, H}
software of risk (c) {E, F, H, K, L} (d) {E}
(4) Spiral development (d) Inflexible
partitioning of the Solution: Find the closure. If it contains all the ele-
project into stages ments in its closure, then it called a candidate key.
{EFH} + = {E, F, G, H, I, J, K, L, M, N}
(a) 1-a, 2-b, 3-c, 4-d (b) 1-d, 2-a, 3-b, 4-c Ans. (b)
(c) 1-d, 2-b, 3-a, 4-c (d) 1-c, 2-a, 3-b, 4-d
22. Given the following statements:
Solution: Waterfall model: Inflexible partitioning S1: A foreign key declaration can always be
of the project into stages (backtracking is not replaced by an equivalent check assertion in SQL.
possible to previous phase)
Evolutionary model: Specifications can be S2: Given the table R(a,b,c) where a and b together
developed form the primary key, the following is a valid table
definition.
Component based software: Requirements compro-
mises are inevitable CREATE TABLE S (
Spiral development: Explicit recognition of risk a INTEGER,
Ans. (b) d INTEGER,

e INTEGER,
19. Suppose a disk has 201 cylinders, numbered from
0 to 200. At some time the disk arm is at cylinder PRIMARY KEY (d),
100, and there is a queue of disk access requests for FOREIGN KEY (a) references R)
cylinders 30, 85, 90, 100, 105, 110, 135 and 145. If Which one of the following statements is
Shortest-Seek Time First (SSTF) is being used for CORRECT?

SET 1 559
(a) S1 is TRUE and S2 is a FALSE 24. Which one of the following are used to generate a
(b) Both S1 and S2 are TRUE message digest by the network security protocols?
(c) S1 is FALSE and S2 is a TRUE
(P) RSA (Q) SHA-1 (R) DES (S) MD5
(d) Both S1 and S2 are FALSE
(a) P and R only (b) Q and R only
Solution: S1: Employee (Name, Dept_ID)
Department (Dept_Name, Deptid) (c) Q and S only (d) R and S only
In the given relation, Employee DeptID is a Solution: Only SHA-1 and MD5 are used as mes-
foreign key referencing Deptid (P.K) of relation sage digest.
Department. Ans. (c)
Now, declare the foreign key by an equivalent
check assertion as follows:- 25. Identify the correct order in which the following
actions take place in an interaction between a web
CREATE TABLE Employee (
browser and a web server.
Name Varchar (10)
1. The web browser requests a webpage using
DeptID INT (6) check (DeptID IN (select Deptid HTTP.
from Department)), 2. The web browser establishes a TCP connection
PRIMARY KEY (Name)); with the web server.
3. The web server sends the requested webpage
The above use of check assertion is good to declare using HTTP.
the foreign key as far as insertion is considered for 4. The web browser resolves the domain name
relation Employee. But it will fail to implement using DNS.
changes done in Department relation in terms of
deletion & updation. So, S1 is false. (a) 4, 2, 1, 3 (b) 1, 2, 3, 4
S2: The given table definition is not valid due (c) 4, 1, 2, 3 (d) 2, 4, 1, 3
to invalid foreign key declaration. Attribute a is
Solution: First of all, browser resolves IP address
declared as foreign key which is a single valued
using DNS server, then connects to web server using
attribute and it is referencing the primary key
TCP connection. After connection establishment,
(a b) of relation R (a, b, c), which is a composite
browser request for the page and web server responds
key. A single value attribute cannot refer a com-
with the page. So, option (a) is correct.
posite key. So, S2 is false.
Ans. (a)
23. Consider the following three statements about link
state and distance vector routing protocols, for a
large network with 500 network nodes and 4000 Q. No. 26 — 55 Carry Two Marks Each
links
[S1] The computational overhead in link state pro- 26. Consider a token ring network with a length of 2 km
tocols is higher than in distance vector protocols. having 10 stations including a monitoring station.
[S2] A distance vector protocol (with split horizon) The propagation speed of the signal is 2 × 108 m/s
avoids persistent routing loops, but not a link state and the token transmission time is ignored. If each
protocol. station is allowed to hold the token for 2 msec,
[S3] After a topology change, a link state protocol the minimum time for which the monitoring sta-
will converge faster than a distance vector protocol. tion should wait (in msec) before assuming that the
token is lost in .
Which one of the following is correct about S1, S2,
and S3? Solution: Length = 2 km, stations = 10, signal speed
(a) S1, S2, and S3 are all true = 2 × 108 m/s and token hold time = 2 ms
(b) S1, S2, and S3 are all false.
(c) S1 and S2 are true, but S3 is false Length 2000 m
(d) S1 and S3 are true, but S2 is false. RTT = = = 10 µs
Speed 2 × 108 m/s
Solution: A distance vector protocol (with split
Minimum wait time = RTT + Stations × Hold time
horizon) avoids persistent routing loops, but not a
link state protocol. So, S2 is false.
Ans. (d) Minimum wait time = 10 + 10 × 2 = 30 ms

27. Let the size of congestion window of a TCP con- 29. Consider the following four schedules due to three
nection be 32 KB when a timeout occurs. The transactions (indicted by the subscript) using read
round trip time of the connection is 100 msec and and write on a data item x, denoted r (x) and
the maximum segment size used is 2 KB. The time w (x) respectively. Which one of them is conflict
taken (in msec) by the TCP connection to get back serializable?
to 32 KB congestion window in .
(a) r1(x); r2(x); w1(x); r3(x); w2(x)
Solution: Given that RTT = 100 ms, congestion (b) r2(x); r1(x); w2(x); r3(x); w1(x)
window size = 32 KB (c) r3(x); r2(x); r1(x); w2(x); w1(x)
(d) r2(x); w2(x); r3(x); r1(x); w1(x)
Congestion window 32
Time out = = = 16 KB
2 2 Solution To test conflict stabilizability, make
precedence graph. If there is a cycle between any
and MSS = 2 KB. Now,
two nodes, then the given schedule is not conflict
2 KB → 1 RTT (Slow Start) serializable.
4 KB → 2 RTT
8 KB → 3 RTT
(a)
T1 T2 Cycle is there
16 KB → 4 RTT (Threshold)
T1 T2 Cycle
So, notisconflict
there
After threshold point is reached, there begins an T
T11 T
T22 Cycle is there
serializable
So, not
Cycle is conflict
there
additive increase phase. T3 So,
So, not conflict
serializable
not conflict
T3
18 KB → 5 RTT T
serializable
serializable
T33
20 KB → 6 RTT (b) T
1 T2
22 KB → 7 RTT T1 T2 Cycle is there
T
T11 T
T22 Cycle is there
˙ T3 Cycle
Cycle is
is there
there
˙
30 KB → 11 RTT T3
T
T33
32 KB T1 T2 Cycle is there
(c) Cycle is there
Therefore, total time = 11 RTTs T1 T2
= 11 × 10 T
T11 T
T22
Cycle
Cycle is
is there
there
= 1100 ms T3
Ans. (1100) T3
T
T33
28. Consider a selective repeat sliding window protocol T1 T2 No cycle, so, it is a
T T2 No cycle,serializable
conflict so, it is a
that uses a frame size of 1 KB to send data on a (d) T 1 T No cycle, so, it
1.5 Mbps link with a one-way latency of 50 msec. T11 T22 No it is
is aa
schedule
cycle,serializable
conflict so,
T3 conflict serializable
To achieve a link utilization of 60%, the minimum schedule
conflict serializable
T3 schedule
number of bits required to represent the sequence T schedule
T33
number field is .
Solution: Given Frame Size = 1 KB, link capacity Ans. (d)

= 1.5 Mbps, one-way latency = 50 msec
RTT = 2 × 50 = 100 msec and link utilization = 60% 30. Given the following two statements:
1000 × 8 S1: Every table with two single-valued attributes is

Transmission Time (T) = = 5.33 ms in 1NF, 2NF, 3NF and BCNF
1.5 × 10 6
n×T
S2: AB→C, D→E, E→C is a minimal cover for
Let window size = n then link utilization = the set of functional dependencies AB→C, D→E,
(T + RTT) AB→E, E→C
0.6(5.33 + 100) Which one of the following is CORRECT?
Therefore, n = = 12
5.33
(a) S1 is TRUE and S2 is FALSE.
Window size in selective repeat = 2n−1 = 12; (b) Both S1 and S2 are TRUE.
so n − 1 = 4 ⇒ n = 5 (c) S1 is FALSE and S2 is TRUE.
Ans. (5) (d) Both S1 and S2 are FALSE.

SET 1 561
Solution: REQ2:
S1: True, the highest normal form achieved by P1 (2,0,0)
any relation with two single valued attributes is With (3, 2, 0), request of P1 can be served. There
BCNF. is one safe sequence (P1, P2, P0). So, REQ2 can
S2: False, because the closure of AB contains be fulfilled.
different set of attributes. Ans. (b)
Ans. (a)
32. Consider the following set of processes that need
31. An operating system uses the Banker’s algorithm for to be scheduled on a single CPU. All the times are
deadlock avoidance when managing the allocation of given in milliseconds
three resource types X, Y, and Z to three processes
P0, P1, and P2. The table given below presents the
Process Arrival Execution
current system state. Here, the allocation matrix
Name Time Time
shows the current number of resources of each type
allocated to each process and the Max matrix shows A 0 6
the maximum number of resources of each type B 3 2
required by each process during its execution.
C 5 4
D 7 6
Allocation Max
X Y Z X Y Z E 10 3
P0 0 0 1 8 4 3
P1 3 2 0 6 2 0 Using the shortest remaining time first scheduling
algorithm, the average process turnaround time (in
P2 2 1 1 3 3 3
msec) is .
There are 3 units of type X, 2 units of type Y Solution
and 2 units of type Z still available. The system
is currently in a safe state. Consider the follow- A B A C E D
ing independent requests for additional resources
in the current state: 0 3 5 8 12 15 21
REQ1: P0 requests 0 units of X, 0 units of Y and

2 units of Z Average turnaround time =
REQ2: P1 requests 2 units of X, 0 units of Y and (8 − 0) + (5 − 3) + (12 − 5) + (21 − 7) + (15 − 10)
0 units of Z
5
Which one of the following is TRUE? 36
= = 7.2 ms
(a) Only REQ1 can be permitted. 5
(b) Only REQ2 can be permitted. Ans. (7.2)
(c) Both REQ1 and REQ2 can be permitted.
(d) Neither REQ1 nor REQ2 can be permitted. 33. Assume that there are 3 page frames which are
initially empty. If the page reference string 1, 2, 3,
Solution: 4, 2, 1, 5, 3, 2, 4, 6, the number of page faults using
the optimal replacement policy is .
Allocation Max Need Available
XYZ XYZ XYZ XYZ Solution:
003 843 840 320 1, 2, 3, 4, 2, 1, 5, 3, 2, 4, 6
320 620 300 3 4 4 4 4 4 4 4

211 333 122 2 2 2 2 2 2 2 2 6
1 1 1 1 1 1 5 3 3 3
REQ1: F F F F H H F F H F
P0 (0, 0, 2)
With (3, 2, 0), only request of P1 can be served. So, Total page faults = 7
REQ1 cannot be fulfilled. Ans. (7)

34. A canonical set of items is given below bags 1 to 5. Their total weight comes out to be
s → L. > R 323 gm. Then the product of the labels of the bags
having 11 gm coins is .
Q → R.
On input symbol < the set has Solution From the given information, we have
(a) a shift-reduce conflict and a reduce-reduce a + 2b + 4c + 8d + 16e = 323

conflict.
(b) a shift-reduce conflict but not a reduce-reduce As other terms are multiples of 2, so a = odd = 11.
conflict. Substituting, we get
(c) a reduce-reduce conflict but not a shift-reduce
conflict.
(d) neither a shift-reduce nor a reduce-reduce 2b + 4c + 8d + 16e = 312
conflict. Þ b + 2c + 4d + 8e = 156
Solution: Input symbol < is not available in the So, b must be even = 10. Solving in similar manner,
given grammar. So, we cannot comment about the we find that c = 11, d = 11 and e = 10. So, the
conflict that has arose. Option (d) is the correct product of the labels of the bags having 11 gm
answer. coins is 1 × 3 × 4 = 12.
Ans. (d) Ans. (12)
35. Let L be a language and L be its complement.

38. Suppose a polynomial time algorithm is discovered
Which of the following is NOT a viable possibility?
that correctly computes the largest clique in a given
(a) Neither L nor L is recursively enumerable graph. In this scenario, which one of the following
(r.e.). represents the correct Venn diagram of the com-
(b) One of L and L is r.e. but not recursive; the plexity classes P, NP and NP Complete (NPC)?
other is not r.e.
(c) Both L and L are r.e. but not recursive.
(d) Both L and L are recursive. (a) NP NP (b)
P P P P NP NP
Solution: Recursively enumerable languages are NP NP
not closed under complement. Thus, both L and L P P P NP NP
P
cannot be both recursive enumerable. NPC NPC
Ans. (c) NPC NPC
P = NP
NPC NPC P = NP
36. Which of the regular expressions given below rep- P =
NPC NP = = NP = NPC
NPC
PNPC
resent the following DFA? (c) P = NP
NPC NPC P = NP (d) P = NP = NPC
P = NP = NPC
(I) 0*1(1+00*1)* NPC NPC
(II) 0*1*1+11*0*1
(III) (0+1)*1
(a) I and II only (b) I and III only
(c) II and III only (d) I, II, and III
Solution: The given DFA accepts all the strings Solution Largest clique problem is known
ending with any number of 1’s. All the three NP-Complete problem. In the question, it is given
expressions are ending with 1 and are accepted by that there exists a polynomial time algorithm for larg-
the given DFA. est clique problem. Thus, P = NP = NP-Complete.
Ans. (d) Ans. (d)
37. There are 5 bags labeled 1 to 5. All the coins in a
given bag have the same weight. Some bags have 39. The minimum number of comparisons required to
coins of weight 10 gm, others have coins of weight find the minimum and the maximum of 100 num-
11 gm. I pick 1, 2, 4, 8, 16 coins, respectively, from bers is .

for(i = 0; i < size; i++)
Y = Y + E[i];
for(j = i; j < size; j++)
{
Z = 0;
for(k = i; k <= j; k++) SET 1 563
Z = Z + E[k];
Solution Number of comparisons for finding if (Z > Y)
3 Y = Z;
minimum and maximum of n numbers = n−2 }
2
return Y;
Here n = 100, so (300/2) − 2 = 148
}
Ans. (148)
40. Consider a hash table with 9 slots. The hash func- The value returned by the function MyX is the
tion is h(k) = k mod 9. The collisions are resolved (a) m aximum possible sum of elements in any sub-
by chaining. The following 9 keys are inserted in array of array E.
the order: 5, 28, 19, 15, 20, 33, 12, 17, 10. The (b) maximum element in any sub-array of array E.
maximum, minimum, and average chain lengths in (c) sum of the maximum elements in all possible
the hash table, respectively, are sub-arrays of array E.
(d) the sum of all the elements in the array E.
(a) 3, 0, and 1 (b) 3, 3, and 3
(c) 4, 0, and 1 (d) 3, 0, and 2
Solution: The following code calculates sum of all
the elements in array.
Solution:
Length Y = Y + E[i];
0 0 This Z = Z + E[k] calculates the sum of sub-array
1 28 19 10 × 3 and compare it with Y. If sum is greater, then it is
2 20 × 1 assigned to the Y. Hence, it calculates maximum
3 12 × 1 possible sum of any subarray.
4 0 Ans. (a)
5 5 × 1
6 15 33 × 2 42. Consider the following pseudo code. What is the
total number of multiplications to be performed?
7 0
17 ×
8 1 D = 2
for i = 1 to n do
9 for j = i to n do
for k = j + 1 to n do
Thus, average length =
9
=1 D = D * 3
9
Ans. (a) (a) Half of the product of the 3 consecutive integers.
(b) One-third of the product of the 3 consecutive
41. Consider the following C function in which size is integers.
the number of elements in the array E: (c) One-sixth of the product of the 3 consecutive
integers.
int MyX(int *E, unsigned int size) (d) None of the above.
{
int Y = 0; Solution:
int Z;
int i, j, k; ∑ ( n − 1) + ∑ ( n − 2) + ∑ 1 = n(n − 1)(n + 1)
for(i = 0; i < size; i++) 6
Y = Y + E[i]; Ans. (c)
for(j = i; j < size; j++) 43. Consider a 6-stage instruction pipeline, where all
{ stages are perfectly balanced. Assume that there
Z = 0; is no cycle-time overhead of pipelining. When an
for(k = i; k <= j; k++) application is executing on this 6-stage pipeline,
Z = Z + E[k]; the speedup achieved with respect to non-pipelined
if (Z > Y)
Y = Z;
}
return Y;
}

execution if 25% of the instructions incur 2 pipeline 46. The function f (x) = x sin x satisfies the follow-
stall cycles is . ing equation. f ˝(x) + f(x) + t cosx = 0. The value
of t is .
Solution: For non-pipeline 6 stages = 6 cycles are
required Solution:
f (x) = x sinx
Pipeline time = 1 + (25/100) × 2 = 1.5
f ˝(x) + f (x) + t cosx = 0
f ´(x) = sinx + x cosx = > f ˝(x) = cosx + x(−sinx)
Non-pipelined 6
Speed-up = = =4
Pipeline time 1.5 + cosx = 2cosx − x sinx
Ans. (4) 2cosx − x sinx + x sinx + tcosx = 0
44. An access sequence of cache block addresses is of
2cosx + tcosx = 0 ⇒ t = −2
length N and contains n unique block addresses.
The number of unique block addresses between Ans. (−2)
two consecutive accesses to the same block address
is bounded above k. What is the miss ratio if 47. A function f(x) is continuous the interval [0,2]. It is
the access sequence is passed through a cache of known that f (0) = f (2) = −1 and f (1) = 1.
associativity A ≥ k exercising least-recently-used Which one of the following statements must be true?
replacement policy?
(a) T here exists a y in the interval (0,1) such that
(a) n/N (b) l/N (c) 1/A (d) k/n f (y) = f (y + 1)
(b) For every y in the interval (0,1), f (y) = f (2 − y)
Solution: The miss ratio is n/N. (c) The maximum value of the function in the
Ans. (a) interval (0.2) is 1
(d) There exists a y in the interval (0,1) such that
45. Consider the 4-to-1 multiplexer with two lines S1
f (y) =f (2 − y)
and S0 given below.
Solution: Let g(x) = f (x) − f (x + 1) defined in [0,1]
0 0 g(0) = f (0) − f (1) = −1 − 1 = −2

1 1 4 - to - 1
F g(1) = f (1) − f (2) = 1 + 1 = +2
R 2 Multiplexer
g(0) is negative and g(1) is positive. Therefore,
R 3 S S there exists some x for which g(x) = 0
1 0
⇒ g(x)= 0 = f (x) − f (x+1) ⇒ f (x) = f (x + 1)
P Q Ans. (a)
The minimal sum-of-products form of the Boolean 48. Four fair six-sided dice are rolled. The probability
expression for the output F of the multiplexer is that the sum of the results being 22 is X/1296. The
value of X is .
(a) PQ + QR + PQR
Solution Sum 22 comes with following possibilities:
(b) PQ + PQR + PQR + PQR
6 + 6 + 6 + 4 = 2 2 (numbers can be arranged in
(c) PQR + PQR + QR + PQR 4 ways)
6 + 6 + 5 + 5 = 22 (numbers can arranged in 6 ways)
(d) PQR
Therefore, the total ways X = 4 + 6 = 10
Solution: PQ0 + PQ1 + PQR + PQR Ans. (10)
PQ + PQR + PQR 49. A pennant is a sequence of numbers, each number

Q(P + PR) + PQR being 1 or 2. An n-pennant is a sequence of
numbers with sum equal to n. For example, (1,1,2)
Q(P + P)(P + R) + PQR is a 4-pennant. The set of all possible 1-pennants
QP + QR + PQR is {(1)}, the set of all possible 2-pennants is {(2),
Ans. (a) (1,1)} and the set of all 3-pennants is {(2,1), (1,1,1),

SET 1 565
(1,2)}. Note that the pennant (1,2) is not the same (c) (3,3,3,1,0,0) (d) (3,2,1,1,1,0)
as the pennant (2,1). The number of 10-pennants 2,2,0,0,0 1,0,0,1,0 ← Reorder
1,−1,0,0 ← not graphic
is .
1,1,0,0,0
Solution 10 pennants: 1 + 9 + 28 + 35 + 15 + 0, 0, 0, 0
1 = 89 Ans. (c)
Ans. (89)
53. Which one of the following propositional logic for-
mulas is TRUE when exactly two of p, q, and r are
50. Let S denote the set of all functions f: {0,1} 4
TRUE?
→{0,1}4. Denote by N the number of functions
from S to the set {0,1}. The value of log2 log2 N is (a) ((p↔q)∧ r)∨(p ∧ q ∧ ~ r)
. (b) (~ (p↔q)∧ r)∨(p ∧ q ∧ ~ r)
(c) ((p→q)∧ r)∨(p ∧ q ∧ ~ r)
Solution: If a function is defined A → B, then
(d) (~ (p↔q)∧ r) ∧ (p ∧ q ∧ ~ r)
Size of A = |A| and Size of B = |B|
A = {0,1}4 ⇒ Number of elements in A, that is, Solution: Consider P = R = true and Q = false.
All the options other than B will become false.
|A| = 216
Number of functions from A → B, N = 2|A|
Ans. (b)
Therefore, log2 log2 2|A| = log2 |A| = log2 216 = 16

54. Given the following schema:
employees(emp-id, first-name, last-name,
Ans. (16) hire-date, dept-id, salary)
departments(dept-id, dept-name,
51. Consider an undirected graph G where self-loops manager-id, location-id)
are not allowed. The vertex set of G is {i,f): 1 ≤ i
≤ 12, 1 ≤ j ≤ 12}. There is an edge between (a,b) You want to display the last names and hire dates
and (c,d) if |a − c| ≤ 1 and |b − d| ≤ 1 The of all latest hires in their respective departments
number of edges in the graph is . in the location ID 1700. You issue the following
query:
Solution According to the given graph:
4 vertices have degree 3, 40 vertices with degree 5 SQL>SELECT last-name, hire-date
and 100 vertices with degree 8. FROM employees
Total degree = 12 + 200 + 800 = 1012 WHERE (dept-id, hire-date) IN
(SELECT dept-id, MAX(hire-date)
So, 1012 = 2 edges ⇒ edges = 506
Ans. (506) FROM employees JOIN departments
USING(dept-id)
WHERE location-id = 1700
52. An ordered n-tuple (d1, d2, …, dn) with d13 d2 …, 3dn
GROUP BY dept-id);
is called graphic if there exists a simple undirected
graph with n vertices having degrees d1, d2, …,
What is the outcome?
dn, respectively. Which of the following 6-tuples is
NOT graphic? (a) It executes but does not give the correct result.
(b) It executes and gives the correct result.
(a) (1, 1, 1, 1, 1, 1) (b) (2, 2, 2, 2, 2, 2) (c) It generates an error because of pairwise
(c) (3, 3, 3, 1, 0, 0) (d) (3, 2, 1, 1, 1, 0) comparison.
(d) It generates an error because the GROUP BY
Solution Applying Havel Hakimi theorem clause cannot be used with table joins in a
(a) (1,1,1,1,1,1) (b) (2,2,2,2,2,2) sub-query.
0,1,1,1,1 ← Reorder 1,1,2,2,2 ← Reorder Solution: Inner query joins employees and depart-
1, 1, 1, 1, 0 2, 2, 2, 1, 1 ments by using clause and apply filter location_id
0, 1, 1, 0 ← Reorder 1, 1, 1, 1 = 1700. Then the list will be grouped by dept_ids.
1, 1, 0, 0 0, 1, 1 ← Reorder This will select all the latest hire dates and dept_id.
Outer query will result into last name and hire date
0, 0, 0 1, 1, 0
of all the latest hires.
0, 0 Ans. (b)

55. Consider two processors P1 and P2 executing the Suppose P1 takes 5 cycles, then P2 takes 20% more,
same instruction set. Assume that under identical that is, 5 + 5 × (20/100) = 6 cycles
conditions, for the same input, a program running P2 takes 25% less time = 5 − 5 × (25/100) = 3.75 ns
on P2 takes 25% less time but incurs 20% more CPI Let us say frequency of P2 = x GHz
(clock cycles per instruction) as compared to the pro- x GHz = x × 109 Hz
gram running on P1. If the clock frequency of P1 is 1 1 cycle ⇒ 1/x ns
6 cycles ⇒ 6/x ns = 3.75 ns
GHz, then the clock frequency of P2 (in GHz) is
Solution: Frequency of P1 = 1 GHz = 109 Hz Therefore x = 6/3.75 = 1.6 GHz
1 cycle ⇒ 1/109 s = 1 ns Ans. (1.6)

SET 2
1. The security system at an IT office is composed pieces of paper are kept in a box. One of the pieces
of 10 computers of which exactly four are work- is drawn at random from the box. The expected
ing. To check whether the system is functional, length of the word drawn is . (The
the officials inspect four of the computers picked answer should be rounded to one decimal place.)
at random (without replacement). The system is
Solution: The given 9 words are as follows:
deemed functional if at least three of the four com-
The, Quick, Brown, Fox, Jumps, Over, The, Lazy, Dog
puters inspected are working. Let the probability
that the system is deemed functional be denoted Words with length 3 are => THE, FOX, THE,
by p then 100p = . DOG (4)
Words with length 4 are => OVER, LAZY (2)
Solution: P is the probability that system is func-
Words with length 5 are => QUICK, BROWN,
tional means at least three systems are working,
JUMPS (3)
p will be calculated as:
Probability for drawn 3 length words = 4/9
P for at least 3 systems are working
Probability for drawn 4 length words = 2/9
All 4 systems working + 3 systems working Probability for drawn 5 length words = 3/9
=
Number of ways to pick 4 systems from 100



4 2
P that 3 systems are working Expected word length =  3 × + 4 × +
= (Number of systems not selected)  9 9
3 
× (Number of ways to pick 3
from 4 selected systems)
5× 
9 

= 6 × 4×
×3×2×4 = 3.88
Ans. (3.88)
Therefore,
3. The maximum number of edges in a bipartite
P for at least 3 systems are working graph on 12 vertices is .
(4 × 3 × 2 × 1) + (6 × 4 × 3 × 2 × 4)
= = 0.1192 Solution: The maximum number of edges in
10 × 9 × 8 × 7
n2
So, 100P = 11.92 bipartite graph =
4
Ans. (11.92) For n = 12, maximum number of edges in bipartite
2. Each of the nine words in the sentence “The 12 × 12
quick brown fox jumps over the lazy dog” graph = = 36 edges
4
is written on a separate piece of paper. These nine Ans. (36)

4. If the matrix A is such that (a) k-bit binary up counter.

(b) k-bit binary down counter.
2
A = −4 [1 9 5]
(c) k-bit ring counter.
  (d) k-bit Johnson counter.
7 
Solution: In case of a k-bit ring encounter, single
Then the determinant of A is equal to . output will be 1 and remaining will be zero at a
time.
Given matrix A is as:
Ans. (c)
2 18 10  8. Consider the equation (123)5 = (x8)y with x and y

Determinate of A = − 4 − 36 − 20  ⇒

as unknown. The number of possible solutions is
 7 63 35  .
Solution: Converting both sides of (123)5 = (x 8)y

{2((−36 × 35) − (−20 × 63))} + 4((18 × 35) −
into decimal number system, we get
(10 × 63)) + 7((−20 × 18) − (−36 × 10))}
= {2(( −1260) − ( −1260)) + 4((630) − 1 × 52 + 2 × 51 + 3 × 50 = x × y1 + 8 × y 0
(630)) + 7(( −360) − ( −360))} 25 + 10 + 3 = xy + 8
= {2(0) + 4(0) + 7(0)} 38 = xy + 8 ⇒ xy = 30
=0
Factors of 30 are −1, 2, 3, 15, 30
Ans. (0) If x = 1 then y = 30
5. A non-zero polynomial f (x) of degree 3 has roots at If x = 2 then y = 15
x = 1, x = 2 and x = 3. Which one of the following If x = 3 then y = 10
must be TRUE? So, we have three solutions to the given problem.
Ans. (3)
(a) f (0) f (4) < 0 (b) f (0) f (4) > 0
(c) f (0) + f (4) > 0 (d) f (0) + f (4) < 0 9. A 4-way set-associative cache memory unit with
a capacity of 16 KB is built using a block size of
Solution: Given that f (x) has degree 3. For 8 words. The word length is 32 bits. The size of the
f (x) = 0, roots are x = 1, 2, 3 which lies between physical address space is 4 GB. The number of bits
0 and 4. So, f (0) and f (4) have opposite sign. for the TAG field is .
Therefore, f (0) f (4) should be less than zero.
Ans. (a) Solution:
A 4-way set associative cache memory has
6. The dual of a Boolean function F (x1 , x2 , … xn , +, , ’) Physical address size = 32 bits
written as F D, is the same expression as that of F Cache size = 16 K Bytes
with + and ⋅ swapped. F is said to be self-dual if
Word length = 32 bits = 4 Bytes
F = F D. The number of self-dual functions with n
Block size = 8 words = 8 × 4 = 32 Bytes
Boolean variables is
n
(a) 2n (b) 2n−1 (c) 22 (d) 22
n−1
Cache size 16 × 1024
Number of blocks = = = 29
Block size 32
Solution: A function is called self-dual if it has
equal number of maxterms and minterms. So, block of f set = 9 bits
n
Total function = 22 29
Number of sets = = 27
Number of mutual exclusive terms = 2n/2 4
Number of possible function by using minterms Set of f set = 7 bits
or maxterms taking any one at a time is (without
Byte of f set = 32 bytes = 25 bits = 5 bits
n−1
mutual excusive terms) = 22
Ans. (d)
TAG = {Physical address bits −
7. Let k = 2n. A circuit is built by giving the output
of an n-bit binary counter as input to an n-to-2n (Set of f set + Byte of f set)}
bit decoder. This circuit is equivalent to a = {32 − (7 + 5)} = 20 bits

SET 2 569
10. Consider the function func shown below: 1 and 7 are inserted into the heap in that order.
int func(int num) {
The level-order traversal of the heap after the
insertion of the elements is:
int count = 0;
(a) 10, 8, 7, 3, 2, 1, 5 (b) 10, 8, 7, 2, 3, 1, 5
while (num) {
(c) 10, 8, 7, 1, 2, 3, 5 (d) 10, 8, 7, 5, 3, 2, 1
count++;
num>>= 1; Solution:
} 10 10
return (count);
} 8 5 8 5
The value returned by func(435) is .

3 2 3 2 1
Solution: The given function has a while loop and
termination condition for loop is, until num will
not be zero. Function have a count variable which 10 10
will count number of times will execute. Every time
loop is executed, a right shift will be performed on 8 5 8 7
the number. Initially, num = 435, binary equiva-
lent = 110110011. 3 2 1 7 3 2 1 5
After first execution num = 011011001 and count = 1
After second execution num = 001101100 and Ans. (a)
count = 2 13. Which one of the following correctly determines the
After third execution num = 000110110 and count = 3 solution of the recurrence relation with T(1) = 1?
After eight execution num = 000000001 and count = 8
2 + log n
n
After nine execution num = 000000000 and count = 9 T (n) = 2T
The function will return count = 9.
Ans. (9) (a) Θ(n) (b) Θ(nlog n)
(c) Θ(n2) (d) Θ(log n)
11. Suppose n and p are unsigned int variables in a
C program. We wish to set p to nC3. If n is large, Solution:
which one of the following statements is most likely Applying Master Theorem,
to set p correctly? T(n) = aT(n/b) + f(n)
(a) p = n × (n − 1) × (n − 2)/6; T(n) = 2T(n/2) + logn;
(b) p = n × (n − 1)/2 × (n − 2)/3; a = 2, b = 2, f (n) = logn
(c) p = n × (n − 1)/3 × (n − 2)/2;
(d) p = n × (n − 1)/2 × (n − 2)/6.0; ⇒ g(n) = (n logb a ) = n log2 2 = Θ(n)
Solution: ⇒ f (n) = log n = O( n log2 2−e ),
P = nC3 = n × (n − 1) × (n − 2)/6 ⇒ f (n) is smaller than g(n) [as per Case 1 of
Master Theorem]
Computing n × (n − 1) × (n − 2) together can go ⇒ T(n) = Θ (n)
beyond the range; so, options (a) and (d) are not So, option (a) is correct.
true. If n is even or odd then n × (n − 1)/2 will Ans. (a)
always result in integer value, thus more accuracy)
whereas in the case of n × (n − 1)/3, it is not cer- 14. Consider the tree arcs of a BFS traversal from a
tain to get integer always so less accuracy. Hence, source node W in an unweighted, connected, undi-
option (b) is more accurate. rected graph. The tree T formed by the tree arcs is
Ans. (b) a data structure for computing
12. A priority queue is implemented as a Max-Heap. (a) the shortest path between every pair of vertices.
Initially, it has 5 elements. The level-order traversal (b) the shortest path from W to every vertex in
of the heap is: 10, 8, 5, 3, 2. Two new elements the graph.

(c) the shortest paths from W to only those nodes (c) Both + and * are right associative
that are leaves of T. (d) Both + and * are left associative
(d) the longest path in the graph.
Solution: From production rule P → Q + P, we
Solution: Application of BFS is to find the short- can decide + is right associative and from T →
est path from u to v. In the given statement, W is T * U, we conclude * is left associative.
the source node. It will calculate shortest path to Ans. (b)
every node from W. So, option (b) is correct.
Ans. (b) 18. Which one of the following is NOT performed
during compilation?
15. If L1 = {an | n ≥ 0} and L2 = {b n | n ≥ 0}, Consider
(a) Dynamic memory allocation
(I) L1⋅L2 is a regular language (b) Type checking
(II) L1⋅L2 = {an bn  n ≥ 0} (c) Symbol table management
(d) Inline expansion
Which one of the following is CORRECT?
Solution: Dynamic memory allocation is done at
(a) Only (I) (b) Only (II)
the run time, not at compile time.
(c) Both (I) and (II) (d) Neither (I) nor (II)
Ans. (a)
In option (1), concatenation of two regular lan-
19. Which one of the following is TRUE?
guages is regular (property).
But in option (2) L1⋅L2 = {an bn  n ≥ 0} is not (a) T he requirements document also describes how
regular because both the variables are independent the requirements that are listed in the docu-
in the given languages L1 and L2. So, only option ment are implemented efficiently.
(a) is correct. (b) Consistency and completeness of functional
Ans. (a) requirements are always achieved in practice.
16. Let A ≤ m B denotes that language A is mapping
(c) Prototyping is a method of requirements
validation.
reducible (also known as many-to-one reducible) to
(d) Requirements review is carried out to find the
language B. Which one of the following is FALSE?
errors in system design.
(a) If A ≤ m B and B is recursive then A is recursive.
Solution: Prototype is constructed to check that
(b) If A ≤ m B and A is undecidable then B is
requirements can be fulfilled or not.
undecidable.
Ans. (a)
(c) If A ≤ m B and B is recursively enumerable
then A is recursively enumerable. 20. A FAT (file allocation table) based file system is
(d) If A ≤ m B and B is not recursively enumerable being used and the total overhead of each entry in
then A is not recursively enumerable. the FAT is 4 bytes in size. Given a 100 × 106 bytes
disk on which the file system is stored and data
Solution: As language A is mapping reducible to block size is 103 bytes, the maximum size of a file
language B, we have f : ∈* → ∈* where for every w, w ∈ A ⇔ fthat
(w) ∈ B be stored on this disk in units of 106 bytes
can
∈* where for every w, w ∈ A ⇔ f (w) ∈ B is .
Hence, if A ≤ m B and B is not recursively enumer-
able then A is not recursively enumerable. Solution:
Ans. (d) Disk size 108
Number of entries = = 3 = 105
17. Consider the grammar defined by the following Block size 10
production rules, with two operators * and + Entry size = 4 B, total space consumed = 4B × 105
S → T *P Maximum file size can be stored = 108 − 0.4 × 106 =
99.6 × 106 B
T → U | T *U
Ans. (99.6)
P → Q+P |Q
Q → Id
21. The maximum number of super keys for the rela-
tion schema R (E, F, G, H) with E as the key is
U → Id .
Which one of the following is TRUE? Solution: Considering E as key, Total 23 = 8 keys
(a) + is left associative, while * is right associative are possible.
(b) + is right associative, while * is left associative Ans. (8)

SET 2 571
22. Given an instance of the STUDENTS relation as shown below:
Student D Student Name Student Email Student Age CPI

2345 Shankar Shankar@math X 9.4
1287 Swati Swati@ee 19 9.5
7853 Shankar Shankar@cse 19 9.4
9876 Swati Swati@mech 18 9.3
8765 Ganesh Ganesh@civil 19 8.7
For (StudentName, StudentAge) to be a key for ISP2

this instance, the value X should NOT be equal to ISP1
.
LAN L2
Solution: Key (Student Name, student Age) should R
always be unique. If X = 19, the duplicate value will S LAN L1
occur. This is not allowed. So, X cannot be 19.
Ans. (19)
Solution:
23. Which one of the following is TRUE about the
Initially, TTL = 32.
interior gateway routing protocols–Routing
Information Protocol (RIP) and Open Shortest Number of routers = 5 and Last is receiver.
Path First (OSPF)? So, TTL value will be 32 − 6 = 26 at R.
(a) RILP uses distance vector routing and OSPF Ans. (26)
uses link state routing
(b) OSPF uses distance vector routing and RIP
uses link state routing Q. No. 26 — 55 Carry Two Marks Each
(c) Both RIP and OSPF use link state routing
(d) Both RIP and OSPF use distance vector routing
26. Consider the store and forward packet switched
Solution: Distance vector routing technique is network given below. Assume that the bandwidth
used by RIP and OSPF uses link state routing. of each link is 106 bytes / sec. A user on host A
Ans. (a). sends a file of size 103 bytes to host B through
routers R1 and R2 in three different ways. In the
24. Which one of the following socket API functions first case, a single packet containing the complete
converts an unconnected active TCP socket into a file is transmitted from A to B. In the second
passive socket? case, the file is split into 10 equal parts, and these
(a) Connect (b) Bind packets are transmitted from A to B. In the third
(c) Listen (d) Accept case, the file is split into 20 equal parts and these
packets are sent from A to B. Each packet con-
Solution: Listen function converts an uncon- tains 100 bytes of header information along with
nected active TCP socket into a passive socket. the user data. Consider only transmission time and
Ans. (c) ignore processing, queuing and propagation delays.
Also assume that there are no errors during trans-
25. In the diagram shown below, L1 is an Ethernet mission. Let T1, T2 and T3 be the times taken
LAN and L2 is a Token-Ring LAN. An IP packet to transmit the file in the first, second and third
originates from sender S and traverses to R, as case, respectively. Which one of the following is
shown. The links within each ISP and across the CORRECT?
two ISPs, are all `point-to-point’ optical links. The
initial value of the TTL field is 32. The maximum
possible value of the TTL field when R receives the
. A R1 R2 B
datagram is

(a) T1 < T2 < T3 (b) T1 > T2 > T3 exactly one static embedded image which is also
(c) T2 = T3, T3 < T1 (d) T1 = T3, T3 > T2 at S. Assuming no caching, which one of the follow-
ing is correct about the HTML webpage loading
Solution: (including the embedded image)?
Given that bandwidth = 106 bytes/sec, file size =
1000 bytes, header size = 100 bytes (a) Q needs to send at least 2 HTTP requests to S,
each necessarily in a separate TCP connection
Case 1: A sends complete file to B
to server S
Total frame size = 1000 + 100 = 1100 bytes (b) Q needs to send at least 2 HTTP requests to
Transmission time = 1100/106 = 1100 ms S, but a single TCP connection to server S is
T1 = 3300 ms
sufficient
(c) A single HTTP request from Q to S is suffi-
Case 2: File size in chunks of 100 bytes cient, and a single TCP connection between Q
Total frame size = 100 + 100 = 200 bytes and S is necessary for this
Transmission time for 1 packet = 200/106 = 200 ms
(d) A single HTTP request from Q to S is suf-
ficient, and this is possible without any TCP
Transmission time for 10 packets = 2000 ms connection between Q and S
T2 = 2000 + 200 + 200 = 2400 ms Solution: Q sends at least 2 HTTP request to S,
Case 3: File size in chunks of 50 bytes single TCP connection is sufficient.
Total frame size = 50 + 100 = 150 bytes Ans. (b)
Transmission time for 1 packet = 150/106 = 150 ms 29. Consider the following schedule S of transactions
Transmission time for 20 packets = 3000 ms
T1, T2, T3, T4:
T3 = 3000 + 150 + 150 = 3300 ms T1 T2 T3 T4

T1 = T3 and T3 > T1
Writes [X] Reads[X]
Ans. (d)
Commit
27. An IP machine Q has a path to another IP machine Writes[X]
H via three IP routers R1, R2, and R3. Commit
Q–R1–R2–R3–H Writes[Y]
H acts as an HTTP server, and Q connects to H via Reads [Z]
HTTP and downloads a file. Session layer encryp- Commit
tion is used, with DES as the shared key encryp- Reads[X]
tion protocol. Consider the following four pieces of
information: Reads[Y]
Commit
[I1] The URL of the file downloaded by Q
[I2] The TCP port numbers at Q and H Which one of the following statements is
[I3] The IP addresses of Q and H CORRECT?
[I4] The link layer addresses of Q and H (a) S is conflict-serializable but not recoverable
(b) S is conflict-serializable but is recoverable
Which of I1, I2, I3, and I4 can an intruder learn
(c) S is both conflict-serializable and recoverable
through sniffing at R2 alone?
(d) S is neither conflict-serializable nor is it
(a) Only I1 and I2 (b) Only I1 recoverable
(c) Only I2 and I3 (d) Only I3 and I4
Solution:
Solution: Intruder can gain only I2 and I3 infor- T1 T2
mation. Because sniffing at R2 will able to capture
only the TCP port numbers which are encapsu-
lated in IP datagram and IP addresses of Q and H.
Ans. (c)
T4 T3
28. A graphical HTML browser resident at a network
client machine Q accesses a static HTML webpage As there is no cycle in the graph, it is conflict
from a HTTP server S. The static HTML page has serializable.

SET 2 573
All the transactions are reading committed values, 32. Three processes A, B and C each execute a loop of
so schedule is recoverable. 100 iterations. In each iteration of the loop, a process
Ans. (c) performs a single computation that requires tc CPU
milliseconds and then initiates a single I/O operation
30. Consider a join (relation algebra) between relations
that lasts for tio milliseconds. It is assumed that the
r(R) and s(S) using the nested loop method. There
computer where the processes execute has sufficient
are 3 buffers each of size equal to disk block size,
number of I/O devices and the OS of the computer
out of which one buffer is reserved for intermediate
assigns different I/O devices to each process. Also,
results. Assuming size(r(R)) < size(s(S)), the join
the scheduling overhead of the OS is negligible. The
will have fewer number of disk block accesses if
processes have the following characteristics:
(a) relation r(R) is in the outer loop.
(b) relation s(S) is in the outer loop. Process id tc tio
(c) join selection factor between r(R) and s(S) is A 100 ms 500 ms
more than 0.5. B 350 ms 500 ms
(d) join selection factor between r(R) and s(S) is
C 200 ms 500 ms
less than 0.5.
Solution: The relation r(R) should be in outer The processes A, B, and C are started at times
loop to have minimum number of disk accesses. 0, 5 and 10 milliseconds, respectively, in a pure
Ans. (a) time sharing system (round robin scheduling) that
uses a time slice of 50 milliseconds. The time in
31. Consider the procedure below for the Producer- milliseconds at which process C would complete its
Consumer problem which uses semaphores: first I/O operation is .
Semaphore n = 0; Solution: Time quantum = 50 ms
Semaphore s = 1; A B C A B C B C B C B
0 50 100 150 200 250 300 350 400 450 500 650
void producer ( ) void consumer() After completion of its CPU operations at 500 ms,
{ { C goes for I/O operations. This will last for 500 ms.
while (true) while(true) Hence, process C will complete its first I/O at
{ semWait(s); 500 + 500 = 1000 ms.
Produce(); semWait(n); Ans. (1000)
SemWait(s); removeFromBuffer(); 33. A computer has twenty physical page frames which
addToBuffer(); semsignal(s); contain pages numbered 101 through 120. Now a
semSignal(s); consume(); program accesses the pages numbered 1, 2, …,
100 in that order, and repeats the access sequence
semSignal(n); }
THRICE. Which one of the following page replace-
} } ment policies experiences the same number of page
} faults as the optimal page replacement policy for
this program?
Which one of the following is TRUE? (a) Least-recently-used (b) First-in-first-out
(a) The producer will be able to add an item to the (c) Last-in-first-out (d) Most-recently-used
buffer, but the consumer can never consume it.
Solution: Reference string = 1, 2, 3…. 100, 1, 2,
(b) The consumer will remove no more than one
3…. 100, 1, 2, 3…. 100
item from the buffer.
(c) Deadlock occurs if the consumer succeeds in In OPR technique, total page faults will be = 300
acquiring semaphore s when the buffer is empty. (100 every time)
(d) The starting value for the semaphore n must be This behaves same as most recently used, so num-
1 and not 0 for deadlock-free operation. bers of page faults are same.
Ans. (d)
Solution: When consumer gets semaphore s and
buffer is empty that means value of n is still zero. 34. For a C program accessing X[i] [j] [k], the follow-
Thus, consumer goes into infinite loop at second ing intermediate code is generated by a compiler.
step resulting in deadlock. Assume that the size of an integer is 32 bits and
Ans. (c) the size of a character is 8 bits.

t0 = i * 1024 between A and B and let y be the number of such

t1 = j * 32 longest common subsequences between A and B.
.
t2 = k * 4 Then x + 10y =
t3 = t1 + t0 Solution:
t4 = t3 + t2 Given that A = “qpqrr” and B = “pqprqrp”
t5 = X[t4] The LCS between A and B is having length 4, so
x = 4 and possible common subsequences are as
Which one of the following statements about the follows:
source code for the C program is CORRECT? (1) qpqr
(a) X is declared as “int X[32] [32] [8]”. (2) pqrr
(b) X is declared as “int X[4] [1024] [32]”. (3) qprr
(c) X is declared as “char X[4] [32] [8]”. So, y = 3. Hence, x + 10y = 4 + 10 × 3 = 34.
(d) X is declared as “char X[32] [16] [2]”. Ans. (34)
Solution: 38. Suppose P, Q, R, S, T are sorted sequences having
Let X be declared as X[A][B][C] lengths 20, 24, 30, 35, 50 respectively. They are
From t0 = i*1024, we have A * B = 1024. to be merged into a single sequence by merging
From t1 = j*32, we have C*(size of Type) = 32. together two sequences at a time. The number of
comparisons that will be needed in the worst case
From t2 = k*4, we have size of Type = 4. by the optimal algorithm for doing this is .
Solution: Merging two lists requires m + n − 1
Therefore, Type = int, A = 32, B = 32 and C = 8.
Ans. (a) comparisons.
35. Let <M> be the encoding of a Turing machine as 20, 24, 30, 35, 50
a string over S = {0,1}. Let L = {<M>  M is a Merge 20, 24 = 44 elements and 44 − 1 = 43
Turning machine that accepts a string of length comparisons
2014}. Then, L is Merge 30, 35 = 65 elements and 65 − 1 = 64
(a) decidable and recursively enumerable comparisons
(b) undecidable but recursively enumerable Merge 44, 50 = 94 elements and 94 − 1 = 93
(c) undecidable and not recursively enumerable comparisons
(d) decidable but not recursively enumerable Merge 94, 65 = 159 elements and 159 − 1 = 158
comparisons
Solution: Languages accepted by Turing machine Total comparisons = 43 + 64 + 93 + 158 = 358
are recursively enumerable. L is undecidable Ans. (358)
because Turing machine can halt or get stuck in
infinite loop. 39. Consider the expression tree shown. Each leaf
Ans. (b) represents a numerical value, which can either be
0 or 1. Over all possible choices of the values at the
36. Let L1 = {w ∈ {0,1}*w has at least as many leaves, the maximum possible value of the expres-
occurrences of (110)’s as (011)’s}. Let L2 = {w Œ sion represented by the tree is .
{0,1}*w has at least as many occurrence of (000)’s
as (111)’s}. Which one of the following is TRUE? Solution:
6
(a) L1 is regular but not L2 +
(b) L2 is regular but not L1
(c) Both L1 and L2 are regular 3 3
(d) Neither L1 nor L2 are regular − +
Solution: Only L1 is regular, construction of DFA
− −1
is possible for L1. L2 is context free grammar, 2 +
requires stack to remember count of 0’s and 1’s. − 1 + 2
Ans. (a)
37. Consider two strings A “= q pqrr" and B =
1 1 0 1 1 0 1 1
“pqprqrp". Let x be the length of the longest
common subsequence (not necessarily contiguous) Ans. (6)

{
static int i = 50;
int k;
if (i == j)
{
printf(“something”); SET 2 575
k = f(i);
40. Consider the following function return 0;
double f (double x) { }
if (abs (x*x – 3) < 0. 01) return x; else return 0;
else return f(x/2 + 1.5/x); }
} Which one of the following is TRUE?
Give a value q (to 2 decimals) such that f (q) will (a) The function returns 0 for all values of j.
return q: . (b) The function prints the string something for all
values of j.
Solution: (c) The function returns 0 when j = 50.
if(abs(p × p−3) < 0.01) return p ; (d) The function will exhaust the runtime stack or
run into an infinite loop when j = 50.
this will be true for x = 1.72
Solution: Condition i = = j becomes true for j = 50
Ans. (1.72) and program will go to infinite loop.
41. Suppose a stack implementation supports an Ans. (d)
instruction REVERSE, which reverses the order of 43. In designing a computer’s cache system, the cache
elements on the stack, in addition to the PUSH block (or cache line) size is an important param-
and POP instructions. Which one of the following eter. Which one of the following statements is cor-
statements is TRUE with respect to this modified rect in this context?
stack?
(a) A
smaller block size implies better spatial locality
(a) A queue cannot be implemented using this (b) A smaller block size implies a smaller cache tag
stack. and hence lower cache tag overhead
(b) A queue can be implemented where ENQUEUE (c) A smaller block size implies a larger cache tag
takes a single instruction and DEQUEUE takes and hence lower cache hit time
a sequence of two instructions. (d) A smaller block size incurs a lower cache miss
(c) A queue can be implemented where ENQUEUE penalty
takes a sequence of three instructions and
Solution: Smaller block size means more number
DEQUEUE takes a single instruction.
of blocks can be accommodated in cache. This
(d) A queue can be implemented where both
improves hit ratio and lower the miss penalty.
ENQUEUE and DEQUEUE take a single
Ans. (d)
instruction each.
44. If the associativity of a processor cache is dou-
Solution: bled while keeping the capacity and block size
An ENQUEUE & DEQUEUE operation takes four unchanged, which one of the following is guaran-
sequences of instructions: teed to be NOT affected?
Enqueue: Reverse, Push, Reverse
(a) Width of tag comparator
Dequeue: POP
(b) Width of set index decoder
(OR) (c) Width of way selection multiplexor
Enqueue: Push (d) Width of processor to main memory data bus
Dequeue: Reverse, POP, Reverse Solution: If associatively is doubled, then the
So, option (c) is correct. index bits will decrease and tag bits will increase.
Ans. (c) So, the options (a), (b) and (c) will get affected.
Ans. (d)
42. Consider the C function given below
45. The value of a float type variable is represented
int f(int j)
using the single-precision 32-bit floating point
{ format of IEEE-754 standard that uses 1 bit for
static int i = 50; sign, 8 bits for biased exponent and 23 bits for
int k; mantissa. A float type variable X is assigned the
if (i == j) decimal value of −14.25. The representation of X
{ in hexadecimal notation is
printf(“something”); (a) C1640000H (b) 416C0000H
k = f(i); (c) 41640000H (d) C16C0000H
return 0;
}
else return 0;
}

Solution: Solution:
Eigen vector is AX = l X
S E M
1 0 0 0 1
1 bit 8 bits 23 bits 0 1 1 1 0
 
(14.25)10 = (1110.01)2 A = 0 1 1 1 0
Now, normalize the binary number 1.11001 × 23 0 1 1 1 0
1 0 0 0 1 
E = e + Bias
Convert the given matrix in diagonal matrix. The
E = 3 + 127 = 130 diagonal entries are called Eigen values.
Binary equivalent of 130 = (10000010)2 Eigen values = 2, 3
Therefore, S = 1 (Since negative number) Product of Eigen values = 2 × 3 = 6
M = (11001000…) Ans. (6)
E = 10000010 48. The probability that a given positive integer lying
between 1 and 100 (both inclusive) is NOT divis-
1100 0001 0110 0100 0000 …. ible by 2, 3 or 5 is .
C 1 6 4 0 000
Solution: Let A event → Divisible by 2, B event →
Ans. (a) Divisible by 3, C event → Divisible by 5
46. In the Newton-Raphson method, an initial guess of P(A′ ∩ B′ ∩ C′) = 1 − P(A ∪ B ∪ C)
x0 = 2 is made and the sequence x0, x1, x2, … is P(A) = 50/100, P(B) = 33/100, P(C) = 20/100
obtained for the function 0.75x3 − 2x2 − 2x + 4 = 0 P(A ∩ B) = 16/100, P(B ∩ C) = 6/100,
P(A ∩ C) = 10/100, P(A ∩ B ∩ C) = 3/100
Consider the statements
P(A ∪ B ∪ C) = P(A) + P(B) + P(C) − P(A ∩ B) −
(I) x3 = 0.
P(B ∩ C) − P(A ∩ C) + P(A ∩ B ∩C)
(II) The method converges to a solution in a finite
number of iterations. P(A ∪ B ∪ C) = (50 + 33 + 20 − 16 − 6 − 10 +
3)/100 = 74/100
Which of the following is TRUE?
P(A′ ∩ B′ ∩ C′) = 1 − 74/100 = 26/100 = 0.26
(a) Only I (b) Only II
(c) Both I and II (d) Neither I nor II Ans. (0.259 to 0.261)
Solution: According to Newton—Raphson method: 49. The number of distinct positive integral factors of
.
x1 = x0 − f0 /f0′
2014 is
f (x) = 0.75x3 − 2x2 − 2x + 4

Solution:
2014 = 2 × 19 × 53 (product of prime factors)
f ′(x) = 2.25x2 − 4x − 2
x0= 2, f0 = −2, f0′ = −1 Number of distinct positive integral factors of 2014 =
(1 + 1) × (1 + 1) × (1 + 1) = 8
x1 = 2 − (−2)/(−1) = 0 Ans. (8)
f1 = 4, f1′ = −2 50. Consider the following relation on subsets of the
x2 = 0 − (4)/(−2) = 2 set S of integers between 1 and 2014. For two dis-
f2 = −2, f2′ = −1 tinct subsets U and V of S we say U < V if the
x3 = 2 − (−2)/(−1) = 0 minimum element in the symmetric difference of
the two sets is in U.
Hence, (I) is true. Also the value diverges, thus
option (II) is false. Consider the following two statements:
Ans. (a) S1: There is a subset of S that is larger than every
other subset.
47. The product of the non-zero eigenvalues of the
matrix S2: There is a subset of S that is smaller than every
1 0 0 0 1 other subset.
0 1 1 1 0
0 0 
Which one of the following is CORRECT?
 1 1 1
0 1 1 1 0 (a) Both S1 and S2 are true
1 0 0 0 1  (b) S1 is true and S2 is false
(c) S2 is true and S1 is false
is . (d) Neither S1 nor S2 is true

SET 2 577
Solution: 1 4
2 2
Both S1 and S2 are true. 2 1 2 1
2 1 2 1
Ans. (a)
51. A cycle on n vertices is isomorphic to2 its comple-
2 2 1 2 2 2 1
ment. The value of n is 1 . 1 1 1
Solution: 2 2 2 2
Consider n = 5, Applying Kruskal’s 5
2
2 2
V22 2 2 2
V2 V3 2 1 2V3 1
2 2 2 1
2 2 1 1
V1 2 1
V1 1 1
V4 2 2
V5 V4 V5 2 2 6
Complement 2
Cycle graph 3 2
of 2 1 2 1
2 cycle graph
1 2 1
By checking properties these two graphs are isomorphic. 2 2 2 1
2 2 2 1 1 1
1 1 Ans. (5)
2 2
52. The number of distinct minimum spanning trees
2 2 Hence, number of distinct minimum spanning trees
for the weighted graph below is are 6.
Ans. (6)
2
2 53. Which one of the following Boolean expressions is
1 2 1
NOT a tautology?
2 2 (a) ((a → b) ∧ (b → c)) → (a → c)

2 1 (b) (a ↔ c) → (~b → (a ∧ c))
1 1 (c) (a ∧ b ∧ c) → (c ∧ a)
(d) a → (b → a)
2 2
Solution: Tautology means all the output values
Solution: should be true. Hence, check the expressions by
1 11 4 44 trying all the possible combinations.
2 2 2 22
2 1 222 111 2 2 1 2 1 2 Ans. (b)
2 1 22 11 22 11
54. SQL allows duplicate tuples in relations, and corre-
2 2 22 22 2 1 22 11 2 2 22 22 2 spondingly
1 22
defines the multiplicity of tuples in the
11
1 1 11 11 1 1 11 11 result of joins. Which one of the following queries
always gives the same answer as the nested query
22 2 2 22 2 2
22 shown 22
below:
ApplyingApplying
Kruskal’sKruskal’s
Applying Kruskal’s 5 55
2
22 Select * from R where a in (select S. a from S)
2 22 2 2 2
2 22
2 2 22 2
2 (a) Select R. * from R, S where R. a = S. a
2 222 11 22 11
2 221 (b) Select distinct R. * from R, S where R. a = S. a
2 2 22 22 2
(c) S1elect
22 R. 11* from R, (select distinct a from S) as
2 2 22 22
2 1 22 111 1 11 11 S1 where R. a = S1. a
1 1 11 11 2 22 (d)
2 Select22R. * from R, S where R. a = S. a and is
2 22 2 622 unique R
66
2 22 Solution:
3 33 2 2 222 11 22 11 Let R
22 1 1
2 222 11
1 1 22 11
2 2 22 22 2 A 1 22 B
11
2 2 22 22 2 1 22 11 1 11 11 A1 3
1
1 1 11 11 A2
2 22 2 22 4
2 22 2 22 A3 4

S (d)
A c A b
A1 5 A1 3
A2 4 A2 4
A2 4 Hence, option (c) is correct.
A3 6 Ans. (c)
Query Output: 55. Consider a main memory system that consists of
8 memory modules attached to the system bus,
A b which is one word wide. When a write request is
A1 3 made, the bus is occupied for 100 nanoseconds (ns)
A2 4 by the data, address, and control signals. During
A2 4 the same 100 ns, and for 500 ns thereafter, the
addressed memory module executes one cycle
(a) accepting and storing the data. The (internal) oper-
A b ation of different memory modules may overlap in
A1 3 time, but only one request can be on the bus at
A2 4 any time. The maximum number of stores (of one
A2 4 word each) that can be initiated in 1 millisecond
is .
A2 4
(b) Solution:
A c For write request bus is occupied for = 100 ns
A1 3 For storing the data requires = 100 ns
A2 4 100 ns → 1 store
(c) 100/109 → 1 store
A b 1 sec → 107
A1 3 10−3 s → 10−3 × 107 stores = 104, that is, 10,000
A2 4 stores
A2 4 Ans. (10000)

SET 3
1. Consider the following statements: Solution:

P: Good mobile phones are not cheap Take example:
Q: Cheap mobile phones are not good X = {1, 2, 3} Y = {b, c}
L: P implies Q S = {1, 2} T = {b}
M: Q implies P Ans. (d)
N: P is equivalent to Q 3. Let G be a group with 15 elements. Let L be a sub-
Which one of the following about L, M, and N is group of G. It is known that L ≠ G and that the
CORRECT? size of L is at least 4. The size of L is .
(a) Only L is TRUE. Solution: According to Lagrange’s theorem, the

(b) Only M is TRUE. order of subgroup divides order of group. Order of
(c) Only N is TRUE. group is 15.
(d) L, M and N are TRUE. 3, 5, 15 divide 15. According to the given condition,
size should be atleast 4 and cannot be equal to that
Solution: of group. So, 5 is the order of subgroup.
P: good mobile phones are not cheap Ans. (5)
Q: Cheap mobile phones are not good. 4. Which one of the following statements is TRUE
P → Q and Q → P about every n × n matrix with only real eigenvalues?
Hence, P ↔ Q. So, L, M and N are true.
(a) If the trace of the matrix is positive and the
Ans. (d)
determinant of the matrix is negative, at least
2. Let X and Y be finite sets and f : X ® Y be a one of its eigenvalues is negative.
function. Which one of the following statements is (b) If the trace of the matrix is positive, all its
TRUE? eigenvalues are positive.
(c) If the determinant of the matrix is positive, all
(a) For any subsets A and B of X, |f (A ∪ B)| =
its eigenvalues are positive.
|f (A)| + |f (B)|
(d) If the product of the trace and determinant of
(b) For any subsets A and B of X, f (A ∩ B) =
the matrix is positive, all its eigenvalues are
f (A) ∩ f (B)
positive.
(c) For any subsets A and B of X, |f (A ∩ B)| =
min{|f (A)|, |f (B)|} Solution: Determinant of matrices = Product of
(d) For any subsets S and T of Y, f −1(S ∩ T) = Eigen values
f −1(S) ∩ f −1(T). Ans. (a)

5. If V1 and V2 are 4-dimensional subspaces of a f (x, y, a, b)

6-dimensional vector space V, then the smallest {
possible dimension of V1 ∩ V2 is . if (x is 1) y = a;
else y = b;
Solution: }
Let V be {e1, e2, e3, e4, e5, e6}
V1 = {e1, e3, e5, e6} Which one of the following digital logic blocks is
V2 = {e2, e3, e4, e5}
the most suitable for implementing this function?
Smallest possible set of V1 ∩ V2 = 2

(a) Full adder (b) Priority encoder
(c) Multiplexor (d) Flip—flop
Ans. (2)
2p Solution:
6. If ∫ x sin x dx = kp , then the value of k is equal y = x’ b + xa
0
to .
b
2:1 y
Solution: a
| sin x | = − sin x [from p to 2p ]
m 2p
= ò x sin x dx + ò −x sin x dx
x
Ans. (c)
0 p
= x(− cos x) − 1(− sin x) | −(−x cos x + sin x) | 9. Consider the following processors (ns stands for
= 4p nanoseconds). Assume that the pipeline registers
Ans. (4) have zero latency.
P1: Four-stage pipeline with stage latencies 1 ns,
7. Consider the following minterm expression for F. 2 ns, 2 ns, 1 ns.
F(P,Q,R,S) = ∑ 0, 2, 5, 7, 8, 10, 13, 15 P2: Four-stage pipeline with stage latencies 1 ns,
The minterms 2, 7, 8 and 13 are do not care terms. 1.5 ns, 1.5 ns, 1.5 ns.
The minimal sum-of-products form for F is P3: Five-stage pipeline with stage latencies 0.5 ns,
1 ns, 1 ns, 0.6 ns, 1 ns.
(a) Q S + Q S
P4: Five-stage pipeline with stage latencies 0.5 ns,
(b) Q S + QS 0.5 ns, 1 ns, 1 ns, 1.1 ns.
(c) Q RS + Q RS + Q R S + QRS Which processor has the highest peak clock
frequency?
(d) P QS + P Q S + PQ S + PQ S
(a) P1 (b) P2 (c) P3 (d) P4
Solution:
Solution:
PQ RS RS RS RS RS Maximum clock pulse will be of duration = 2 ns
for P1
PQ 1 X Minimum clock pulse duration = 1 ns for P3
PQ 1 X Frequency ∝ 1/Clock pulse
Highest frequency will be of P3.
PQ X 1 Ans. (c)
X 1
PQ 10. Let A be a square matrix size n × n. Consider
the following pseudocode. What is the expected
QS + Q S output?
Ans. (b)
C = 100;
8. Consider the following combinational function for i = 1 to n do
block involving four Boolean variables x, y, a, b for j = 1 to n do
where x, a, b are inputs and y is the output. {
Temp = A[i] [j] + C;
A[i][j] = A [j][i];
A[j][i] = Temp – C;
}
for i = 1 to n do
for j = 1 to n do
GATE(CS) 2014 Set 3 solutions.indd 580
output (A[i][j]); 4/9/2015 10:28:37 AM
C = 100;
for i = 1 to n do
for j = 1 to n do SET 3 581
{
Temp = A[i] [j] + C; The order in which the nodes are visited during an
A[i][j] = A [j][i]; in-order traversal of the tree is
A[j][i] = Temp – C; (a) SQPTRWUV (b) SQPTUWRV
} (c) SQPTWUVR (d) SQPTRUWV
for i = 1 to n do
for j = 1 to n do Solution:
output (A[i][j]); Left Root Middle Right
(a) The matrix A itself SQPTRWUV
(b) Transpose of the matrix A Ans. (a)
(c) Adding 100 to the upper diagonal elements and 13. Suppose depth first search is executed on the graph
subtracting 100 from lower diagonal elements below starting at some unknown vertex. Assume
of A that a recursive call to visit a vertex is made only
(d) None of these after first checking that the vertex has not been
visited earlier. Then the maximum possible recur-
Solution: A[i][j] and A[j][i] exchanging their
sion depth (including the initial call) is .
values. Perform the operation for atleast three
iterations.
Ans. (a)
11. The minimum number of arithmetic operations
required to evaluate the polynomial P(X) = X5 +
4X3 + 6X + 5 for a given value of X, using only
one temporary variable is .
Solution: Solution: In depth first search, the nodes are tra-

3 2
P(X) = X (X + 4) + 6X + 4 versed in depth. Maximum possible depth is 19.
Ans. (19)
X2 = 1
X3 = 1 14. You have an array of n elements. Suppose you
implement quick sort by always choosing the cen-
X3 × (X2 + 4) = One multiplication + One addition
tral element of the array as the pivot. Then the
= 2 operations tightest upper bound for the worst case perfor-
6X = 1 multiplication mance is
5 + 2 additions = 7 operations (a) O(n2) (b) O(n log n)
Ans. (7) (c) Θ(n log n) (d) O(n2)
12. Consider the following rooted tree with the vertex Solution: Worst case is if the list is already sorted.
labelled P as the root So, worst case complexity is O(n2).
P Ans. (a)
15. The length of the shortest string NOT in the lan-

guage (over S = {a,b}) of the following regular
expression is . a*b* (ba)* a*
Q R Solution: String of length 3 `bab’ is not present.
Ans. (3)
16. Let ∑ be a finite non-empty alphabet and let 2∑*
be the power set of ∑ *. Which one of the following
T U V is TRUE?
(a) Both 2∑* and ∑* are countable
S
(b) 2∑* is countable ∑* is uncountable

(c) 2∑* is uncountable and ∑* is countable
W (d) Both 2∑* and ∑* are uncountable

Solution: ∑* is countable infinite. The power set Solution:

of countable infinite sets is uncountable. LRU with 3 frames
Ans. (c) 4→1
17. One of the purposes of using intermediate code in 7→2
compilers is to 6→7
(a) make parsing and semantic analysis simpler. Page faults = 6
(b) improve error recovery and error reporting Ans. (6)
(c) increase the chances of reusing the machine-
independent code optimizer in other compliers. 21. What is the optimized version of the relation alge-
(d) improve the register allocation. bra expression (pA1(pA2(sF1(r)))), where A1, A2
are sets of attributes in with A1 ⊂ A2 and F1, F2
Solution: Increase the chances of reusing the are Boolean expressions based on the attributes
machine-independent code optimizer in other in r?
(a) pA1(s(F1 ∧ F2) (r)) (b) pA1(s(F1 ∨ F2) (r))
compliers.
(c) pA2(s(F1 ∧ F2) (r)) (d) pA2(s(F1 ∨ F2) (r))
Ans. (c)
18. Which of the following statements are CORRECT?

Solution: p select column and s select row. As we
(1) Static allocation of all data areas by a compiler can see from relational algebra expression, p select
makes it impossible to implement recursion. the final A1 attribute set. Row is selected which
(2) Automatic garbage collection is essential to fulfills both the conditions of F1 and F2.
implement recursion. So, equivalent expression that can be written is:
(3) Dynamic allocation of activation records is pA1(s(F1 ∧ F2(r)))
essential to implement recursion. Ans. (a)
(4) Both heap and stack are essential to implement
recursion. 22. A prime attribute of a relation scheme R is an
attribute that appears
(a) 1 and 2 only (b) 2 and 3 only
(c) 3 and 4 only (d) 1 and 3 only (a) in all candidate keys of R.
(b) in some candidate key of R.
Solution: For recursion, dynamic memory alloca- (c) in a foreign keys of R.
tion is required. (d) only in the primary key of R.
Ans. (d)
Solution: An attribute that appears in candidate
19. In the context of modular software design, which key is known as prime attribute.
one of the following combinations is desirable? Ans. (b)
(a) High cohesion and high coupling
23. In the following pairs of OSI protocol layer/sub-
(b) High cohesion and low coupling
layer and its functionality, the INCORRECT pair is
(c) Low cohesion and high coupling
(d) Low cohesion and low coupling (a) Network layer and Routing
(b) Data Link Layer and Bit synchronization
Solution: Cohesion measures the internal depen- (c) Transport layer and End-to-end process
dency within module whereas coupling presents communication
inter module dependence. So, there should be high (d) Medium Access Control sub-layer and Channel
cohesion and low coupling. sharing
Ans. (b)
Solution: The task of data link layer is to create
20. A system uses 3 page frames for storing process frames. Bit synchronization is handled by physical
pages in main memory. It uses the Least Recently layer.
Used (LRU) page replacement policy. Assume that Ans. (b)
all the page frames are initially empty. What is the
total number of page faults that will occur while 24. A bit-stuffing based framing protocol uses an 8-bit
processing the page reference string given below? delimiter pattern of 01111110. If the output bit-
string after stuffing is 01111100101, then the input
4, 7, 6, 1, 7, 6, 1, 2, 7, 2 bit-string is

SET 3 583
(a) 0111110100 (b) 0111110101 Third Entry:

(c) 0111111101 (d) 0111111111 IP: 131. 00010111.151.76
Solution: 01111110 is the delimiter pattern given. Mask: 131. 00010011. 0.0 [16 n/w bits]
In output string 011111100101, 0 is stuffed. Input n/w address: 131.19.0.0 [Not Matched]
string is 01111110101.
Ans. (b) Fourth Entry:
25. Host A (on TCP/IP v4 network A) sends an IP IP: 131. 00010111.151.76
datagram D to host B (also on TCP/IP V4 net- Mask: 131. 00010110. 0.0 [15 n/w bits]
work B). Assume that no error occurred during the n/w address: 131.20.0.0 [Matched]
transmission of D. When D reaches B, which of the
following IP header field(s) may be different from according to longest mask, packet will be forwarded
that of the original datagram D? to interface 1.
Ans. (1)
(i) TTL (ii) Checksum (iii) Fragment Offset
27. Every host in an IPv4 network has a 1-second
(a) (i) only (b) (i) and (ii) only resolution real-time clock with battery backup.
(c) (ii) and (iii) only (d) (i), (ii) and (iii) Each host needs to generate up to 1000 unique
Solution: TTL, checksum and fragment offset identifiers per second. Assume that each host has a
changes at every node. globally unique IPv4 address. Design a 50-bit glob-
Ans. (d) ally unique ID for this purpose. After what period
(in seconds) will the identifiers generated by a host
wrap around?
Q. No. 26 — 55 Carry Two Marks Each
Solution:
50 bit unique id = 32 bit ip + 18 bit extra
26. An IP router implementing Classless Inter-domain With 18 bits = 218 ids can be generated.
routing (CIDR) receives a packet with address
131.23.151.76. The router’s routing table has the 1000 ids take = 1 sec
following entries: 218 id will take = 218/1000 = 256 sec.
Ans. (256)
Prefix Output Interface
Identifier 28. An IP router with a Maximum Transmission Unit
(MTU) of 1500 bytes has received an IP packet
131.16.00/12 3 of size 4404 bytes with an IP header of length
131.28.0.0/14 5 20 bytes. The values of the relevant fields in the
header of the third IP fragment generated by the
131.19.0.0/16 2
router for this packet are
131.22.0.0/15 1
(a) MF bit: 0, Datagram Length: 1444; Offset: 370
(b) MF bit: 1, Datagram Length: 1424; Offset: 185
The identifier of the output interface on which this
. (c) MF bit: 1, Datagram Length: 1500; Offset: 370
packet will be forwarded is
(d) MF bit: 0, Datagram Length: 1424; Offset: 2960
Solution: Solution:
Calculating the network address: First fragment: 1480 + 20 MF = 1
First entry:
Second fragment: 1480 + 20 MF = 1
IP: 131. 00010111.151.76
Third fragment: 1444 + 20 MF = 0
Mask: 131. 00010000. 0.0 [12 n/w bits] Ans. (a)
n/w address: 131.16.0.0 [Matched]
29. Consider the transactions T1, T2, and T3 and the
Second Entry: schedules S1 and S2 given below.
IP: 131. 00010111.151.76 T1 : r1 (X) ; r1 (Z) ; w1 (X) ; w1 (Z)
Mask: 131. 00011100. 0.0 [14 n/w bits] T2 : r2 (X) ; r2 (Z) ; w2 (Z)
n/w address: 131.20.0.0 [Not Matched] T3 : r3 (X) ; r3 (X) ; w3 (Y)

S1: r1(X); r3(Y); r3(X); r2(Y); r2(Z); w3(Y); Solution:

w2(Z); r1(Z); w1(X); w1(Z) Part A — Part B
S2: r1(X); r3(Y); r2(Y); r3(X); r1(Z); r2(Z); w3(Y); Part B selects those employee ids whose age is less
w1(X); w2(Z); w1(Z) than his dependents. When Part B is subtracted
Which one of the following statements about the from all the employee ids, result will be set of
schedules is TRUE? those employee eids who are greater than all their
dependents.
(a) Only S1 is conflict-serializable. Ans. (d)
(b) Only S2 is conflict-serializable.
31. A system contains three programs and each requires
(c) Both S1 and S2 are conflict-serializable.
three tape units for its operation. The minimum
(d) Neither S1 nor S2 is conflict-serializable.
number of tape units which the system must have
Solution: Precedence graph according to Schedule 1: such that deadlocks never arise is .
Solution:
T1 T2
Exp: Maximum Allocate Need Available
P1 −3 2 1 1
P2 −3 2 1
T3
P3 −3 2 1
No cycle so S1 is conflict serializable. With the above given data, after allocating 2 units
Precedence graph according to Schedule 2: of tape to each process, with 1 available unit any
of the 3 process can be satisfied in such a way, that
No dead lock will be there.
T1 T2 So, answer is 7 tape units.
There are three processes, each require 3 resources.
If we have 7 resources, we can allocate 2 to each
process and 1 would be available. Any process can
T3 start executing.
Ans. (7)
32. An operating system uses shortest remaining time
Cycle exists, so S2 is not conflict serialzable.
first scheduling algorithm for pre-emptive sched-
Ans. (a)
uling of processes. Consider the following set of
30. Consider the relational schema given below, where processes with their arrival times and CPU burst
eId of the relation dependent is a foreign key refer- times (in milliseconds):
ring to empId of the relation employee. Assume
that every employee has at least one associated Process Arrival Time Burst Time
dependent in the dependent relation. P1 0 12
employee (empId, empName, empAge) P2 2 4
dependent (depId, eId, depName, depAge) P3 3 6
Consider the following relational algebra query P4 8 5
Π empId (employee) − Π empId (employee (empId = eID) ∧The average

(empAge ≤ depA
waiting
ge) time (in milliseconds) of the
dependent)
processes is .
− Π empId (employee (empId = eID) ∧ (empAge ≤ depAge) dependent)
Solution:
The above query evaluates to the set of empIds of According to SRJF
employees whose age is greater than that of P1 P2 P3 P4 P1
(a) some dependent. 0 2 6 12 17 27
(b) all dependents.
(c) some of his/her dependents. Waiting time:
(d) all of his/her dependents. P1: 17 − 2 = 15

SET 3 585
P2: 0 Solution: There is no algorithm to check ambigu-

P3: 6 − 3 = 3 ity of CFG.
Ans. (a)
P4: 12 − 8 = 4
Average waiting time = 15 + 3 + 4 = 22/4 = 5.5 ms 36. Consider the following languages over the alphabet
= {0,1,c}
Ans. (5.5)
L1 = {0n1n | n ≥ 0}
33. Consider a paging hardware with a TLB. Assume
that the entire page table and all the pages are in L2 = {wcwr | w ∈ {0, 1}*} { }
the physical memory. It takes 10 milliseconds to L3 = {wwr | w ∈ {0, 1}*}
search the TLB and 80 milliseconds to access the
physical memory. If the TLB hit ratio is 0.6, the Here wr is the reverse of the string w. Which of
effective memory access time (in milliseconds) is these languages are deterministic context-free
. languages?
Solution: Effective memory access time = Cache (a) None of the languages
hit × (TLB access + Main Memory) + Cache miss (b) Only 1 L
× (TLB access + 2 × Main Memory) (c) Only 1 L and 2 L
(d) All the three languages
= 0.6 × (10 + 80) + 0.4 × (10 + 160)
=122 ms Solution: L1 and L2 are accepted by DPDA, L3 is
Ans. (122) accepted by NDPDA.
Ans. (c)
34. Consider the basic block given below.
37. Suppose you want to move from 0 to 100 on the
a=b+c
number line. In each step, you either move right by
c=a+d a unit distance or you take a shortcut. A shortcut is
d=b+c simply a pre-specified pair of integers i, j with i < j.
e=d−b Given a shortcut i, j if you are at position i on the
a=e+b number line, you may directly move to j. Suppose
The minimum number of nodes and edges present T(k) denotes the smallest number of steps needed
in the DAG representation of the above basic block to move from k to 100. Suppose further that there
respectively are is atmost 1 shortcut involving any number, and
in particular from 9 there is a shortcut to 15. Let
(a) 6 and 6 (b) 8 and 10 y and z be such that T(9) = 1 + min(T(y), T(z)).
(c) 9 and 12 (d) 4 and 4 Then the value of the product yz is .
Solution: Solution:
a=b+c T(9) = 1 + min(T(x), T(y))
c=a+d T(x) and T(y) can be two values: either move right
d = 2b + c + d or short cut
e=c
Moving right means: 10 and shortcut from 9 is 15
a=d
So product is 15 × 10 = 150.
The maximum number of nodes and edges are
Ans. (150)
(6, 6).
Ans. (a) 38. Consider the decision problem 2CNFSAT defined
as follows:
35. Which one of the following problems is undecidable?
{f f is a satisfiable propositional formula in CNF
(a) Deciding if a given context-free grammar is with at most two literal per clause}
For example, φ = (x1 ∨ x2 ) ∧ (x1 ∨ x3 ) ∧ (x2 ∨ x4 )
ambiguous.
(b) Deciding if a given string is generated by a
is a Boolean formula and it is in 2CNFSAT. The
given context-free grammar.
decision problem 2CNFSAT is
(c) Deciding if the language generated by a given
context-free grammar is empty. (a) NP-Complete.
(d) Deciding if the language generated by a given (b) solvable in polynomial time by reduction to
context-free grammar is finite. directed graph reachability.

Struct treeNode
{
Treeptr leftMostchild, rightSibiling;
};
Int Dosomething (treeptr tree)
{
586 SOLVED GATE (CS) 2014 int value =0;
if (tree ! = NULL) {
(c) solvable in constant time since any input If (tree -> leftMostchild = = NULL)
instance is satisfiable. else
(d) NP-hard, but not NP-complete. value = Dosomething
(tree->leftMostchild);
Solution: The decision problem 2CNFSAT is value = value + Dosometing
Solvable in polynomial time by reduction to (tree->rightsibiling);
directed graph reachability. }
Ans. (b) return (value);
}
39. Suppose we have a balanced binary search tree T
holding n numbers. We are given two numbers L and When the pointer to the root of a tree is passed as
H and wish to sum up all the numbers in T that lie the argument to DoSomething, the value returned
between L and H. Suppose there are m such num- by the function corresponds to the
bers in T. If the tightest upper bound on the time
to compute the sum is O(na logb n + mc logd n), (a) number of internal nodes in the tree.
the value of a + 10b + 100c + 1000d is . (b) height of the tree.
(c) number of nodes without a right sibling in the
Solution: tree.
Time complexity is O (log n + m) (d) number of leaf nodes in the tree.
According to this, a = 0, b = 1, c = 1 and d = 0
Solution: Pseudocode recursively finds the leaf
Put these in given equation: a + 10b + 100c + nodes and counts them.
1000d = 0 + 10 + 100 + 0 = 110 Ans. (d)
Ans. (110)
42. Consider the C function given below. Assume that
40. Consider a hash table with 100 slots. Collisions are the array listA contains n (> 0) elements, sored in
resolved using chaining. Assuming simple uniform ascending order.
hashing, what is the probability that the first
3 slots are unfilled after the first 3 insertions? int ProcessArray (int * listA,
int x, int n)
(a) (97 × 97 × 97)/100 3
{
(b) (99 × 98 × 97)/1003 int 1, j, k;
(c) (97 × 96 × 95)/1003 i = 0;
(d) (97 × 96 × 95)/(3! × 1003) j = n – 1;
Solution: do {
k = (i + j) /2;
97 × 97 × 97
Using same index: if (x < = listA [k])
100 × 100 × 100 j = k – 1;
Ans. (a) If (listA [k] < = x)
41. Consider the pseudocode given below. The function i = k+1;
Dosomething () takes as argument a pointer to the }while (1 < = j);
root of an arbitrary tree represented by the left- If (listA [k] = = x)
MostChild-rightSibling representation. Each node return (k) ;
of the tree is of type treeNode. else
return -1;
typedef struct treeNode* treeptr; }
Struct treeNode
Which one of the following statements about the
{
function ProcessArray is CORRECT?
Treeptr leftMostchild, rightSibiling;
}; (a) It will run into an infinite loop when x is not
Int Dosomething (treeptr tree) in listA.
{ (b) It is an implementation of binary search.
int value =0; (c) It will always find the maximum element in listA.
if (tree ! = NULL) { (d) It will return — 1 even when x is present in
If (tree -> leftMostchild = = NULL) listA.
else
value = Dosomething
(tree->leftMostchild);
value = value + Dosometing
(tree->rightsibiling);
}
return (value);
}
SET 3 587
Solution: Checking the logic of code clearly show Stall cycle = 5

that binary search has been implemented. Stall frequency = 20%
Ans. (b) Maximum clock period = 1
43. An instruction pipeline has five stages, namely, Average access time (Q) = (0.8 + 0.12) × 1 = 2 ns
instruction fetch (IF), instruction decode and P/Q = 3.08/2 = 1.54
register fetch (ID/RF), instruction execution Ans. (1.54)
(EX), memory access (MEM), and register write
back (WB) with stage latencies 1 ns, 2.2 ns, 44. The memory access time is 1 nanosecond for a
2 ns, 1 ns, and 0.75 ns, respectively (ns stands read operation with a hit in cache, 5 nanoseconds
for nanoseconds). To gain in terms of frequency, for a read operation with a miss in cache, 2 nano-
the designers have decided to split the ID/RF seconds for a write operation with a hit in cache
stage into three stages (ID, RF1, RF2) each of and 10 nanoseconds for a write operation with a
latency 2.2/3 ns. Also, the EX stage is split into miss in cache. Execution of a sequence of instruc-
two stages (EX1, EX2) each of latency 1 ns. The tions involves 100 instruction fetch operations,
new design has a total of eight pipeline stages. 60 memory operand read operations and 40 memory
A program has 20% branch instructions which operand write operations. The cache hit-ratio is
execute in the EX stage and produce the next 0.9. The average memory access time (in nano-
instruction pointer at the end of the EX stage in seconds) in executing the sequence of instructions
the old design and at the end of the EX2 stage in is .
the new design. The IF stage stalls after fetching
a branch instruction until the next instruction Solution:
pointer is computed. All instructions other than 100 fetch operations: 90 × 1 ns + 10 × 5 = 140 ns
the branch instruction have an average CPI of Memory read operations (60): 0.9 × 60 × 1 + 0.1
one in both the designs. The execution times of × 60 × 5 = 84 ns
this program on the old and the new design are
Memory write operations (40): 0.9 × 40 × 1 + 0.1
P and Q nanoseconds, respectively. The value of
. × 40 × 5 = 112 ns
P/Q is
Average memory access time = 336/200 = 1.68
Solution: Ans. (1.68)
Old Design:
45. The above synchronous sequential circuit built
Stages = 5 using JK flip-flops is initialized with Q2Q1Q0 =
Stall cycle = 2 000. The state sequence for this circuit for the next
Stall frequency = 20% 3 clock cycles is
Maximum clock period = 2.2 (a) 001, 010, 011
Average access time (P) = (0.8 + 0.6) × 2.2 = 3.08 ns (b) 111, 110, 101
(c) 100, 110, 111
New Design: (d) 100, 011, 001
Stages = 8
J Q2 J Q1 J Q0
C C C
K Q2 K Q1 K Q0

Solution:
Present State Next State
Q2 Q1 Q0 Q2(J = Q1′, K = Q0) Q1(J = Q2, K = Q2′) Q0(J = Q1, K = Q0′)
0 0 0 1 0 0
1 0 0 1 1 0
1 1 0 1 1 1
Ans. (c)
46. With respect to the numerical evaluation of the 47. The value of the integral given below is
p
definite integral, K = ∫
b 2
∫x
x dx , where a and b are 2
a cos x dx
given, which of the following statements is/are 0
TRUE? (a) −2p (b) p (c) −p (d) 2p
(I) The value of K obtained using the trapezoidal Solution:
rule is always greater than or equal to the p
òx cos x dx = [x2 sin x − 2x(− cos x)

2
exact value of the definite integral.
(II) The value of K obtained using the Simpson’s 0 + 2(− sin x)]0 to p
rule is always equal to the exact value of the = −2p
definite integral.
= [x2 sin x − 2x(− cos x)
(a) I only (b) II only
+ 2(− sin x)]0 to p
(c) Both I and II (d) Neither I nor II
= −2p
Solution: Ans. (a)
1
x3
∫x dx =
2 1 48. Let S be a sample space and two mutually exclu-
3
= = 0. 33333
3 sive events A and B be such that A ∪ B = S. If P(.)
0
denotes the probability of the event, the maximum
As N = 4, we have value of P(A)P(B) is .
Solution:
x 0 0.25 0.5 0.75 1 Maximum value of P(A) P(B) = ?
y = x2 0 0.0625 0.25 0.5625 1 P(A)P(B) = P(A)[1 − P(A)]
y0 y1 y2 y3 y4 Let P(A) = x
f (x) = x − x2
Trapezoidal rule: f ’(x) = 1 − 2x = 0
1 So, x = 1/2 and f ”(x) = −2 and x is maximum
∫x dx = [(0 + 1) + 2(0.0625 + 0.25 + 0.5625)] = 0.34375 at 1/2
2 0.25
0
2
P(A) P(B) = 0.25
) + 2(0.0625 + 0.25 + 0.5625)] = 0.34375 49. Consider the set of all functions f :{0, 1, …, 2014}
→ {0, 1 ,…, 2014} such that f (f (i)) = i, for 0 ≤ i
Simpson’s 1/3 rule: ≤ 2014. Consider the following statements.
1 P. For each such function it must be the case that
∫x dx = [(0 + 1) + 2 (0.25) + 4(0.0625 + 0.5625)] = 0.3333 for every i, f(i) = i,
2 0.25
0
2
Q. For each such function it must be the case that
(0.25) + 4(0.0625 + 0.5625)] = 0.3333

for some i, f(i) = i,
R. Each such function must be onto.
Ans. (c) Which one of the following is CORRECT?

SET 3 589
(a) P, Q and R are true We know that e = n + r − 2

(b) Only Q and R are true n + r − 2 ≥ 3n/2
(c) Only P and Q are true
(d) Only R is true or r ≥ n/2 + 2
Ans. (a)
Solution: It is not necessary that for every f(i) = i,
53. The CORECT formula for the sentence, “not all
so statement P is incorrect.
rainy days are cold” is
Ans. (b)
(a) ∀d (Rainy(d) ∧ ~Cold(d))
50. There are two elements x,y in a group (G,*) such (b) ∀d (~Rainy(d) → Cold(d))
that every element in the group can be written as (c) ∃d (~Rainy(d) → Cold(d))
a product of some number of x’s and y’s in some (d) ∃d (Rainy(d) ∧ ~Cold(d))
order. It is known that
x * x = y * y = x * y *x * y = y* x * y *x = e Solution:
where e is the identity element. The maximum “Not all rainy days are cold”
∼ ∀d ( rainy ( d ) → cold ( d )) →

number of elements in such a group is .
Solution: Elements of this group are: {x, y, e,

∃d ( rainy ( d ) ∼ cold ( d ))
x * y}
Ans. (4) Ans. (d)
51. If G is a forest with n vertices and k connected 54. Consider the following relational schema:
components, how many edges does G have?
Employee (empId, empName, empDept)
(a) [n/k] (b) [n/k]
(c) n − k (d) n − k +1 Customer (custId,custName, salesRepId,
rating)
Solution: In forest each component is a tree.
SalesRepId is a foreign key referring to empId of
There are G1, G2, … G k components. Each has
the employee relation. Assume that each employee
n − 1 edges.
makes a sale to at least one customer. What does
k
∑n − i = n − k edges
the following query return?
i =1 SELECT empName
Ans. (c) FROM employee E
52. Let d denote the minimum degree of a vertex in a WHERE NOT EXISTS (SELECT custId
graph. For all planar graphs on n vertices with d ≥ FROM customer C
3, which one of the following is TRUE?
WHERE C. salesRepId = E. empId
(a) In any planar embedding, the number of faces AND C. rating < > `GOOD’)
n
is at least + 2 (a) N ames of all the employees with at least one of
2
their customers having a `GOOD’ rating.
(b) In any planar embedding, the number of faces
n (b) Names of all the employees with at most one of
is less than + 2 their customers having a `GOOD’ rating.
2 (c) Names of all the employees with none of their
(c) There is a planar embedding in which the customers having a `GOOD’ rating.
n (d) Names of all the employees with all their cus-
number of faces is less than + 2
2 tomers having a `GOOD’ rating.
(d) There is a planar embedding in which the
n Solution: Inner query will result in the customer
number of faces is at most
d +1 ids of those whose rating is not good. Outer query
results in those employee names whose customer
Solution:
ids are not in the inner query. So, the result will
Degree ≥ 3 is of n vertices. display employees with good rating.
Means 3n ≤ 2e → e ≥ 3n/2 Ans. (d)

55. Let ⊕ denote the Exclusive OR (XOR) opera- Solution:

tion. Let `1’ and `0’ denote the binary constants. (P ′ ⊕ ( P ⊕ Q)) ⊕((P ⊕ Q) ⊕ Q)
Consider the following Boolean expression for F
over two variables P and Q. (P ′ ⊕ ( PQ′ + P ′Q)) ⊕ (( PQ′ + P ′Q) ⊕ Q)
F(P,Q) = ((1 ⊕ P) ⊕ (P ⊕ Q)) ⊕ ((P ⊕ Q) (P ′( PQ′ + P ′Q) + P (PQ′ + P ′Q)′ ) ⊕ (Q(PQ′ + P ′Q)′ + Q′(PQ′ + P ′Q
⊕ (Q ⊕ 0)) (P ′( PQ′ + P ′Q) + P (PQ′ + P ′Q)′ ) ⊕ (Q(PQ′ + P ′Q)′ + Q′(PQ′ + P ′Q))
Q′ ⊕ P
The equivalent expression for F is
(P ⊕ Q)′
(a) P + Q (b) P + Q
(c) P ⊕ Q (d) P ⊕ Q
Ans. (d)

SET 1
Q. NO. 1 – 25 Carry One Mark Each
x g(h(x)) 3. Match the following:

1. If g(x) =1 — x and h(x) = , then is
x −1 h(g(x))
P: Prim’s algorithm for (i) Backtracking
h(x) −1
(a) (b) minimum spanning tree
g(x) x Q: Floyd—Warshal algo- (ii) Greedy method
g(x) x rithm for all pairs shortest
(c) (d) paths
h(x) (1 − x)2
R: Merge sort (iii) Dynamic programming
x S: Hamiltonian circuit (iv) Divide and conquer
Solution: We have g(x) = 1 − x, h(x) = ,
x −1
(a) P-iii, Q-ii, R-iv, S-i
 x  x 1
g(h(x)) = g  = 1−
 x − 1
= (b) P-i, Q-ii, R-iv, S-iii
x −1 1− x (c) P-ii, Q-iii, R-iv, S-i
x −1 (d) P-ii, Q-i, R-iii, S-iv
h(g(x)) = h(1 − x) =
x Solution: Merge sort firstly divides the list and
g(h(x)) x h(x) then merges it to sort. Hence, it uses divide and
= =
h(g(x)) (x − 1)(1 − x) g(x)
conquer strategy.
All pair shortest path algorithm uses the concept
Hence, (a) is the correct answer. of overlapping while computing the shortest path.
Ans. (a) Hence, it uses dynamic programming.
2. lim x→∞ x1/x is Hamiltonian circuit means covering all the vertices
and while doing so backtracking is required.
(a) ∞ (b) 0 (c) 1 (d) Not defined
Prim’s algorithm for minimum spanning tree is
1/x greedy method.
Solution: In lim x →∞x let us assume y =
lim x→∞ x1/x Ans. (c)
Taking logarithm on both sides, 4. Which one of the following is the recurrence equa-
1/x tion for the worst case time complexity of the
log y = log (lim x→∞ x ) = lim x→∞ (log x)/x
Quicksort algorithm for sorting n (≥ 2) numbers?
RHS is ∞/∞ form, therefore L-Hopital rule can In the recurrence equations given in the options
be applied below, c is a constant.
Thus, log y = lim x→∞ 1/x = 0. Hence, y = 1. (a) T(n) = 2T(n/2) + cn
Ans. (c) (b) T(n) = T(n−1) + T(1) + cn
GATE 2015 Set 1 solutions.indd 591 4/9/2015 10:21:19 AM

(c) T(n) = 2T(n−1) + cn 7. Which of the following is/are correct inorder tra-
(d) T(n) = T(n/2) + cn versal sequence(s) of binary search tree(s)?
Solution: Quicksort’s worst case is when the list I. 3, 5, 7, 8, 15, 19, 25
is already sorted or when it is in reverse order. In II. 5, 8, 9, 12, 10, 15, 25
both cases, the pivot element is already in sorted III. 2, 7, 10, 8, 14, 16, 20
order and list will always be divided into two lists IV. 4, 6, 7, 9, 18, 20, 25
containing 1 and n−1 elements.
(a) I and IV only (b) II and III only
Thus, the recurrence relation formed is as follows: (c) II and IV only (d) II only
T(n) = T(n−1) + T(1) + cn Solution: Inorder traversal of tree is the sorted
Ans. (b) order of nodes of the tree. Check the given
5. The height of a tree is the length of the longest sequences; the one which is in sorted order is the
root-to-leaf path in it. The maximum and minimum correct inorder traversal.
number of nodes in a binary tree of height 5 are Ans. (a)
8. Which one of the following is TRUE at any valid
(a) 63 and 6, respectively
state in shift — reduce parsing?
(b) 64 and 5, respectively
(c) 32 and 6, respectively (a) Viable prefixes appear only at the bottom of
(d) 31 and 5, respectively the stack and not inside
(b) Viable prefixes appear only at the top of the
Solution: If height of the binary tree is h, then stack and not inside
Minimum number of nodes in a tree = h + 1 (c) The stack contains only a set of viable prefixes
Maximum number of nodes = 2h+1 − 1 (d) The stack never contains viable prefixes
Here, h = 5 Solution: The set of prefixes of right sentential

forms that can appear on the stack of a shift-reduce
Therefore, minimum nodes = 5 + 1 = 6
parser are called viable prefixes. They come at the
maximum nodes = 26 − 1 = 63 top of stack.
Hence, answer is 63 and 6, respectively. Ans. (b)
Ans. (a) 9. Which one of the following is NOT equivalent to
6. Match the following: p↔q?
(a) (p∨q) ∧(p∨q) (b) (p∨q) ∧(q→p)
(c) (p∧q) ∨(p∧q) (d) (p∧q) ∨(p∧q)
P: Condition coverage (I) Black-box testing
Q: Equivalence class
(II) System testing
partitioning Solution: Drawing the truth table of all the
R: Volume testing (III) White-box testing options, we can find the equivalency of expressions.
S: Alpha testing (IV) Performance testing Ans. (c)
(a) P-II, Q-III, R-I, S-IV 10. For a set A, the power set of A is denoted by 2A. If
(b) P-III, Q-IV, R-II, S-I A = {5, {6}, {7}}, which of the following options
(c) P-III, Q-I, R-IV, S-II are TRUE?
(d) P-III, Q-I, R-II, S-IV I. Φ ∈ 2A II. Φ ⊆ 2A
Solution: Condition coverage means covering all III. {5, {6}} ∈ 2A IV. {5{6}} ⊆ 2A
the possibilities of both true and false conditions.
White box testing covers the internal structure of (a) I and III only (b) II and III only
software. (c) I, II and III only (d) I, II and IV only
Equivalence class testing is dividing the input of Solution: A: The given set and 2A: Power set of
software into small partitions of equivalent data, A, that is, set of subsets of A.
from which various test cases are derived. This is
black box testing. Volume or stress testing is a type Φ is the subset of every set.
of system testing. Therefore, statements I and II are true.
Alpha testing is done by the user/customer for Since 5 does not belong to 2A. Hence, statement IV
checking its performance. is not true.
Ans. (d) Ans. (c)

SET 1 593
11. Consider a 4-bit Johnson counter with an ini-

Clock Flip- Flip- Flip- Flip-
tial value of 0000. The counting sequence of this
Pulse Flop A Flop B Flop C Flop D
counter is
Number
(a) 0, 1, 3, 7, 15, 14, 12, 8, 0 Initial 0 0 0 0
(b) 0, 1, 3, 5, 7, 9, 11, 13, 15, 0
1 1 0 0 0
(c) 0, 2, 4, 6, 8, 10, 12, 14, 0
(d) 0, 8, 12, 14, 15, 7, 3, 1, 0 2 1 1 0 0
3 1 1 1 0
Solution: This inversion of Q is feedback to input
4 1 1 1 1
D of flip-flop A and this causes the counter to
“count” in a unique way. Counting Pattern is as 5 0 1 1 1
follows: 6 0 0 1 1
7 0 0 0 1
4-bit Parallel output

QA QB QC QD
D Q D Q D Q D Q
Flip-Flop Flip-Flop Flip-Flop Flip-Flop

A B C D
Q
CLK
CLR
Ans. (d)
12. For computers based on three-address instruction II. If the estimated round trip time at any given
formats, each address field can be used to specify point of time is t sec the value of the retrans-
which of the following. mission timeout is always set to greater than or
equal to t sec.
S1: A memory operand
III. The size of the advertised window never changes
S2: A processor register
during the course of the TCP connection.
S3: An implied accumulator register
IV. The number of unacknowledged bytes at the
(a) Either S1 or S2 (b) Either S2 or S3 sender is always less than or equal to the
(c) Only S2 and S3 (d) All of S1, S2 and S3 advertised window.
Solution: Computer with three-address instruc-
tion format can use each address field to specify (a) III only (b) I and III only
either processor register or memory operand. (c) I and IV only (d) II and IV only
Ans. (a)
Solution: It is not necessary to have the subse-
13. Suppose two hosts use a TCP connection to trans-
quent sequence number as m + 1.
fer a large file. Which of the following statements
is/are FALSE with respect to the TCP connection? Size of window is not constant during TCP connec-
tion. It changes once the threshold value of conges-
I. If the sequence number of a segment is m,
tion window reaches. So, statement III is false.
then the sequence number of the subsequent
segment is always m + 1. Ans. (b)

14. Suppose that everyone in a group of N people wants not recursive, which of the following is/are neces-
to communicate secretly with the N−1 others using sarily true?
symmetric key cryptographic system. The commu- I. L1 (complement of L1) is recursive
nication between any two persons should not be
II. L 2 (complement of L2) is recursive
decodable by the others in the group. The number
of keys required in the system as a whole to satisfy III. L1 is context-free
the confidentiality requirement is IV. L1 ∪ L2 is recursively enumerable
(a) 2N (b) N(N−1) (a) I only (b) III only
(c) N(N−1)/2 (d) (N−1)2 (c) III and IV only (d) I and IV only
Solution: In symmetric key cryptography, for n Solution: Context-free language is also recur-
users the number of possible keys is n(n−1)/2. sive and recursive is closed under complement.
Therefore, statement I is true.
Ans. (c)
Complement of recursive enumerable is not recur-
15. Which of the following statements is/are FALSE? sive. Hence, statement II is false.
I. XML overcomes the limitations in HTML to Context-free language is not closed under comple-
support a structured way of organizing content. ment. Hence, statement III is false.
II. XML specification is not case sensitive while ∼L1 is recursive and hence also recursive enumer-
HTML specification is case sensitive. able. L2 is recursive enumerable. RE languages are
III. XML supports user defined tags while HTML closed under union. Hence, statement IV is true.
uses pre-defined tags. Ans. (d)
IV. XML tags need not be closed while HTML tags
19. Consider a system with byte-addressable memory,
must be closed
32-bit logical addresses, 4 kilobyte page size and
(a) II only (b) I only page table entries of 4 bytes each. The size of the
(c) II and IV only (d) III and IV only page table in the system in megabytes is .
Solution: HTML is case insensitive, while XML is Solution: Logical address space = 32 bits
case sensitive. Therefore, statement II is incorrect. Page size = 4 KB = 212
XML tags must be closed. Hence, statement IV is Each page table entry size = 4 B
also false.
Number of pages = 232/212 = 220
Ans. (c)
Page table size = (Number of pages) × (Page table
16. Which one of the following fields of an IP header is entry size) = 4 B × 220 = 4 MB
NOT modified by a typical IP router? Ans. (4)
(a) Checksum (b) Source address 20. The following two functions P1 and P2 that share
(c) Time to Live (TTL) (d) Length a variable B with an initial value of 2 execute
concurrently.
Solution: Source address is the only thing that
remains fixed and cannot be varied. P1() P2()
Ans. (b) { {
C = B – 1; D = 2 * B;
17. In one of the pairs of protocols given below, both the B = 2 * C; B = D – l;
protocols can use multiple TCP connections between } }
the same client and the server. Which one is that?
The number of distinct values that B can possible
(a) HTTP, FTP (b) HTTP, TELNET take after the execution is .
(c) FTP, SMTP (d) HTTP, SMTP
Solution: Initial value of B = 2
Solution: HTTP can use multiple TCP connec- There are four possibilities in which two processes
tions for each resource. can be executed:
FTP uses Telnet protocol for control information
P1 is executed first and then P2:
on a TCP connection and another TCP connection
for data exchange. After this execution B = 3
Ans. (a) P2 is executed first and then P1
18. For any two languages L1 and L2 such that L1 is After this execution B = 4
context-free and L2 is recursively enumerable but P1 is preempted by P2

SET 1 595
After this execution B = 2 int main()

{
P2 is preempted by P1
int a=4, b=5, c=6;
After this execution B = 3 f1(a, b);
Therefore, three possible values of B = 2, 3, 4 f2(&b, &c);
Ans. (3) printf(“%d”, c–a–b);
}
21. SELECT operation in SQL is equivalent to
(a) the selection operation in relational algebra Solution: f1 (a, b): This is call by value. Values
(b) the selection operation in relational algebra, will not change.
except that SELECT in SQL retains duplicates f2 (&b, &c): This is call by reference. This swaps
(c) the projection operation in relational algebra values of b and c.
(d) the projection operation in relational algebra,
except that SELECT in SQL retains duplicates Now a = 4, b = 6, c = 5
Solution: In SQL, the SELECT operation is used printf will print c − a − b = 5 − 6 − 4 = −5

for selecting the columns. In a similar way, rela- Ans. (−5)
tional algebra uses project to select the columns.
But SQL retains duplicates as well which is not so 25. What are the worst-case complexities of insertion
in relational algebra. and deletion of a key in a binary search tree?
Ans. (d)
(a) Θ(log n) for both insertion and deletion
22. A file is organized so that the ordering of data
(b) Θ(n) for both insertion and deletion
records is the same as or close to the ordering of
(c) Θ(n) for insertion and Θ(log n) for deletion
data entries in some index. Then that index is called
(d) Θ(log n) for insertion and Θ(n) for deletion
(a) dense. (b) sparse.
(c) clustered. (d) unclustered. Solution: Worst case for insertion and deletion
could be if we have to traverse all the nodes. In
Solution: A clustered index is a type of index
that case, the complexity is Θ(n) in both.
where the table records are physically re-ordered
to match the index. Ans. (b)
Ans. (c)
2 2
23. In the LU decomposition of the matrix 
9
,
4 Q. NO. 26 – 55 Carry Two Marks Each
if the diagonal elements of U are both 1, then the
lower diagonal entry l22 of L is –———–—.
26. Suppose that the stop-and-wait protocol is used on
Solution: In LU decomposition, we have a link with a bit rate of 64 kilobits per second and
é2 2 ù él11 0 ù é1 u12 ù 20 milliseconds propagation delay. Assume that
êë4 9 úû = êl úê ú the transmission time for the acknowledgement
ë 21 l22 û ë0 1 û and the processing time at nodes are negligible.
Solving the above matrix, we have l11 = 2 . Now, Then the minimum frame size in bytes to achieve
l11u12 = 2 ⇒ u12 = 1 a link utilization of at least 50% is .
Also, l21 = 4 . So, l21u12 + l22 = 9 ⇒ l22 = 5
Solution: Minimum frame size means 100%
Ans. (5)
utilization.
24. The output of the following C program is .
RTT = 20 × 2 = 40 ms
void f1(int a, int b)
{ Bit rate = 64 Kbps
int c; 64 × 10 bytes ⇒ 1 sec
3
c=a; a=b; b=c;

40 ms ⇒ 64 × 103 × 40 × 10−3 = 2560 bits
}
void f2(int *a, int *b) 2560
{ Bytes ⇒ = 320 bytes
8
int c;
c=*a; *a=*b; *b=c; Hence, to achieve 50% utilization, bytes = 160
} Ans. (160)

27. Consider a max heap, represented by the array: 40, Now consider that a value 35 is inserted into this
30, 20, 10, 15, 16, 17, 8, 4. heap. After insertion, the new heap is
(a) 40, 30, 20, 10, 15, 16, 17, 8, 4, 35

Array 1 2 3 4 5 6 7 8 9 (b) 40, 35, 20, 10, 30, 16, 17, 8, 4, 15
Index
(c) 40, 30, 20, 10, 35, 16, 17, 8, 4, 15
Value 40 30 20 10 15 16 17 8 4 (d) 40, 35, 20, 10, 15, 16, 17, 8, 4, 30
Solution:
40 40
30 20 Insert 30 20
35
10 15 16 17 10 15 16 17
8 4 8 4 35
40 40
35 20 30 20
10 30 16 17 10 35 16 17
8 4 15 8 4 15
40, 35, 20, 10, 30, 16, 17, 8, 4, 15
Ans. (b)
28. Consider the following C program segment. Therefore, cyclomatic complexity = 4 + 1 = 5.

while(first <= last)
Ans. (5)
{
if(array[middle] < search)
first = middle + 1; 29. Consider a LAN with four nodes S1, S2, S3 and S4.
else if (array[middle] = = search) Time is divided into fixed-size slots, and a node
found = TRUE; can begin its transmission only at the beginning of
else last = middle – 1; a slot. A collision is said to have occurred if more
middle = (first + last)/2; than one node transmit in the same slot. The prob-
}
if (first > last) notPresent = TRUE; ability of generation of a frame in a time slot by S1,
S2, S3 and S4 are 0.1, 0.2, 0.3 and 0.4, respectively.
The cyclomatic complexity of the program segment The probability of sending a frame in the first slot
is ————. without any collision by any of these four stations
Solution: Cyclomatic complexity = Number of is .
predicate nodes + 1 Solution:
Predicate node: Decision making node, that is, P = P(S1) P(~S2) P(~S3) P(~S4) + P(~S1) P(S2)
wherever we check for some condition. P(~S3) P(~S4) + P(~S1) P(~S2) P(S3) P(~S4)
Here in our code, the number of predicate nodes = 4 + P(~S1) P(~S2) P(~S3) P(S4)

SET 1 597
= 0.1 × 0.8 × 0.7 × 0.6 + 0.9 × 0.2 × 0.7 × of the elements of . Let pr be the probability that
0.6 + 0.9 × 0.8 × 0.3 × 0.6 + 0.9 × 0.8 × an element (x, y, z) ∈ 3 chosen equiprobably satis-
0.7 × 0.4 fies x ∨ (y ∧ z) = (x ∨ y) ∧ (x ∨ z). Then
= 0.4404 (a) pr = 0 (b) pr = 1
Ans. (0.4404)
(c) 0 < pr £
1 1
(d) < pr < 1
30. The binary operator ≠ is defined by the following 5 5
truth table.
3
Solution: Number of elements in
p q p≠q
= Number of ways to choose 3 elements from 5
0 0 0
= 5 × 5 × 5 = 125 ways
0 1 1
Now, when we take x = t, then also the given con-
1 0 1
dition is satisfied for any y and z.
1 1 0
Number of ways y and z can be taken = 5 × 5 =
Which one of the following is true about the binary 25 ways.
operator ≠?
If we take x = p, y = p, z = r, the given condition
(a) Both commutative and associative is satisfied. So, pr > (25 / 125) > 1/5
(b) Commutative but not associative
(c) Not commutative but associative For x = t, y = r, z = p, the given condition is not
(d) Neither commutative nor associative satisfied. So, pr ≠ 1.
Solution: Looking at the truth table, we analyze So, option (d) is correct.
that this is the same as XOR operation. XOR is Ans. (d)
both commutative (i.e., x*y = y*x) and associative
33. Consider the operations
(i.e., a*(b*c) = (a*b)*c).
Ans. (a) f(X,Y,Z) = XÝZ + XY´ + Y´Z´ and
99 g(X,Y,Z) = XÝZ + XÝZ´ + XY.
31. ∑
1
= _______. Which one of the following is correct?
x =1 x(x + 1)
(a) Both {f} and {g} are functionally complete
Solution: (b) Only {f} is functionally complete
99 (c) Only {g} is functionally complete
å x(x + 1) = 1 × 2 + 2 × 3 + 3 × 4 + + 99 × 100
1 1 1 1 1
(d) Neither {f} nor {g} is functionally complete
x =1
2− 1 3− 2 100 − 99 Solution: A system of Boolean functions is func-
= + + + tionally complete iff function satisfies following
1× 2 2 × 3 99 × 100
properties:
1 1 1 1 1 1
= 1− + − + − + − ••at
2 2 3 3 99 100 least one function that does not preserve zero
1 ••at least one function that does not preserve one
= 1− = 0.99
100 ••at least one function that is not linear
Ans. (0.99) ••at least one function that is not monotone
32. Suppose = {p,q,r,s,t} is a lattice represented by ••at least one function that is not self-dual
the following Hasse diagram:
{f} satisfies all the above properties.
t Ans. (b)
34. Let G be a connected planar graph with 10 verti-
q r s ces. If the number of edges on each face is three,
then the number of edges in G is .
p Solution: Using formula: R + V = E+ 2

For any x, y ∈ , not necessarily distinct, x ∨ y and 3R= 2E
x ∧ y are join and meet of x, y, respectively. Let 3 =
{(x,y,z): x,y,z ∈ } be the set of all ordered triplets R = (2/3)E

(2/3)E +V = E+ 2  E− (2/3)E= 8 Solution:
E= 24 p =q +r
Ans. (24) 1 s =p +q
u =s ∗v
35. Let an represent the number of bit strings of length
n containing two consecutive 1’s. What is the
recurrence relation for an?
(a) an-2 + an-1 + 2n-2 2 v =r +u 3 q =s ∗u
(b) an-2 + 2an-1 + 2n-2
(c) 2an-2 + an-1 + 2n-2
(d) 2an-2 + 2an-1 + 2n-2
q =v +r
Solution: Suppose n = 3, possible strings are ⇒
4
011, 110, 111
Checking the three conditions given in question,
For n = 2 ⇒ 11 is the only possible string
r is live at block2 since r is used in block4 and has a
a3 = a1 + a2 + 21 = 0 + 1 + 2 = 3 ⇒ True path from block2 to block4 in the control flow graph.
a3 = a1 + 2a2 + 21 = 0 + 2 + 2 = 4 ⇒ False Also there isn’t any modification of r in the path.
a3 = 2a1 + a2 + 21 = 0 + 1 + 2 = 3 ⇒ True
Hence, r is live at block2. Similarly, check for u.
Ans. (c)
a3 = 2a1 + 2a2 + 21 = 0 + 2 + 2 = 4 ⇒ False
Now for options (a) and (c), check by considering n 37. The least number of temporary variables required
= 4. The answer comes out to be (a). to create a three-address code in static single
assignment form for the expression q+r/3+s—
Ans. (a)
t*5+u*v/w is –––.
36. A variable x is said to be live at a statement Si in Solution: We will need a temporary variable for
a program if the following three conditions hold storing the result of each binary operation.
simultaneously:
q + r / 3 + s — t * 5 + u * v/w
I. There exists a statement Sj that uses x t1 = r/3;
II. There is a path from Si to Sj in the flow graph t2 = t*5;
corresponding to the program t3 = u*v;
III. The path has no intervening assignment to x t4 = t3/w;
including at Si and Sj t5 = q + t1;
t6 = t5 + s;
t7 = t5 — t2;
p =q +r
1
t8 = t7 + t4
s =p +q So, a total of 8 temporary variables are required.
u =s ∗v Ans. (8)
38. Consider an Entity-Relationship (ER) model in
which entity sets E1 and E2 are connected by an
m:n relationship R12. E1 and E3 are connected by
2 v =r +u 3 q =s ∗u a 1: n (1 on the side of E1 and n on the side of E3)
relationship R13.
E1 has two single-valued attributes a11 and a12
q =v +r
of which a11 is the key attribute. E2 has two single-
4
valued attributes a21 and a22 of which is the key
attribute. E3 has two single-valued attributes a31
The variables which are live at the statement in and a32 of which a31 is the key attribute. The rela-
basic block 2 and at the statement in basic block 3 tionships do not have any attributes.
of the above control flow graph are If a relational model is the derived from the
above ER model, then the minimum number of
(a) p, s, u (b) r, s, u relations that would be generated if all the rela-
(c) r, u (d) q, v tions are in 3NF is .

SET 1 599
Solution: The relations are as shown: (a) 10110 (b) 10010

E1: <a11, a12> (c) 01010 (d) 01001
E2: <a21, a22> Solution: This PDA is accepts all strings of the
E3: <a31, a32, a11> form x0x’ or x1x’, where x’ is the reverse of the
1’s complement of x. The given string 101100 has
E1−E3: <a11, a21>for m:n relationship E1 − E2 6 letters and we are given 5 letter strings. So, x0
We cannot combine any relation here as it will give is done, with x = 10110. Now, the stack contains
rise to partial functional dependency and thus vio- 10110 and remaining string should be opposite and
late 3NF. reverse to it, so that the stack becomes empty. So,
Ans. (4) 1’s complement of x = 01001 and x’ = 10010.
Ans. (b)
39.
41. Let G = (V, E) be a simple undirected graph, and
b a a b s be a particular vertex in it called the source.
a
N:
b For x ∈ V, let d(x) denote the shortest distance
M:
in G from s to x. A breadth first search (BFS) is
performed starting at s. Let T be resultant BFS
b a tree. If (u, v) is an edge of G that is not in T, then
which one of the following CANNOT be the value
Consider the DFAs M and N given above. The
of d(u) — d(v)?
number of states in a minimal DFA that accepts
the languages L(M) ∩ L(N) is . (a) —1 (b) 0 (c) 1 (d) 2
Solution: M accepts the strings ending with à’. Solution: In BFS traversal of given graph, we can
N accepts the strings ending with `b’. Thus, their either visit node u first or node v. Let us assume
intersection should accept an empty string. node u if visited first. Then all neighbors of node u
will be in the queue to be visited. Since v is also a
neighbor of u and v is not visited before, d(v) will
become d(u) + 1.
So, d(u) − d(v) = 2 is not possible.
a,b Ans. (d)
Ans. (1)
42. Consider a uniprocessor system executing three
40. Consider the NPDA tasks T1, T2 and T3, each of which is composed
of an infinite sequence of jobs (or instances) which
〈Q = {q0, q1, q2}, ∑ = {0, 1}, Γ = {0, 1, Γ}⊥ d, q0, arrive periodically at intervals of 3, 7 and 20
⊥, F = {q2}〉, milliseconds, respectively. The priority of each task
where (as per usual convention) Q is the set of is the inverse of its period, and the available tasks
states, ∑ is the input alphabet, d is the stack alpha- are scheduled in order of priority, with the highest
bet, d is the state transition function, q0 is the ini- priority task scheduled first. Each instance of T1,
tial state, ⊥ is the initial stack symbol, and F is T2 and T3 requires an execution time of 1, 2 and
the set of accepting states. The state transition is 4 milliseconds, respectively. Given that all tasks
as follows: initially arrive at the beginning of the 1st milli-
second and task preemptions are allowed, the first
instance of T3 completes its execution at the end of
1, Z → 1 Z 0, 1Z → Z milliseconds.
0, Z → 0 Z 1, 0Z → Z
q0 q1 q2 Solution:
T1 T2 T1 T3 T1 T3 T2 T1 T2 T3
0/1/∈, Z → Z
1 2 4 5 7 8 9 10 11 12 13
∈, ⊥ → ∈
Ans. (13)
Which one of the following sequences must follow 43. A positive edge-triggered D flip-flop is connected
the string 101100 so that the overall string is to a positive edge-triggered JK flip-flop as follows.
accepted by the automaton? The Q output of the D flip-flop is connected to both

the J and K inputs of the JK flip-flop, while the Q that is, 60 seconds = 10000 rotations
output of the JK flip-flop is connected to the input 60
of the D flip-flop is connected to both the J and K 1 rotation = = 6 ms
10000
inputs of the JK flip-flop, while the Q output of Rotational latency = 6/2 = 3 ms
the JK flip-flop is connected to the input of the D
flip-flop. Initially, the output of the D flip-flop is One rotation = One track
set to logic one and the output of the JK flip-flop Number of sectors in one track = 600 ⇒ 600 sectors
is cleared. Which one of the following is the bit = 6 ms
sequence(including the initial state) generated at the One sector ⇒ 6/600 = 0.01 ms
Q output of the JK flip-flop when the flip-flops are
The file contains 2000 sectors.
connected to a free-running common clock? Assume
that J = K = 1 is the toggle mode and J = K = 0 is Therefore, 2000 sectors take = 2000 × 0.01 = 20 ms
the state-holding mode of the JK flip-flop. Both the Hence, total time needed to read the file
= (Number of sectors) × (Seek time + Rotational
flip-flops have non-zero propagation delays.
(a) 0110110….. (b) 0100100…. latency) + (Time for 2000 sectors)
(c) 011101110…. (d) 011001100… = 2000 × (4 + 3) + 20 =14020
Solution: Ans. (14000—14020)
45. Consider a non-pipelined processor with a clock
rate of 2.5 gigahertz and average cycles per instruc-
tion of four. The same processor is upgraded to a
D Q J Q1 pipelined processor with five stages; but due to the
internal pipelined delay, the clock speed is reduced
FF1 FF2 to 2 gigahertz. Assume that there are no stalls in
the pipeline. The speed up achieved in this pipe-
K lined processor is .
Solution: Non-pipelined method:
Clock Q Q1
Frequency = 2.5 GHz
0 (initial) 1 0
Time required for one cycle = 1/(2.5) = 0.4 ms
1 0 1
Each instruction requires 4 cycles. Therefore, time
2 1 1 for each instruction = 0.4 × 4 = 1.6 ms
3 1 0 Pipelined structure:
4 0 1 Frequency = 2 GHz
5 1 1 Time = 0.5 ms
Enhanced time for each stage in pipeline structure =
6 1 0
time for one stage of pipeline = 0.5 ms
7 0 1 Non-pipelined structure time 1.6
Speed up = = = 3.2
Ans. (a) Enhanced time in pipeline 0.5
Non-pipelined structure time 1.6
44. Consider a disk pack with a seek time of 4 = = 3.2
Enhanced time in pipeline
milliseconds and rotational speed of 10000 rotations 0.5
Ans. (3.2)
per minute (RPM). It has 60 sectors per track and
each sector can store 512 bytes of data. Consider a 46. Suppose the following disk request sequence (track
file stored in the disk. The file contains 2000 sec- numbers) for a disk with 100 tracks is given: 45,
tors. Assume that every sector access necessitates a 20, 90, 10, 50, 60, 80, 25, 70. Assume that the ini-
seek, and the average rotational latency for access- tial position of the R/W head is on track 50. The
ing each sector is half of the time for one complete additional distance that will be traversed by the
rotation. The total time (in milliseconds) needed to R/W head when the Shortest Seek Time First
read the entire file is . (SSTF) algorithm is used compared to the SCAN
(Elevator) algorithm (assuming that SCAN algo-
Solution: Seek time = 4 ms rithm moves towards 100 when it starts execution)
RPM = 10000 rotations is tracks.

SET 1 601
Solution: SSTF:
SCAN: 1 10 20 25 45 50 60 70 80 90 100
1 10 20 25 45 50 60 70 80 90 100 10 15
10 10 10 10
10 10
10 5 65
10 10
55
5 20 Total Head Moments = 10 + 15 + 10 + 10 + 10 + 65
+ 5 + 10 = 130
10
Total Head Moments = 10 + 10 + 10 + 10 + 10 + 55 Additional distance traversed = 140 — 130 = 10

+20 + 5 + 10 = 140 Ans. (10)
47. Consider a main memory with five page frames (a) Both incur the same number of page faults.
and the following sequence of page references: 3, (b) FIFO incurs 2 more page faults than LRU
8, 2, 3, 9, 1, 6, 3, 8, 9, 3, 6, 2, 1, 3. Which one of (c) LRU incurs 2 more page faults than FIFO
the following is true with respect to page replace- (d) FIFO incurs 1 more page faults than LRU
ment policies First In First Out (FIFO) and Least
Recently Used (LRU)?
Solution:
FCFS:
1 1 1 1 1 1 1 1 1 1
9 9 9 9 9 9 9 9 2 2 2
2 2 2 2 2 2 8 8 8 8 8 8 8
8 8 8 8 8 8 3 3 3 3 3 3 3 3
3 3 3 3 3 3 6 6 6 6 6 6 6 6 6
Hit Hit Hit Hit Hit Hit
3 8 2 3 9 1 6 3 8 9 3 6 2 1 3
Page faults for FIFO = 9
LRU:
1 1 1 1 1 1 1 2 2 2
9 9 9 9 9 9 9 9 9 9 9
2 2 2 2 2 2 8 8 8 8 8 1 1
8 8 8 8 8 6 6 6 6 6 6 6 6 6
3 3 3 3 3 3 3 3 3 3 3 3 3 3 3
Hit Hit Hit Hit Hit Hit
3 8 2 3 9 1 6 3 8 9 3 6 2 1 3
Page faults for LRU = 9

Numbers of page faults are equal.
So, option (a) is correct.
Ans. (a)

2 /p
∫1/p
cos(1 / x) Insertion will take a constant time, that is, Θ(1).
48. dx = _____.
x2 N insert operations need = N × Θ(1) = Θ(N) time.
2 /p
Solution: Substituting 1/x = t in ∫
cos(1 / x) Delete operations will also be done in constant time
dx
1 /p
x2 since the pointer to be deleted is provided and no
Differentiating both sides, we get −1/x2 dx = dt time will be required for search.
t = 1/x = p and p /2 (log N)1/2 delete operations will be completed in =
p Θ((log N)1/2) time
Therefore, ∫p /2 cos(t) dt = sin p − sin p / 2 = −1 Decrease key operation will also take = Θ((log
Ans. (−1) N)1/2) time
If we use sorted order array, insertion will take
49. Consider the following 2 × 2 matrix A where two more time since we need to find its exact position
elements are unknown and are marked by a and b. in the list where to insert the data. And in the case
The eigenvalues of this matrix are −1 and 7. What of min heap, more time will be required in inser-
are the values of a and b? tion and deletion since heapify algorithm needs to
 1 4 be implemented after each insertion or deletion.
A=
 b a  Ans. (a)
(a) a = 6, b = 4 (b) a = 4, b = 6 51. Consider the following relations:

(c) a = 3, b = 5 (d) a = 5, b = 3 Students
Solution: Eigenvalues of matrix, l1 = −1 and
l2 = 7 Roll-No Student—Name
We know that, l1 + l2 = Trace (i.e., sum of diago- 1 Raj
nal elements) 2 Rohit
l1 × l2 = Determinant 3 Raj
l1 + l2 = −1 + 7 = 6 = 1 + a Þ a = 5
Performance
l1 × l2 = −7 = a − 4b Þ −7 = 5 − 4b Þ b = 3
Roll—No Course Marks
Ans. (d)
1 Math 80
1/2
50. An algorithm performs (log N) find operations, 1 English 70
N insert operations, (log N)1/2 delete operations, 2 Math 75
and (log N)1/2 decrease-key operations on a set
of data items with keys drawn form a linearly 3 English 80
ordered set. For a delete operation, a pointer is 2 Physics 65
provided to the record that must be deleted. For 3 Math 80
the decrease-key operation, a pointer is provided
to the record that has its key decreased. Which one
Consider the following SQL query.
of the following data structures is the most suited
for the algorithm to use, if the goal is to achieve SELECT S. Student—Name, sum (P.Marks)
the best asymptotic complexity considering all the FROM Student S, Performance P
operations?
WHERE S.Roll—No = P.Roll—No
(a) Unsorted array GROUP BY S.Student—Name
(b) Min—heap The number of rows that will be returned by the
(c) Sorted array SQL query is .
(d) Sorted doubly linked list
Solution: Since here group by Name is done,
Solution: Suppose if we use an unsorted array therefore marks for roll nos. 1 and 3 are added
then, search operation will require Θ(N) time. together giving 310 and marks for roll no. 2, that is,
(logN)1/2 search operations will take = Θ(N(log Rohit are added up to give 140. Thus, the resulting
N)1/2) time. output table is:

SET 1 603
Solution:
Student_Name Marks
15
Raj 310 B E
Rohit 140 4
Ans. (2)
A 2 F
52. What is the output of the following C code?
Assume that the address of x is 2000 (in decimal) 9 6
and an integer requires four bytes of memory? C D
int main()
{ B
15
E
unsigned int x[4][3] = {(1,2,3),
{4,5,6}, {7, 8, 9}, {10, 11, 12}}; 10 4
printf(“%u, %u, %u”, x+3, *(x+3),
*(x+2)+3); A 2 7 F
}
9 6
(a) 2036, 2036, 2036
(b) 2012, 4, 2204 C D
16
Sum = 10 + 9 + 2 + 15 + 16 + 7 + 4 + 6 = 69
(c) 2036, 10, 10
(d) 2012, 4, 6
Ans. (69)
Solution:
1 2 3 54. Consider the following C function.
4 5 6 int fun1 (int n)
7 8 9 {
10 11 12 int i, j, k, p, q = 0;
for (i = 1; i < n; ++i)
As given in question, integer takes 4 bytes to {
store p = 0;
for(j = n; j > 1; j = j/2)
Here, X + 3 ⇒ 2000 + (3 × 3) × 4 = 2036 ++p;
*(X + 3) ⇒ Value at address 2036, %u will print for(k = 1; k < p; k = k*2)
++q;
the address only, that is, 2036 }
*(X + 2) + 3 ⇒ 2000 + (3 × 2)4 + 3 × 4 = 2036 return q;
}
Ans. (a)
Which one of the following most closely approxi-
53. The graph shown below has 8 edges with distinct mates the return value of the function fun1?
integer edge weights. The minimum spanning tree
(MST) is of weight 36 and contains the edges: (a) n3 (b) n(log n)2
{(A, C), (B, C), (B, E), (E, F), (D, F)}. The edge (c) n log n (d) n log(log n)
weights of only those edges which are in the MST Solution: First for loop will execute n times, the
are given in the figure shown below. The minimum others are sequential not nested, so both execute
possible sum of weights of all 8 edges of this graph log(n) and log(p) times but the value returned by
is . fun1 is “q” which is calculated in the second loop.
15 So, n × (log(p)), [where p = log (n)] ⇒ n log (log
B E n)times.
4 Ans. (d)
A 2 F 55. Consider the following pseudo code, where x and y

are positive integers.
9 6
begin
C D q: = 0
r: = x

while r ≥ y do (c) {y = qx + r ∧ 0 < r < y}

being (d) {q + 1 < r — y ∧ y > 0}
r: = r – y
q: = q + 1 Solution: In this program, q counts the number of
end times y is subtracted from x and at the remaining
end r value will act as remainder.
The post condition that needs to be satisfied after So, x = qy + r and r < y.
the program terminates is Ans. (b)
(a) {r = qx + y ∧ r < y}
(b) {x = qy + r ∧ r < y}

SOLVED GATE(CS) 2015
SET 2
Q. No. 1 – 25 Carry One Mark Each
1. Consider the following two statements. (a) R is symmetric and reflexive but not transitive
S1: If a candidate is known to be corrupt, then he (b) R is reflexive but not symmetric and not transitive
will not be elected (c) R is transitive but not reflexive and not symmetric
(d) R is symmetric but not reflexive and not transitive
S2: If a candidate is kind, he will be elected
Which one of the following statements follows from Solution: R is not reflexive since a and b are dis-
S1 and S2 per sound interference rules of logic? tinct as given in the question. Now, the given rela-
tion is true for a = 4 and b = 8; this is also true for
(a) If a person is known to be corrupt, he is kind
a = 8 and b = 4. Hence, it satisfies the condition
(b) If a person is not known to be corrupt, he is
for symmetric relation, that is, if (a,b) satisfies then
not kind
(b,a) should satisfy the relation.
(c) If a person is kind, he is not known to be corrupt
(d) If a person is not kind, he is not known to be R does not satisfy transitivity. For example, a = 5,
corrupt b = 10, c = 12. So, (5,10) satisfies the relation,
(10,12) also satisfies the relation, but (5,12) does
Solution: Let p: Candidate is known to be corrupt not. Hence, the relation is not transitive.
q: Candidate is kind Ans. (d)
r: Candidate is elected
4. The number of divisors of 2100 is .
S1: p → ~r => r → ~p
Solution:
S2: q → r
2100 = 22 + 3 × 52 × 7
Therefore, q → ~p, that is, if a candidate is kind,
he is not known to be corrupt. Number of divisors = (2 + 1) × (1 + 1) × (2 + 1) ×
Ans. (c) (1 + 1) = 36
Ans. (36)
2. The cardinality of the power set of {0, 1, 2, … 10}
is . 5. The larger of the two eigenvalues of the matrix
Solution: The cardinality of the power set of 4 5
{0, 1, 2, … , 10} is 2 number of elements in set, that is, 2 1 is .
211 = 2048. Solution:

Ans. (2048) (4−l) (1−l) − 10 = 0 ⇒ 4 − 4l − l + l2 = 10
l2 − 5l − 6 = 0 ⇒ l = 6, −1
3. Let R be the relation on the set of positive integers
such that aRb if and only if a and b are distinct and
have a common divisor other than 1. Which one of Hence, the larger value of l = 6
the following statements about R is true? Ans. (6)

6. An unordered list contains n distinct elements. The III. If L is a language in NP, L is Turing decidable.
number of comparisons to find an element in this Which of the above statements is/are true?
list that is neither maximum nor minimum is (a) Only II (b) Only III
(a) Θ(n log n) (b) Θ(n) (c) Only I and II (d) Only I and III
(c) Θ(log n) (d) Θ(1)
Solution: Turing decidable languages are recur-
Solution: In an unordered list, if we consider any sive languages. Recursive languages are closed
three elements, one of them would neither be maxi- under complement, that is, complement of every
mum or minimum. Hence, in constant time we can recursive language is recursive.
find the required element. Ans. (d)
Ans. (d)
11. Consider the following function written in the C
7. The minimum number of JK flip-flops required to programming language.
construct a synchronous counter with the count
sequence (0, 0, 1, 1, 2, 2, 3, 3, 0, 0, …….) is . void foo (char *a) {
if (*a && *a != ' ')
Solution: Number of distinct states = 4. Hence,
{
minimum flip-flops required is 2.
Ans. (2) foo (a+1);
putchar *a;
8. Assume that for a certain processor, a read request
takes 50 nanoseconds on a cache miss and 5 nano- }
seconds on a cache hit. Suppose while running a }
program, it was observed that 80% of the proces- The output of the above function on input “ABCD
sors read requests result in a cache hit. The aver- EFGH” is
age and access time in nanoseconds is . (a) ABCD EFGH (b) ABCD
Solution: Given that cache hit rate = 0.8, so (c) HGFE DCBA (d) DCBA
cache miss rate = 0.2 Solution: If condition fails when the space is
Time on cache miss = 50 ns and time on cache hit encountered, then output will be DCBA.
= 5 ns
Ans. (d)
So, average access time = 0.8(5) + 0.2(50) = 14 ns
Ans. (14) 12. Consider a complete binary tree where the left and
the right subtrees of the root are max-heaps. The
9. A computer system implements a 40-bit virtual lower bound for the number of operations to con-
address, page size of 8 kilobytes, and a 128-entry vert the tree to a heap is
translation look-aside buffer (TLB) organized into
(a) Ω(log n) (b) Ω(n)
32 sets each having four ways. Assume that the
(c) Ω(n log n) (d) Ω(n2)
TLB tag does not store any process id. The mini-
mum length of the TLB tag in bits is . Solution: Left and right subtrees of binary tree are
max heaps. Thus, to convert the tree into heap we
Solution:
only need to compare the roots. This would always
40 be done on the number of levels which is log(n).
Ans. (a)
22 5 13 13. A binary tree T has 20 leaves. The number of nodes
in T having two children is .
TAG SET-OFFSET WORD-OFFSET
Page size = 8 KB = 210 × 23 = 213 ⇒ 13 byte Solution: The number of nodes with two children
word-offset. is always one less than the number of leaf nodes.
Number of sets = 32 = 25 ⇒ 5 bytes set-offset Here, Leaf nodes = 20 ⇒ Number of two children
nodes = 19.
Remaining bytes are for tag = 40 − 18 = 22 Ans. (19)
Ans. (22)
14. Consider the following C function.
10. Consider the following statements
int fun(int n)
I. The complement of every Turing decidable
{
language is Turing decidable.
II. There exists some language which is in NP but int x=1, k;
is not Turing decidable. if (n==1) return x;

SET 2 607
for (k=1; k<n; ++k) Register allocation can be used in graph coloring.
x = x + fun (k) * fun (n-k); Post-order traversal is used for expression
return x; evaluations.
} Ans. (c)
The return value of fun (5) is . 18. In the context of abstract-syntax-tree (AST) and
Solution: control-flow-graph (CFG), which one of the follow-
ing is TRUE?
For n = 1, f (n) = 1
n -1 (a) In both AST and CFG, let node, N2 be the
For n > 1, f (n) = 1 + ∑k
=1 f(k) f(n - k) successor of node N1. In the input program,
f(5) = 51 the code corresponding to N2 is present after
Ans. (51) the code corresponding in N1.
15. A software requirements specification (SRS) doc- (b) For any input program, neither AST nor CFG
ument should avoid discussing which one of the will contain a cycle.
following? (c) The maximum number of successors of a node
in an AST and a CFG depends on the input
(a) User interface issues program.
(b) Non-functional requirements (d) Each node is AST and CFG corresponds to
(c) Design specification atmost one statement in the input program.
(d) Interfaces with third party software
Solution: Option (a) is false in the case when
Solution: Design specifications are not discussed in CFG contains a cycle. Hence, option (b) is also
SRS document. Only design constraints are discussed false as it says CFG cannot contain cycle. Also,
Ans. (c) single block can contain multiple statements. Thus,
16. Consider two decision problems Q1, Q2 such that option (d) is also false.
Q1 reduces in polynomial time to 3-SAT and Ans. (c)
3 -SAT reduces in polynomial time to Q2. Then 19. Consider the basic COCOMO model where E is the
which one of following is consistent with the above effort applied in person-months, D is the develop-
statement? ment time in chronological months, KLOC is the
(a) Q1 is in NP, Q2 is NP hard estimated number of delivered lines of code (in
(b) Q1 is in NP, Q2 is NP hard thousands) and ab, bb, cb, db have their usual mean-
(c) Both Q1 and Q2 are in NP ings. The basic COCOMO equations are of the form
(d) Both Q1 and Q2 are NP hard
(a) E = ab (KLOC ) exp (bb ), D = cb (E ) exp (db )
Solution: 3-SAT is NPC problem. (b) D = ab (KLOC ) exp (bb ), E = cb (D ) exp (db )
Q1 ≤p 3-SAT and 3-SAT ≤p Q2 (c) E = ab exp (bb ), D = cb (KLOC ) exp (db )
If an NP problem A is polynomial time reducible to (d) E = ab exp (Db ), D = cb (KLOC ) exp (bb )
another NP problem B, then B is NPC. But if we
do not know the type of B problem, then B is said to Solution: Basic COCOMO model equations are of
be NP hard. Hence, Q1 is in NP and Q2 is NP hard. the form:
Ans. (a) Effort = ab(KLOC ) exp(bb)
17. Match the following: Development time = cb(Effort) exp(db)
Ans. (a)
(P) Lexical analysis (1) Graph coloring
20. A system has 6 identical resources and N processes
(Q) Parsing (2) DFA minimization
competing for them. Each process can request at
(R) Register allocation (3) Post-order traversal most 2 resources. Which one of the following values
(S) Expression (4) Production tree of N could lead to a deadlock?
evaluation
(a) 1 (b) 2
(a) P − 2, Q − 3, R −1, S − 4 (c) 3 (d) 4
(b) P − 2, Q − 1, R − 4, S − 3
(c) P − 2, Q − 4, R − 1, S − 3 Solution: No deadlock occurs when N = 1, N = 2
(d) P − 2, Q − 3, R − 4, S − 1 and N = 3. When N = 4, deadlock may occur as
two processes will get complete resources; and two
Solution: Lexical analyzer uses DFA. will wait for the completion of other two.
Parsing is done though production tree. Ans. (d)

21. Consider the following transaction involving two transmission delay, and that its propagation delay
bank accounts x and y. is the same as the data propagation delay. Also
read (x) ; x : = x — 50; write (x); read (y); assume that the processing delays at the nodes are
y: = y + 50; write (y) negligible. The efficiency of the stop-and-wait pro-
The constraint that the sum of the accounts x and tocol in this setup is exactly 25%. The value of
y should remain constant is that of the one-way propagation delay (in milliseconds) is
.
(a) Atomicity (b) Consistency
(c) Isolation (d) Durability 1 1
Solution: In stop and wait, efficiency = =
1 1 (1 + 2a) 4
Solution: Consistency means the overall state of = ⇒ a = 3/2
system should remain constant before and after the (1 + 2a) 4
transaction. Hence, the sum remains constant with 1000 × 8
Transmission time, Tt = = 8 ms
the property of consistency. 106
8
Ans. (b) Propagation time, Tp = a × Tt = 3 × = 12 ms
2
22. With reference to the B+ tree index of order 1 Ans. (12)
shown below, the minimum number of nodes
(including the Root node) that must be fetched 25. Which one of the following statements is NOT cor-
in order to satisfy the following query: “Get all rect about HTTP cookies?
records with a search key greater than or equal to (a) A
cookie is a piece of code that has the potential
7 and less than 15” is . to compromise the security of an internet user
(b) A cookie gains entry to the user’s work area
9
through an HTTP header
(c) A cookie has an expiry date and time
5 13 17 (d) Cookies can be used to track the browsing
pattern of a user at a particular site
Solution: A cookie cannot compromise the secu-

1 3 5 7 9 11 13 15 17 rity of an internet user.
Solution: Ans. (a)
1 9
Q. No. 26 – 55 Carry Two Marks Each
2 5 4 13 17
26. Consider the following routing table at an IP router:
1 3 5 7 9 11 13 15 17 Network No. Net Mask Next Hop

128.96.170.0 255.255.254.0 Interface 0
3 5 6
128.96.168.0 255.255.254.0 Interface 1
Ans. (6)
128.96.166.0 255.255.254.0 R2
23. Identify the correct order in which a server process
128.96.164.0 255.255.252.0 R3
must invoke the function calls accept, bind, listen,
and recv according to UNIX socket APL 0.0.0.0 Default R4
(a) listen, accept, bind, recv For each IP address in Group I identify the correct
(b) bind, listen, accept, recv choice of the next hop from Group II using the
(c) bind, accept, listen, recv entries from the routing table above.
(d) accept, listen, bind, recv
Group I Group II
Solution: The correct order is bind, listen, accept,
recv. First three are used for connection establish- (i) 128.96.171.92 (a) Interface 0
ment and recv is used for data transfer operation. (ii) 128.96.167.151 (b) Interface 1
Ans. (b) (iii) 128.96.163.151 (c) R2
6
24. A link has a transmission speed of 10 bits/sec. (iv) 128.96.165.121 (d) R3
It uses data packets of size 1000 bytes each. (e) R4
Assume that the acknowledgement has negligible

SET 2 609
(a) i − a, ii − c, iii − e, iv − d 8880

(b) i − a, ii − d, iii − b, iv − e Number of datagrams = =7
1472
(c) i − b, ii − c, iii − d, iv − e
(d) i − b, ii − c, iii − e, iv − d and offset =
8872
= 1109 + 1 = 1110
8
Solution: Ans. (c)
(i) 28. Assume that the bandwidth for a TCP connection
is 1048560 bits/sec. Let a be the value of RTT
IP Address: 128.96.171.92
in milliseconds (rounded off to the nearest inte-
Subnet Mask: 255.255.254.0 ger) after which the TCP window scale option is
needed. Let b be the maximum possible window
Network Address: 128.96.170.0 size the window scale option. Then the values of a
and b are
This matches with Interface 0.
(a) 63 milliseconds, 65535 × 214
(ii) (b) 63 milliseconds, 65535 × 216
IP Address: 128.96.167.151 (c) 500 milliseconds, 65535 × 214
(d) 500 milliseconds, 65535 × 216
Subnet Mask: 255.255.254.0
Solution: Bandwidth = 1048560 bits/sec
Network Address: 128.96.166.0 TCP window scale option is needed when size
increases more than 65535 B.
This matches with R2.
65535 × 8
Time to send 65535 B = = 500 ms
1048560 × 1000
(iii)
IP Address: 128.96.163.151 So, a = 500 ms.
Subnet Mask: 255.255.252.0 Maximum window size with window scale option
(b) = 65535 × 214
Network Address: 128.96.170.0 Ans. (c)
This matches with R4. 29. Consider a simple checkpointing protocol and the
(iv) following set of operations in the log.
IP Address: 128.96.165.121 (start, T4); (write, T4, y, 2, 3); (start, T1);

Subnet Mask: 255.255.254.0 (commit, T4); (write, T1, z, 5, 7);
(checkpoint);
Network Address: 128.96.164.0 (start, T2); (write, T2, x, 1, 9);
This matches with R3. (commit, T2); (start, T3),
So, option (a) is correct. (write, T3, z, 7, 2);
If a crash happens now and the system tries to
Ans. (a)
recover using both undo and redo operations,
27. Host A sends a UDP datagram containing 8880 what are the contents of the undo lists and the
bytes of user data to host B over an Ethernet LAN. redo list?
Ethernet frames may carry data up to 1500 bytes
(i.e. MTU = 1500 bytes). Size of UDP header is (a) Undo: T3,T1; Redo: T2
8 bytes and size of IP heard is 20 bytes. There is no (b) Undo: T3,T1; Redo: T2,T4
option field in IP header. How many total number (c) Undo: none; Redo: T2,T4,T3,T1
of IP fragments will be transmitted and what will (d) Undo: T3,T1,T4; Redo: T2
be the contents of offset field in the last fragment?
Solution: The transactions that are not committed
(a) 6 and 925 (b) 6 and 7400 must not be saved in database. Hence, we undo
(c) 7 and 1110 (d) 7 and 8880 transactions T1 and T3 here as they are not
yet committed. Transaction that is after the
Solution: Datagram size = 1500 bytes checkpoint must be redone. Hence, we redo the
Data in one datagram = 1500 − 20 − 8 = 1472 bytes transaction T2.
Total data to send = 8880 bytes Ans. (a)

30. Consider two relations R1(A,B) with the tuples 50 × 106 bytes/sec. if the average seek time of the
(1,5), (3,7) and R2 (A,C) = (1,7), (4,9). Assume disk is twice the average rotational delay and the
that R(A,B,C) is the full natural outer join of controller’s transfer time is 10 times the disk trans-
R1 and R2. Consider the following tuples of the fer time, the average time (in milliseconds) to read
form (A,B,C): a = (1,5,null), b = (1,null,7) c = or write a 512-byte sector of the disk is .
(3,null,9), d = (4,7,null), e = (1,5,7), f = (3,7,null),
g = (4,null,9).Which one of the following state- Solution:
ments is correct? RPM = 15000
(a) R contains a, b, e, f, g but not c, d. 15000 rotations ⇒ 60 sec
(b) R contains all of a, b, c, d, e, f, g 1 rotation ⇒ 4 ms = average seek time
(c) R contains e, f, g but not a, b Rotational latency = 4/2 = 2 ms
(d) R contains e but not f, g
Transfer rate = 50 × 106 bytes/sec
Solution: In option (c) 50 × 106 bytes ⇒ 1 sec
512
A B A C 512 bytes ⇒ = 0.01 ms ⇒ Disk transfer time
(50 × 106 )
R1 ⇒ 1 5 R2 ⇒ 1 7 time
3 7 4 9 Controller’s transfer time = 10 × (Disk
transfer
time) = 10 × 0.01 = 0.1 ms
A B C Average time = 4 + 2 + 0.1 = 6.1 ms
Ans. (6.1)
R⇒ 1 5 7
33. A computer system implements 8 kilobyte pages
3 7 Null
and a 32-bit physical address space. Each page
4 Null 9 table entry contains a valid bit, a dirty bit, three
permission bits, and the translation. If the maxi-
a = (1,5,null), b = (1,null,7) c = (3,null,9), d = (4,7,null), mum size of the page table of a process is 24 mega-
e = (1,5,7), f = (3,7,null), g = (4,null,9). bytes, the length of the virtual address supported
a, b, c and d are not the rows of relation R but e, f, g by the system is bits.
are the rows in relation R. Hence, option (c) is correct.
Solution: Page size = 8 KB = 213
Ans. (c) Physical address space = 32 bits
31. Consider six memory partitions of sizes 200 KB, Number of frames = 232/213 = 219 frames
400 KB, 600 KB, 500 KB, 300 KB and 250 KB,
Number of bits for page table entry = 19 + 1 V +
where KB refers to kilobyte. These partitions need
1 D + 3 P = 24 bits
to be allotted to four processes of sizes 357 KB,
210 KB, 468 KB and 491 KB in that order. If the Page table size = 24 MB = (24 × 220 × 8) bits
best fit algorithm is used, which partitions are = 24 × (Number of pages)
NOT allotted to any process?
So, number of pages = 223
(a) 200 KB and 300 KB (b) 200 KB and 250 KB Therefore, virtual address apace contains
(c) 250 KB and 300 KB (d) 300 KB and 400 KB (23 + 13) = 36 bits
Solution: Partitions needed for the processes = Ans. (36)
357 KB, 210 KB, 468 KB and 491 KB
34. Consider the intermediate code given below.
Using best fit strategy, the places allocated to the
processes are: (1) i = 1 (2) j = 1
357 KB 491 KB 468 KB 210 KB (3) t1 = 5 * i (4) t2 = t1+ j
(5) t3 = 4 * t2 (6) t4 = t3
200 400 600 500 300 250 (7) a[t4] = −1 (8) j = j + 1
(9) if j < = 5 goto (3) (10) i = i + 1
Thus, two memory slots are never used.
(11) if i < 5 goto (2)
Ans. (a)
The number of nodes and edges in the control-flow-
32. Consider a typical disk that rotates at 15000 rota- graph constructed for the above code, respectively,
tions per minute (RPM) and has a transfer rate of are

SET 2 611
(a) 5 and 7 (b) 6 and 7 L2: {anbm | m ≠ n and m, n ≥ 0}

(c) 5 and 5 (d) 7 and 8 L3: {apbqcr | p, q, r ≥ 0}
Solution: (a) L1 and L3 only (b) L2 only
Number of nodes = 6 (c) L2 and L3 only (d) L3 only
Number of edges = 7
Solution: L1 and L3 are regular since we can make
i=1 a regular expression for them.
Ans. (a)
j=1 37. Given below are some algorithms, and some algo-
rithm design paradigms.
t1 = 5 * i (1) D
ijkstra’s shortest path (i) Divide and conquer
t 2 = t1 + j (2) F
loyd-Warshall algo- (ii) Dynamic programming
t 3 = 4 * t2 rithm to compute all
pair shortest path
t 4 = t3
(3) B
inary search on a (iii) Greedy design
a[t4] = −1 sorted array
j = j+1 (4) B
acktracking search (iv) Depth-first search
on a graph
Yes (v) Breadth-first search
if j < 5
No Match the above algorithms on the left to the corre-
sponding design paradigm they follow.
i = i+1
(a) 1 − i, 2 − iii, 3 − i, 4 − v
(b) 1 − iii, 2 − iii, 3 − i, 4 − v
if i < 5 (c) 1 − iii, 2 − ii, 3 − i, 4 − iv
Yes (d) 1 − iii, 2 − ii, 3 − i, 4 − v
Ans. (b)
Solution: Dijkstra uses the Greedy approach.
35. The number of states in the minimal determinis-
All pair shortest path algorithm uses the concept
tic finite automaton corresponding to the regular
. of overlapping while computing the shortest path.
expression (0 + 1)*(10) is
Hence, it uses dynamic programming.
Solution: Binary search divides the list and then searches for an
0,1 element. Hence, it uses divide and conquer strategy.
Backtracking is used in DFS. Once we have cov-
1 0 ered one path, we need to backtrack to move on to
A B C another.
Ans. (c)
The equivalent DFA is
0 38. A Young tableau is a 2D array of integers increas-
1 ing from left to right and from top to bottom. Any
A
1 unfilled entries are marked with ∞, and hence there
B
cannot be any entry to the right of, or below a ∞.
The following Young tableau consists of unique
0 1 entries.
0
C 1 2 5 14
Ans. (3) 3 4 6 23
36. Which of the following languages is/are regular? 10 12 18 25
L1:{wxwR|w, x ∈ {a,b}* and |w|, |x| > 0}, wR is 31 ∞ ∞ ∞
the reverse of string w

When an element is removed from a Young tab- the array, and all elements greater than the pivot is
leau, other elements should be moved into its place in the right part. In addition, it moves the pivot so
so that the resulting table is still a Young tableau that the pivot is the last elements of the left part.
(unfilled entries may be filled in with a ∞). The The return value is the number of elements in the
minimum number of entries (other than 1) to be left part.
shifted, to remove 1 from the given Young tableau The following partially given function in the C pro-
is . gramming language is used to find the k th smallest
element in an array a [ ] of size n using the parti-
Solution: tion function. We assume k ≤ n.
int kth_smallest(int a[], int n, int k)
1 2 5 14 After removing ∞ 2 5 14
element 1 {
3 4 6 23 3 4 6 23 int left_end partition (a,n);
10 12 18 25 10 12 18 25 if (left_end + 1 == k)
{
31 ∞ ∞ ∞ 32 ∞ ∞ ∞
return a [left_end];
Intial state }
Shift element 2
if ((left_end + 1) > k)
with ∞
{
return kth_smallest (__________);
2 4 5 14 Shift 4 2 ∞ 5 14 }
with ∞ else
3 ∞ 6 23 3 4 6 23
return kth_smallest (__________);
10 12 18 25
10 12 18 25 }
31 ∞ ∞ ∞ 31 ∞ ∞ ∞ }
Shift 6 The missing argument lists are respectively

with ∞ (a) ( a, left _ end, k) and (a + left _ end +1, n −
left _ end −1, k − left _ end −1)
2 4 5 14 Shift 18 2 4 5 14 (b) (a, left _ end, k) and (a, n − left _ end −1,
3 6 ∞ 23 with ∞ 3 6 18 23 k − left _ end −1)
10 12 18 25 10 12 ∞ 25 (c) (a + left _ end +1, n − left end −1, k − left _
end −1) and (a, left _ end, k)
31 ∞ ∞ ∞ 31 ∞ ∞ ∞
(d) (a, n − left _ end −1, k − left _ end −1) and
Shift 25 (a, left _ end, k)
with ∞ Solution: The given code is implementation of
kth smallest element.
2 4 5 14
Replace the blanks with (a, left_end, k) and (a +
3 6 18 23 left_end + 1, n − left_end − 1, k − left_end − 1).
Final table 10 12 25 ∞ Ans. (a)
31 ∞ ∞ ∞ 40. Which one of the following hash functions on integers
will distribute keys most uniformly over 10 buckets
So, minimum number of entries to be shifted = 5. numbered 0 to 9 for i ranging from 0 to 2020?
Ans. (5) (a) h(i) = i2 mod 10
(b) h(i) = i3 mod 10
39. Suppose you are provided with the following func- (c) h(i) = (11 * i2) mod 10
tion declaration in the C programming language (d) h(i) = (12 * i) mod 10
int partition (int a[], int n); Solution: If we take the first 10 elements, number
of collisions taken by the hash function is given by
The function treats the first element of a[ ] as a
option (b), which is less when compared to others.
pivot, and rearranges the array so that all elements
less than or equal to the pivot is in the left part of Ans. (b)

SET 2 613
41. The secant method is used to find the root of an {

equation f(x) = 0. It is started from two distinct static int size =0, stkTop=0;
estimates, xa and xb for the root. It is an itera-
switch (opcode)
tive procedure involving linear interpolation to
a root. The iteration stops if f(xb) is very small {
and then xb is the solution. The procedure is given case-1: size = val; break;
below. Observe that there is an expression which is
missing and is marked by ? Which is the suitable case 0: if (stkTop < size) A (stktop++]
expression that is to be put in place of ? so that it = val; break;
follows all steps of the secant method? default: if (stktop) return A [--stkTop];
Secant }
return -1;
Initialize: xa, xb, ε, N
// ε =convergence indicator }
// N = maximum no. of iterations int main()
fb = f(xb) {
i = 0 int B[20]; A = B; stkTop = -1;
while (i < N and |fb| > ε) do stkFunc(-1, 10);
i = i + 1 // update counter
stkFunc(0, 5);
xt = ? // missing expression for
stkFunc(0, 10);
xa = xb
printf(“%d/n”, stkFunc(1, 0) +
xb = xt // intermediate value
stkfunc(1, 0);
fb = f(xb) // function value at new xb
}
end while
The value printed by the above program is .
if |fb| > ε then // loop is terminated
with i=N Solution: stkFunc(−1,0) returns size = 10. Then,
write “Non-convergence” stkFunc(0,5) returns 5. Then StrFunc(0,10) returns
else
10. StkFunc(1,0) returns 5 and then 10, so answer
is 15.
write “return xb” Ans. (15)
end if
43. Consider the sequence of machine instruction given
below:
(a) xb–(fb–f(xa))fb/(xb– xa)
(b) xa–(fa–f(xa))fa/(xb–xa) MUL R5, R0, R1
(c) xb–(xb–xa)fb/(fb–f(xa)) DIV R6, R2, R3
(d) xa–(xb–xa)fa/(fb–f(xa))
ADD R7, R5, R6
Solution: The given pseudocode implements SUB R8, R7, R4
secant method. Option (c) provides the direct In the above sequence, R0 to R8 are general pur-
formula. pose registers. In the instructions shown, the first
Secant method: xn + 1 = (f(xn−1) − fn−1(xn))/fn − fn−1 register stores the result of the operation per-
from (c), xb −((xb − xa) fb)/(fb − fa) = ( fb xa − faxb)/ formed on the second and the third registers. This
(fb − fa) sequence of instructions is to be executed in a
pipelined instruction processor with the following
Ans. (c)
4 stages (1) Instruction Fetch and Decode (IF), (2)
Operand Fetch (OF), (3) Perform Operation (PO)
42. Consider the C program below. and (4) Write back the result (WB) . The IF,OF
#include <stdio.h> and WB stages take 1 clock cycle each for any
instruction The PO stage takes 1 clock cycle for
int *A, stkTop;
ADD or SUB instruction, 3 clock cycles for MUL
int stkFunc (int opcode, int val) instruction and 5 clock cycles for DIV instruction.

The pipelined processor uses operand forwarding Solution: F(A, B, C, D) = [D′ + AB ′+ A′C + AC ′D
from the PO stage to the OF stage. The number + A′C ′D]′
of clock cycles taken for the execution of the above
sequence of instructions is . CD
AB 00 01 11 10
Solution:
00 0 0 0 0
1 2 3 4 5 6 7 8 9 10 11 12 13
01 0 0 0 0
I I P P P W
11 0 0 1 0
O O - - P P P P P W
- - - - - -
I O P W 10 0 0 0 0
I - - O - - - - P W The number of min-terms = 1.

Ans. (1)
Ans. (13) −(1/ 3)
46. Let f (x) = x and A denote the area of the
44. Consider a processor with byte-addressable region bounded by f (x) and the X-axis, when x
memory. Assume that all registers, including varies from −1 to 1. Which of the following state-
Program Counter (PC) and Program Status Word ments is/are TRUE?
(PSW), are of size 2 bytes. A stack in the main I f is continuous in [−1,1]
memory is implemented from memory location II f is not bounded in [−1,1]
(0100)16 and it grows upward. The stack pointer III A is nonzero and finite
(SP) points to the top element of the stack. The
(a) II only (b) III only
current value of SP is (016E)16. The CALL instruc-
(c) II and III only (d) I, II and III
tion is of two words, the first word is the op-code
and the second word is the starting address of Solution:
the subroutine. (one word = 2 bytes). The CALL 1 0 1
∫ f ( x ) dx = ∫ x dx + ∫x dx = 0
instruction is implemented as follows: −1 / 3 −1/ 3
Store the current value of PC in the stack −1 −1 0
Store the value of PSW register in the stack So, only statement II is true.
Load the starting address of the subroutine in PC Ans. (c)
The content of PC just before the fetch of a CALL 47. Perform the following operations on the matrix
instruction is (5FA0)16. After execution of the
CALL instruction, the value of the stack pointer is  3 4 45 
 7 9 105
(a) (016A)16 (b) (016C)16  
(c) (0170)16 (d) (0172)16 13 2 195
Solution: (i) Add the third row to the second row
Given that memory is byte addressable. (ii) Subtract the third column from the first column.
Implementation of CALL instruction is as follows: The determinant of the resultant matrix is .
••Store the current value of PC in the stack:
Solution: After applying row and column opera-
Current stack pointer +2 = (016E)16 +2
tions, the resultant matrix is:
••Store the value of PSW register in the stack:
(016E)16 +2 +2 = =(0172)16  −42 4 45 
Ans. (d) B = −280 11 300 
 
45. The number of min-terms after minimizing the −182 2 195 
following Boolean expression is —————.
|B| = 0
[D′ + AB ′ + A′C + AC ′D + A′C ′D]′ Ans. (0)

SET 2 615
48. The number of onto functions (surjective functions) (a) A multiple of 4

from set X = {1,2,3,4} to set Y = {a,b,c} is . (b) Even
(c) Odd
Solution: The number of onto functions from X
(d) Congruent to 0 mod 4, or, 1 mod 4
to Y is
n −1
∑ (−1) knck (n - k)m, where½X½= m and½Y½= n
Solution: In a self-complementary graph, the
number of vertices is equal to exactly half of
k =0
number of edges in a complete graph, that is,
Here, X = {1, 2, 3, 4}, Y = {a, b, c}, m = 4 and n(n−1)/4. Hence, answer is (d).
n=3 Ans. (d)
Applying the formula we get, the number of onto
functions = 36 52. In a connected graph, a bridge is an edge whose
Ans. (36) removal disconnects a graph. Which one of the fol-
lowing statements is true?
49. Let X and Y denote the sets containing 2 and 20
(a) A tree has no bridges
distinct objects respectively and F denote the set
(b) A bridge cannot be part of a simple cycle
(c) Every edge of a clique with size ≥ 3 is a bridge
of all possible functions defined from X to Y. Let
f be randomly chosen from F. The probability of f
(A clique is any compete subgraph of a
being one-to-one is .
graph)
(d) A graph with bridges cannot have a cycle
Solution: |X | = 2 and |Y | = 20
Total number of functions from Solution:
Option (a): False as in tree, every edge acts as a bridge.
X to Y = 202 = 400
Option (b): True
Total number of one to one functions Option (c): False as every edge of complete graph
20 is not bridge. Clique is also a complete graph.
= P2 = 380
Option (d): False as the graph is having a cycle
380 and bridge.
Probability = = 0.95
400
Ans. (0.95)
50. Consider the alphabet Σ = {0.1}, the null/empty

string l and the sets of strings X0, X1, and X2 Ans. (b)
generated by the corresponding non-terminals of
a regular grammar X0, X1, and X2 are related as 53. Which one of the following well formed formulae is
follows. tautology?
X0 = 1 X1 (a) ∀x ∃y R(x, y) ↔ ∃y ∀x R(x, y)
X1 = 0 X1 + 1 X2 (b) (∀x [∃y R(x,y) → S(x,y)]) → ∀x∃yS(x,y)
X2 = 0 X1 + {l} (c) [∀x∃y (P(x,y) → R(x,y)] ↔ [∀x∃y(¬P(x,y)
∨R(x,y)]
Which one of the following choices precisely repre-
sents the strings in X0? (d) ∀x∀y P(x,y) → ∀x∀yP(y,x)
(a) 10(0*+(10)*)1
Solution: Tautology is that which is always true
(b) 10(0*+(10)*)*1
in all cases. Check the expressions for this.
(c) 1(0 +10)*1
(d) 10(0 +10*)*1+110(0 +10)*1 Ans. (c)
Solution: Strings formed by the given equations 54. Which one of the following assertions concerning
are represented by regular expression: 1(0 +10)*1. code inspection and code walkthrough is true?
Ans. (c)
(a) C
ode inspection is carried out once the code has
51. A graph is self-complementary if it is isomorphic to been unit tested
its complement. For all self-complementary graphs (b) Code inspection and code walkthrough are
on n vertices, n is synonyms

(c) Adherence to coding standards is checked adders and one OR gate. The propagation delay of
during code inspection an XOR gate is twice that of an AND/OR gate.
(d) Code walkthrough is usually carried out by an The propagation delay of an AND/OR gate is 1.2
independent test team microseconds. A 4-bit ripple-carry binary adder is
implemented by using four full adders. The total
Solution: Code inspection generally checks that propagation time of this 4-bit binary adder in
coding standards are followed or not. microseconds is .
Ans. (c)
Solution: Total delay = 4 × 4.8 = 19.2 ms
55. A half adder is implemented with XOR and AND
gates. A full adder is implemented with two half Ans. (19.2)

SOLVED GATE(CS) 2015
SET 3
1. Consider the following C program segment. 2. Suppose U is the power set of the set S = {1, 2, 3, 4,
5, 6}. For any T ∈ U, let |T| denote the number of
#include <stdio.h>
elements in T and T1 denote the complement of T.
int main() For any T, R ∈ U, let T \R be the set of all
{ elements in T which are not in R. Which one of the
following is true?
char sl[7] = “1234”, *p;
p = sl + 2; (a) ∀X ∈ U (|X| = |X ′|)
(b) ∃X ∈ U ∃ Y ∈ U (|X | = 2, |Y | = 5 and X ∩ Y
*p = ‘0’;
= 0)
printf(“%s”, s1); (c) ∀X ∈ U ∀ Y ∈ U (|X | = 2, |Y | = 3 and X \Y
} = 0)
(d) ∀X ∈ U ∀ Y ∈ U (X\Y = Y ′\X ′)
What will be printed by the program?
(a) 12 (b) 120400 Solution:
(c) 1204 (d) 1034 S = {1, 2, 3, 4, 5, 6}
Let X = {1,2}
X ′ = S − X = {3, 4, 5, 6}
Solution: The given character array
representation:
|X | ≠ |X ′|
1 2 3 4 \0
100 101 102 103 Option (a) is false.
Let X = {1, 2, 4, 5, 6} and Y = {1, 2, 3, 4, 5}
s1: Base address = 100 X ∩ Y = {1, 2, 4, 5} ≠ ∅
p = 100 + 2 = 102
Option (b) is incorrect.
*p = 102 = 0
Let X = {1, 2} and Y = {2, 4, 5}
1 2 0 4 \0 X/Y = {1} ≠ ∅
100 101 102 103 Option (c) is incorrect.
Output will be: 1204 So, option (d) is the correct answer.
Ans. (c) Ans. (d)

3. Consider the relation X(P, Q, R, S, T, U) with the (a) Only S1 is true

following set of functional dependencies (b) Only S2 is true
F={
(c) Both S1 and S2 are true
(d) Neither S1 nor S2 are true
{P, R}→ {S, T},
Solution:
{P, S, U}→ {Q, R} X#Y = X ′ + Y ′
} S1: (P#Q)#R = P#(Q#R)
Which of the following is the trivial functional LHS
dependency in F+, where F+ is closure of F? (P ′ + Q ′)′ + R ′
(a) {P, R} → {S, T} (b) {P, R} → {R, T} PQ + R ′
(c) {P, S} → {S} (d) {P, S, U} → {Q} RHS
P ′ + (Q ′ + R ′)′
Solution: Trivial function dependency X → Y
holds if Y ⊆ X. P′ + QR
From the given set of functional dependencies, only LHS ≠ RHS
{P, S} → {S} holds true. S1 is False.
Because S ⊆ PS. S2: Q#R = R#Q
Ans. (c) Q′ + R′ = R′ + Q′
4. The maximum number of processes that can be in S2 is True.
Ready state for a computer system with n CPUs is Ans. (b)
2
(a) n (b) n 7. Consider a software project with the following
(c) 2n (d) independent of n information domain characteristics for calculation
of function point metric.
Solution: Any number of processes can wait in
Ready queue. It does not depend on number of Number of external inputs (I) = 30
CPUs available. Number of external output (O) = 60
Ans. (d) Number of external inquiries (E) = 23
5. Among simple LR (SLR), canonical LR, and look- Number of files (F) = 08
ahead LR (LALR), which of the following pairs Number of external interfaces (N) = 02
identify the method that is very easy to implement
It is given that the complexity weighting factors for
and the method that is the most powerful, in that
I, O, E, F and N are 4, 5, 4, 10 and 7, respectively.
order
It is also given that, out of fourteen value adjust-
(a) SLR, LALR ment factors that influence the development effort,
(b) Canonical LR, LALR four factors are not applicable, each of the other
(c) LSR, canonical LR four factors have value 3, and each of the remain-
(d) LALR, canonical LR ing factors have value 4. The computed value of
function point metric is .
Solution: CLR is the most powerful parsing.
Solution:
Simple LR is easy to implement as there is no over-
Number of external inputs (I) = 30 × 4 = 120
head of look ahead symbols.
Ans. (c) Number of external outputs (O) = 60 × 5 = 300
6. Let # be a binary operator defined as Number of external inquiries (E) = 23 × 4 = 92
Number of files (F) = 08 × 10 = 80
X # Y = X ′+Y ′ where X and Y are Boolean
variables. Number of external interfaces (N) = 02 × 7 = 14
Consider the following two statements Total unadjusted function points = 606
S1: (P # Q) # R = P # (Q # R) Ci = 4 × 3 + 6 × 4 = 36
S2: Q # R = R # Q VAF = 0.65 + [(Ci)/100] = 0.65 + 0.36 = 1.01
Which of the following is/are true for the Boolean FP = UFP × VAF ⇒ FP = 606 × 1.01 = 612.06
variables P, Q and R? Ans. (612.06)

SET 3 619
8. In a web server, ten WebPages are stored with the (a) Only I
URLs of the form (b) Only II
(c) I or III or IV but not II
http://www.yourname.com/var.html; where, var is
(d) II or III or IV but not I
a different number from 1 to 10 for each Webpage.
Suppose, the client stores the Webpage with var
Solution:
= 1 (say W1) in local machine, edits and then
tests. Rest of the WebPages remains on the web
n2 (n + 1)2
server. W1 contains several relative URLs of the X = 13 + 23 + 33 + 43 + L + n3 =
form “var.html” referring to the other WebPages. 4
Which one of the following statements needs to
be added in W1, so that all the relative URLs This is equivalent to Q (n4), O(n5) and W(n3).
in W1 refer to the appropriate WebPages on the Ans. (c)
web server?
11. Consider a binary tree T that has 200 leaf nodes.
(a) <a href: “http://www.yourname.com/”, href: Then, the number of nodes in T that have exactly
“…var.html”> two children are .
(b) <base href: “http://www.yourname.com/”>
(c) <a href: “http://www.yourname.com/”> Solution:
(d) <base href: “http://www.yourname.com/”, Leaf (p) = 200
range “…var.html”> Total nodes = 2p − 1 = 399
Solution: Base tag with href is used: < base href :“ Nodes with degree exactly two children = 399 −
http : // www.yourname.com / ”, range :“...var. 200 = 199
html” > Ans. (199)
Ans. (b)
12. Given a hash table T with 25 slots that stores 2000
9. Consider the following statements elements, the load factor a for T is .
I. TCP connections are full duplex Solution:

II. TCP has no option for selective acknowledgement Size
Load factor (a) =
III. TCP connections are message streams Complexity
(a) Only I is correct Size: Number of elements = 2000
(b) Only I and III are correct Capacity: Number of slots = 25
(c) Only II and III are correct
2000
(d) All of I, II, and III are correct a= = 80
25
Ans. (80)
Solution:
I. True, TCP connection is full duplex connection  1 −1 2 
because sender and receiver can exchange data 13. In the given matrix 0 1 0  , one of the Eigen
 
1 2 1 
in the same connection.
II. False, there is no selective acknowledgement in values is 1. The Eigen vectors corresponding to the
TCP, concept of cumulative acknowledgement Eigen value 1 are
is used. (a) {a(4, 2, 1)|a ≠ 0, a ∈ R}
III. False, Data is sent in byte streams. (b) {a(-4, 2, 1)|a ≠ 0, a ∈ R}
(c) {a(2, 0, 1)|a ≠ 0, a ∈ R}
Ans. (a) (d) {a(-2, 0, 1)|a ≠ 0, a ∈ R}
n
10. Consider the equality ∑ i3 = X and the following Solution: Let X be the eigen vector corresponding
choices for X: i=0 to the eigen value l = 1
I. Q(n )
4
II. Q (n5) Apply AX = lX
III. O(n5) IV. Ω(n3) Thus, we get Eigen vectors as {a (−4,2,1)|a ≠ 0,
The equality above remains correct if X is a ∈ R}
replaced by Ans. (b)

(
14. The value of lim x →∞ 1 + x2 e−x is ) .
Option (c): If person is of Type 2, then he
is telling lie that toss is head. Option (c) is
1 incorrect.
(a) 0 (b)
2 Option (d): If person is of Type 1, then he is
(c) 1 (d) ∞ telling the truth that toss is head. Hence, this is
incorrect.
Solution: Ans. (a)
( )
lim x→∞ 1 + x2 e−x = (∞)0 = 1 17. While inserting the elements 71, 65, 84, 69, 67,
83 in an empty binary search tree (BST) in
the sequence shown, the element in the lowest
Ans. (c)
level is
15. The number of 4 digit numbers having their digits
(a) 65 (b) 67
in non-decreasing order (from left to right) con-
(c) 69 (d) 83
structed by using the digits belonging to the set
{1, 2, 3} is .
Solution: 67 is at the lowest level.
Solution: 4 digit numbers having their digits in 71

non-decreasing order (from left to right) from set
{1, 2, 3}
Starting with 1: 1111, 1112, 1113, 1122, 1123, 1133, 65 84
1222, 1223, 1233, 1333 = 10
Starting with 2: 2222, 2223, 2233, 2333 = 4
Starting with 3: 3333 = 1
69 83
Total = 10 + 4 + 1 =15
Ans. (15) 67
16. In a room there are only two types of people, Ans. (b)
namely Type 1 and Type 2. Type 1 people always
18. The result evaluating the postfix expression
tell the truth and Type 2 people always lie. You
10 5 + 60 6 / * 8 - is
give a fair coin to a person in that room, without
knowing which type he is from and tell him to toss (a) 284 (b) 213
it and hide the result from you till you ask for it. (c) 142 (d) 71
Upon asking, the person replies the following
Solution:
“The result of the toss is head if and only if I am \
telling the truth.”
Which of the following options is correct? 6
(a) The result is head 5 60 60 10 8
(b) The result is tail 10 10 15 15 15 15 150 150 142
(c) If the person is of Type 2, then the result is tail
(d) If the person is of Type 1, then the result is tail 10 5 + 60 6 / ∗ 8 −
\
Ans. (c)
Solution: Type 1 people always tell the truth.
Type 2 people always lie. 19. Consider the following relation
Statement: “The result of the toss is head if and Cinema (theater, address, capacity)
only if I am telling the truth.”
Which of the following options will be needed at
Option (a): This statement extracts that the the end of the SQL query?
person is of Type 1 and the result is head.
SELECT P1. address
Option (b): We cannot draw conclusion about
the Type of person. So, it is incorrect. FROM Cinema P1

SET 3 621
Such that it always finds the addresses of theaters while(varQ == true)

with maximum capacity? {
/* Critical Section */
(a) WHERE P1. capacity > = All (select P2.
varP = false;
capacity from Cinema P2)
(b) WHERE P1. capacity > = Any (select P2. }
capacity from Cinema P2) }
(c) WHERE P1. capacity > All (select max (P2. /* other code for process X */
capacity) form Cinema P2)
(d) WHERE P1. capacity > Any (select max (P2. Process Y
capacity) from Cinema P2) /* other code for process Y */
while (true)
Solution: Inner query selects the capacity of {
Theaters, > = All compares it with the all the
varQ = true;
values. So, option (a) will find address of theaters
with maximum capacity. while(varP == true)
Ans. (a) {
20. Consider the following array of elements /* Critical Section */
varQ = false;
<89, 19, 50, 17, 12, 15, 2, 5, 7, 11, 6, 9, 100>
}
The minimum number of interchanges needed to }
convert it into a max-heap is
/* other code for process Y */
(a) 4 (b) 5
(c) 2 (d) 3 Here, varP and varQ are shared variables and both
are initialized to false. Which one of the following
Solution: statements is true?
89 89 (a) The proposed solution prevents deadlock but

fails to guarantee mutual exclusion
19 50 Swap 19 50 (b) The proposed solution guarantees mutual
exclusion but fails to prevent deadlock
17 12 15 2 100 & 15 17 12 100 2 (c) The proposed solution guarantees mutual exclu-
sion and prevents deadlock
5 7 11 6 9 100 5 7 11 6 9 15 (d) The proposed solution fails to prevent deadlock
and fails to guarantee mutual exclusion
Swap 100 & 50
Solution: Proposed solution does not ensure
100 89 mutual exclusion, because both varP and varQ
can be true at the same time and enter to the
19 89 19 100
critical section. But there is no deadlock state in
Swap
17 12 50 20 17 12 50 2 this code, because X and Y do not lock required
89 & 100 variables.
5 7 11 6 9 15 5 7 11 6 9 15 Ans. (a)
Ans. (d)
22. Let L be the language represented by the regu-
21. Two processes X and Y need to access a critical lar expression ∑*0011∑* where ∑ = {0, 1}. What
section. Consider the following synchronization is the minimum number of states in a DFA that
construct used by both the processes recognizes L (complement of L)?
Process X (a) 4 (b) 5
(c) 6 (d) 8
/* other code for process X */
while (true) Solution: 5 (n + 1) states are sufficient to repre-
{ sent 0011.
varP = true; Ans. (b )

23. Consider a software program that is artificially Q. No. 26 — 55 Carry Two Marks Each
seeded with 100 faults. While testing this program,
159 faults are detected, out of which 75 faults are
from those artificially seeded faults. Assuming that 26. The velocity v (in kilometer/minute) of a motor-
both real and seeded faults are of same nature and bike which starts from rest, is given at fixed inter-
have same distribution, the estimated number of vals of time t(in minutes) as follows:
undetected real faults is .
t 2 4 6 8 10 12 14 16 18 20
Solution: v 10 18 25 29 32 20 11 5 2 0
N: Non-seeded errors = 159 − 75 = 84 The approximate distance (in kilometers) rounded

S: Seeded errors =100 to two places of decimals covered in 20 minutes
s: Seeded errors detected = 75 using Simpson’s 1/3rd rule is .
Solution: Simpson’s 1/3rd rule:

N (S − s) 84(100 − 75)
= = 28 H=2
s 75
2
Ans. (28) S= = 309.33
3[(0+0)+2(18+29+20+5)
24. Consider a machine with a byte addressable main +4(10+25+32+11+2)]
memory of 220 bytes, block size of 16 bytes and a
direct mapped cache having 212 cache lines. Let Ans. (309.33)
the addresses of two consecutive bytes in main 27. Assume that a mergesort algorithm in the worst
memory be (E201F)16 and (E2020)16. What are case takes 30 seconds for an input of size 64.
the tag and cache line address (in hex) for main Which of the following most closely approximates
memory address (E201F)16? the maximum input size of a problem that can be
solved in 6 minutes?
(a) E, 201 (b) F, 201
(c) E, E20 (d) 2, 01F (a) 256 (b) 512
(c) 1024 (d) 2048
Solution: Total bits = 20
Solution: Merge sort worst case time complexity
Word = O(nlog n)
4 12 4 For input of size 64, it takes 30 seconds

n log n = 30 for n = 64
Ans. (a) 64 log 64 = 30 seconds
25. Consider a CSMA/CD network that transmits For 6 minutes, factor is 12. So, multiply by 12 on
data at a rate of 100 Mbps (108 bits per second) both sides
over a 1 km (kilometer) cable with no repeaters. If 12 × (64 log 64) = 12 × 30 seconds
the minimum frame size required for this network 512 log 512 generates the factor of 12
is 1250 bytes, what is the signal speed (km/sec) in
So, maximum 512 size array can be sorted.
the cable?
Ans. (b)
(a) 8000 (b) 10000 28. Consider the following recursive C function.
(c) 16000 (d) 20000
void get(int n)
Solution: Frame size = 1250 bytes, Bandwidth = {

100 Mbps, Distance = 1 km, Signal Speed = ? if (n<1)return;
In CSMA/CD, if (n<1) return;
get (n–1);
L=
2dB
Þ V = 2 dB/L = 20,000 km/sec get (n–3);
V
printf(“%d”, n);
Ans. (d) }

SET 3 623
If get (6) function is being called in main() then (S2) F = ∑ (0, 1, 2, 3, 7)

how many times will the get() function be invoked (S3) F = ∏ (4, 5, 6)
before returning to the main()?
(S4) F = ∏ (0, 1, 2, 3, 7)
(a) 15 (b) 25
(c) 35 (d) 45 Which of the following is true?
(a) (S1)−False, (S2)−True, (S3)−True, (S4)−False
Solution:
(b) (S1)−True, (S2)−False, (S3)−False, (S4)−True
get(6) (c) (S1)−False, (S2)−False (S3)−True, (S4)−True
(d) (S1)−True, (S2)−True, (S3)−False, (S4)−False
Solution:
g(5) g(3)
F = P′ + QR = P′(Q + Q′) (R + R′) + (P+P′)QR
F = P′Q + P′Q′(R + R′) + PQR + P′QR
g(4) g(2) g(2) g(10)
F = P′QR + P′QR′ + P′Q′ R+ P′Q′R′+ PQR +
P′QR
g(3) g(1) g(1) g(−1) g(1) g(−1)
P QR
00 01 11 10
g(2) g(0) g(0) g(−2) g(0) g(−2)
0 1 1 1 1
1 0 0 1 0
g(1) g(−1)
F = ∑(0, 1, 2, 3, 7)
g(0) g(−2)
= ∏ (4, 5, 6)
Ans. (b)
S2 and S3 are true.
29. Consider a B+ tree in which the search key is Ans. (a)
12 bytes long, block size is 1024 bytes, record
pointer is 10 bytes long and block pointer is 8 31. Language L1 is polynomial time reducible to lan-
bytes long. The maximum number of keys that guage L2. Language L3 is polynomial time reducible
can be accommodated in each non-leaf node of to L2, which in turn is polynomial time reducible to
the tree is . language L4.
Solution: In B+ trees, if order of tree is p then for Which of the following is/are true?
internal nodes: I. if L4 ∈ P, then L2 ∈ P
p(child pointer) + (p − 1)(Key value) ≤ Block size II. if L1 ∈ P or L3 ∈ P, then L2 ∈ P
III. L1 ∈ P, if and only if L3 ∈ P
p(8) + (p − 1)12 ≤ 1024
IV. if L4 ∈ P, then L1 ∈ P and L3 ∈ P
20p − 12 ≤ 1024
(a) II only (b) III only
p ≤ 51 ⇒ p = 51 (c) I and IV only (d) I only
Solution: Given, L1 ≤P L2
So, for 51 order tree, we can save maximum
50 keys.
L3 ≤P L2
Ans. (50)
L2 ≤P L4
30. Given the function F = P′ + QR, where F is a If L4 is polynomial, then it implies that L2 is also
function in three Boolean variables P, Q and R and polynomial. Similarly, if L2 is polynomial this also
P′ = !P, consider the following statements. implies that L1 and L3 are also polynomial.
(S1) F = ∑ (4, 5, 6) Ans. (c)

32. Consider the following C program. p is a pointer to array. 1st element of p points to 10
#include <stdio.h> 2nd element points to 40
int f1(void); 3rd element points to 50
int f2(void); 4th element points to 20
int f3(void); 5th element points to 30
int x = 10; Thus, implementing the code gives us 140.
int main() Ans. (140)
{
34. Which of the following languages are context-free?
int x = 1;
L1 = {ambnanbm|m, n ≥ 1}
x + = f1() + f2() + f3() + f2();
printf (“%d”, x); L2 = {ambnambn|m, n ≥ 1}
return 0; L3 = {ambn|m = 2n + 1}
} (a) L1 and L2 only (b) L1 and L3 only
(c) L2 and L3 only (d) L3 only
int f1() {int x = 25; x++; return x;}
int f2() {static int x = 50; x++; return Solution: For identifying the context-free lan-
x;} guage, check whether the stack implementation is
int f3() {x *= 10; return x}; possible or not.
The output of the program is . L1 = {ambnanbm|m, n ≥ 1}
Solution: Here when we put m number of a’s and n number

of b’s in stack, top of stack will contain n number
x = x + f 1() + f 2() + f 3() + f 2()
of b’s. Thus, when n number of a’s comes in the
After evaluating the functions through the given string then pop b’s and when m number of b’s
code, we get comes then pop a’s from stack. Thus, it is possible
f1() ⇒ 26 to count the a’s and b’s properly. Hence, this is a
context-free language.
f2() ⇒ 51
Similarly, check for others. Thus, L1 and L3 are
f3() ⇒ 100 context free.
f2() ⇒ 52 [Static variable is there in f2(). Thus, its Ans. (b)
value is initialized only once and the value com-
puted in 1st call of f2() is retained in the next call 35. Consider the following policies for preventing
of function] deadlock in a system with mutually exclusive
Ans. (230) resources.
33. Consider the following C program. I. Processes should acquire all their resources at
the beginning of execution. If any resource is
#include <stdio.h> not available, all resources acquired so far are
int main() released.
{ II. The resources are numbered uniquely, and pro-
static int a[] = {10, 20, 30, 40, 50}; cesses are allowed to request for resources only
in increasing resource numbers
static int *p[] = {a, a+3, a+4, a+1,
III. The resources are numbered uniquely, and pro-
a+2};
cesses are allowed to request for resources only
int **ptr = p; in decreasing resource numbers
ptr++; IV. The resources are numbered uniquely. A pro-
printf(“%d%d”, ptr-p, **ptr); cess is allowed to request only for a resource
} with resource number larger than its currently
held resources
The output of the program is .
Which of the above policies can be used for pre-
Solution: a[] = {10,20,30,40,50} venting deadlock?

SET 3 625
(a) Any one of I and III but not II or IV 38. Consider the following reservation table for a pipe-
(b) Any one of I, III, and IV but not II line having three stages S1, S2, and S3.
(c) Any one of II, and III but not I or IV
(d) Any one of I, II, III, and IV Time→
Solution: All the four methods can be used to pre- 1 2 3 4 5
vent deadlock in the system.
S1 X X
Ans. (d)
S1 X X
36. In the network 200.10.11.144/27, the fourth octet
(in decimal) of the last IP address of the network S3 X
which can be assigned to a host is .
The minimum average latency (MAL) is .
Solution:
IP address: 200.20.11.144 => 200.20.11.10010000 Solution: For S1 latency = 0 + 4
Subnet mask address: 225.225.225.11100000
For S2 latency = 1 + 3
Looking at the subnet address, we get to know that
For S3 latency = 2
1st three bits gives the subnet number and remain-
ing are for host. Average latency = 10/3 = 3.33
Thus, 100 10000 Ans. (3.33)
Last host bits will be: 100 11110 ⇒ 158 39. Consider the following code sequence having five
Ans. (158) instructions I1 to I5. Each of these instructions has
37. Consider a network connecting two systems located the following format.
8000 kilometers apart. The bandwidth of the net- OP Ri, Rj, Rk
work is 500 × 106 bits per second. The propagation where operation Op is performed on contents of
speed of the media is 4 × 106 meters per second. It registers Rj and Rk and the result is stored in
is needed to design a Go Back − N sliding window register Ri.
protocol for this network. The average packet size
is 107 bits. The network is to be used to its full I1: ADD R1, R2, R3
capacity. Assume that processing delays at nodes I2: MUL R7., R1, R3
are negligible. Then, the minimum size in bits of
the sequence number field has to be . I3: SUB R4, R1, R5
I4: ADD R3, R2, R4
Solution:
I5: MUL R7, R8, R9
Distance = 8000 km = 8 × 106 m
Speed = 4 × 106 m/s
Consider the following three statements.
Transmission time = Distance/Speed = 2 s S1:There is an anti-dependence between instruc-

tion I2 and I5
Bandwidth = 500 × 10 bps 6
S2:There is an anti-dependence between instruc-
Packet size = 107 bits tions I2 and I4
500 × 106 bits ⇒1 s S3:Within an instruction pipeline an anti-depen-
107 bits ⇒ 107 / 500 × 106 = 1/50 s dence always creates one or more stalls
Propagation time = 1/50 s ⇒ RTT = 2/50 = 1/25 s Which one of above statements is/are correct?
Full capacity network utilization is done. Therefore, (a) Only S1 is true
efficiency is 100%. (b) Only S2 is true
w (c) Only S1 and S3 are true
h= = 1 ⇒ w = 1 + 2a (d) Only S2 and S3 are true
(1 + 2a)
2 Solution: There occurs a name dependency here
a= = 100
0.02 (write after read). This is anti-dependency and it
2n = 1 + 200 = 201 ⇒ n = 8 always causes a stall of one or more cycles.
Ans. (8) Ans. (b)

40. Consider the following two C code segments. Y and

Time instance Transaction -id
X are one- and two-dimensional arrays of size n
and n × n, respectively, where 2 ≤ n ≤ 10. Assume T1 T2
that in both code segments, elements of Y are ini- 1 read(A)
tialized to 0 and each element X[i] [j] of array X is
initialized to i+j. Further assume that when stored 2 write(A)
in main memory all elements of X are in same main 3 recd(C)
memory page frame.
4 write(C)
Code segment 1:
5 read(B)
//initialize elements of Y to 0
6 write(B)
//initialize elements X[i] [j] of X to i+j
7 read(A)
for (i = 0; i < n; i++)
8 commit
Y[i] += X[0] [i];
9 read(B)
Code segment 2:
//initialize elements of Y to 0 Schedule S
Suppose that the transaction T1 fails immediately
//initialize elements X[i] [j] of X to i+j
after time instance 9. Which one of the following
for (i = 0; i < n; i++) statements is correct?
Y[i] += X[i][0]; (a) T2 must be aborted and then both T1 and
Which of the following statements is/are correct? T2 must be re-started to ensure transaction
atomicity
S1: Final contents of array Y will be same in both (b) Schedule S is non-recoverable and cannot
code segments ensure transaction atomicity
S2: Elements of array X accessed inside the for (c) Only T2 must be aborted and then restarted to
loop shown in code segment 1 are contiguous in ensure transaction atomicity
main memory (d) Schedule S is recoverable and can ensure atomi-
city and nothing else needs to be done
S3: Elements of array X accessed inside the for
loop shown in code segment 2 are contiguous in Solution: T2 transaction reads the value written
main memory by T1 transaction. T2 is committed before T1 and
(a) Only S2 is correct T1 fails. Thus, T2 has read that value of A which
(b) Only S3 is correct no more exists in the database. Hence, the sched-
(c) Only S1 and S2 are correct ule S is non-recoverable schedule. Also, it doesn’t
(d) Only S1 and S3 are correct ensure atomicity.
Ans. (b)
Solution: Y is initialized with all zeros. X is ini- 42. If the following system has non-trivial solution
tialized with sum of its indices. px + qy + rz = 0
Y[i] + = x[0] [i] and Y[i] + = x[i] [0] will con- qx + ry + pz = 0
tain same values. Statement S1 is true. S2 is true rx + py + qz = 0
because array is stored as x[0][0], x[0][1], x[0]
then which one of the following options is TRUE?
[2]…x[1][0], x[1][2] and so on.
(a) p − q + r = 0 or p = q = −r
Ans. (c)
(b) p + q − r = 0 or p = −q = r
41. Consider the following partial Schedule S involving (c) p + q + r = 0 or p = q = r
two transactions T1 and T2. Only the read and (d) p − q + r = 0 or p = −q = −r
the write operations have been shown. The read
Solution: For non-trivial solution, |A| = 0. Applying
operation on data item P is denoted by read (P)
this on the given matrix, we get answer (c).
and the write operation on data item P is denoted
by write (P). Ans. (c)

SET 3 627
43. Consider the following C program: Replacing x by 1/x, we get

# include<stdio.h> af (1/x) + bf (x) = x – 25 (2)
int main( )
{ Multiply Eq. (1) by a and Eq. (2) by b, and then
subtract them, to get
int i, j, k = 0;
j = 2 * 3 / 4 + 2.0 / 5 + 8 / 5; f (x) = [1/(a2 − b2) ][a/x − bx + 25(b − a)]
k –= ––j; 2  
∫ f(x) dx = (a2 − b2 )  ×
for (i = 0; i < 5: i++)
1
{ 1  
switch (i + k)   x2 
2
{ a(ln x − 25x) − b  − 25x 

  2  1
case 1:
case 2: printf (“ n%d”, i+k); Solving this equation, we get the answer (c).
case 3: printf (“n%d”, i+k); Ans. (a)
default: printf (“n%d”, i+k);
45. Let G be a connected undirected graph of 100
} vertices and 300 edges. The weight of a minimum
} spanning tree of G is 500. When the weight of each
return 0; edge of G is increased by five, the weight of a mini-
mum spanning tree becomes .
}
The number of times printf statement is executed Solution: In a spanning tree containing 100 verti-
is . ces, number of edges will be 99.
Solution: j and k will be evaluated to be 2 and Therefore, weight is increased by = 500 + 5 × 99
−1, respectively. After getting in the loop, and = 995
implementing every iteration, the printf is called Ans. (995)
10 times.
46. Two hosts are connected via a packet switch with
Ans. (10) 107 bits per second links. Each link has a propaga-
1 1
tion delay of 20 microseconds. The switch begins
44. If for non-zero x, af (x) + bf   = − 25 where forwarding a packet 35 microseconds after it receives
x x the same. If 10000 bits of data are to be transmitted
2
a ≠ b then ∫ f (x) dx is
between the two hosts using a packet size of 5000 bits,
the time elapsed between the transmission of the first
1
bit of data and the reception of the last bit of the data
in microseconds is .
1 47b 
(a) a(ln 2 − 25) + 2 
a −b
2 2
Solution: Transmission time of 5000 bits = 500
 47b 
1 microseconds
(b) a(2 ln 2 − 25) − 2 
a2 − b2  Total transmission time of first bit = 500 + 20 +
35 + 20 = 575 microseconds
1 47b 
 a(2 ln 2 − 25) +
2 
(c)
a −b 
2 2 Total transmission time of last bit = 500 + 500 =
1000 microseconds
1 47b 
(d) a(ln 2 − 25) − 2  Total time = 1000 + 575 = 1575 microseconds
a −b
2 2
Ans. (1575)
Solution:
47. For the processes listed in the following table,
1 which of the following scheduling schemes will give
af (x) + bf (1/x) = (1)
x − 25 the lowest average turnaround time?

while((i< value) && (result < = maxint))

Process Arrival Time Processing Time
}
A 0 3 i = i +1;
B 1 6 result = result + 1;
C 4 4 }
if (result < = maxint)
D 6 2
{
printf(result);
(a) First Come First Serve
(b) Non-preemptive Shortest Job First }
(c) Shortest Remaining Time else
(d) Round Robin with Quantum value two {
printf(“large”);
Solution:
}
Process Arrival Time Processing Time printf(“end of program”);
}
A 0 3
B 1 6 Control Flow Diagram of Program — Y:
C 4 4
D 6 2
Shortest remaining time first:

A B C D B
0 3 4 8 10 15
Turnaround time = Completion time - Arrival
time
A⇒3-0=3
B ⇒ 15 - 1 = 14
C⇒8-4=4
D ⇒ 10 - 6 = 4
3 + 14 + 4 + 4
Average turnaround time = = 6.25
4 Control Flow Diagram of Program — Z:
Similarly checking for other cases, we will find that
average turnaround time is minimum in SRTF. Control flow diagram of
Ans. (c) Program–X
48. Consider three software items: Program—X, Control
Flow Diagram of Program—Y and Control Flow
Diagram of Program—Z as shown below Control flow diagram of
Program–Y
Program X:
sumcal(int maxint, int value) The values of McCabe’s Cyclomatic complexity of
{ Program—X, Program—Y, and Program—Z, respec-
int result =0, i = 0; tively, are
if(value < 0) (a) 4, 4, 7
{ (b) 3, 4, 7
value = -value; (c) 4, 4, 8
} (d) 4, 3, 8

SET 3 629
Solution: From Program—X: Pr[Xi = 0] = Pr[Xi = 1] = 1/2 for i = 1, 2, 3.

Number of conditions and control statements = 3 Define another random variable Y = X1 X2 ⊕ X3,
where ⊕ denotes XOR. Then Pr[Y = 0|X3 = 0] =
Complexity = 3 + 1 = 4 .
Control Flow Diagram of Program—Y:
Complexity = Edges − Number of vertices + 2 Solution:
= 10 − 8 + 2 = 4 X3 = 0 and Y = X1 X2 XOR X3 => X1 X2 = 0
Control Flow Diagram of Program—Z: X1 = 0 and X2 = 0, X1 = 1 and X2 = 0, X1 = 0

and X2 = 1
Complexity = X + Y − 1 = 4 + 4 − 1 = 7
P[Y=0/X3 = 0] = P[(Y=0)∩( X3 = 0)]/P[X3 = 0]
Ans. (a)
= (1/2)3 × 3/(1/2) = 3/4 = 0.75
49. Consider the equation (43)x = (y3)8 where x and y
Ans. (0.75)
are unknown. The number of possible Solutions is
. 52. The total number of prime implicants of the
function f (w, x, y, z) = ∑(0, 2, 4, 5, 6, 10)
Solution: is .
(43)x = (y3)8
⇒ 4x + 3 = 8y + 3 Solution:
⇒ x = 2y wx yz
00 01 11 10
Moreover, x ≥ 5 and y ≤ 7
Hence, possible values of x and y are: (14,7), (12,6), 00 ∗1 0 0 1
(10,5), (8,4), (6,3)
Ans. (5) 01 1 1∗ 0 1∗
50. Let R be a relation on the set of ordered pairs of 11 0 0 0 0

positive integers such that ((p,q), (r,s)) ∈ R if and
only if p−s = q−r. Which one of the following is 10 0 0 0 1∗
true about R?
(a) Both reflexive and symmetric
(b) Reflexive but not symmetric Ans. (3)
(c) Not reflexive but symmetric
(d) Neither reflexive nor symmetric 53. Suppose c = 〈c[0], …., c[k-]〉 is an array of length k,
where all the entries are from the set {0, 1}. For
Solution: any positive integers a and n, consider the follow-
((p,q) , (r,s)) Î R ing pseudocode.
R⇒p-s=q-r DOSOMETHING (c, a, n)

z ← 1
As we know, for reflexive we need (p,q) and (q,p)
for i ← 0 to k – 1
in relation R.
do z ← z2 mod n
But p − q ≠ q − p, for example, (4, 2) and (2, 4)
if c[i] = 1
⇒ 4 − 2 = 2 and 2 − 4 = −2
then z ← (z × a) mod n
Hence, relation is not reflexive.
return z
Similarly, for symmetric if (p,q) R (q,s) ⇒ p − s
=q−r If k = 4, c = 〈1, 0, 1, 1〉, a = 2 and n = 8, then the
.
Also, r − q = s − p ⇒ (r,s) R (p,q) output of DOSOMETHING (c, a, n) is
Hence, it is symmetric. Solution: k = 4, c = <1 0 1 1>, a = 2 and n = 8

Ans. (c) We call the function DOSOMETHING(c, a, n) and
implement the code in the given steps.
51. Suppose Xi for i = 1, 2, 3 are independent and
identically distributed random variables whose After evaluation, we get answer 0.
probability mass functions are Ans. (0)

54. Let f (n) = n and g(n) = n(1 + sin n), where n is a 55. Consider the following grammar G
positive integer. Which of the following statements
S→FH
is/are correct?
F→pc
I. f (n) = O(g(n))
H→dc
II. f (n) = Ω(g(n))
where S, F, and H are non-terminal symbols, p, d,
(a) Only I and c are terminal symbols. Which of the following
(b) Only II statement(s) is/are correct?
(c) Both I and II S1: LL(1) can parse all strings that are generated
(d) Neither I nor II using grammar G
Solution: S2: LR(1) can parse all strings that are generated
using grammar G
f (n) = n and g(n) = n1 + sin n
(a) Only S1 (b) Only S2
Minimum value of g(n) = n1−1 = 1 (c) Both S1 and S2 (d) Neither S1 nor S2
Maximum value of g(n) = n2 Solution: When we try to parse the given gram-
mar with LL(1) parser or LR(1) parser, there
In either case, it is not equal to f (n). Hence, f (n)
occurs a conflict in both cases. Hence, it cannot be
can’t be best or worst case for g(n).
parsed with any of these parsers.
Ans. (d) Ans. (d)

INDEX
A important points for, 110 level, 125

Abstract data types, 119 initialisation of one-dimensional, 110 node, 125
Abstraction, 107 two-dimensional, 111 parent, 125
Addressing modes, 59—60 Asymptotic notation, 178—180 path, 125
auto-increment or auto-decrement big-O, 179 root, 125
mode, 60 complexities, 180 sibling, 125
base register, 60 omega, 179 strictly, 125
direct mode, 60 properties of, 180 subtree, 125
immediate mode, 59 small-o, 180 Binding, 118—119
implied mode, 59 theta, 178 early, 118—119
index mode, 60 Asynchronous counter, 22 late, 119
indirect mode, 60 Asynchronous interrupts, 66 Bistable multivibrator, 16
register (direct) mode, 59 4-bit synchronous counter, 23
register (indirect) mode, 59 logic diagram for a, 23—24
relative mode, 60 B up counter, 22
Adjacency list, 195 Banker’s algorithm, 342—343 Boolean algebra
Adjacency matrix, 195 Base register addressing mode, 60 basic operations in, 6—7
Algorithm Bellman—Ford algorithm, 201 consensus theorem, 7
analysis, 178 Binary heap, 137 duality theorem, 7
asymptotic notation, 178—180 max-heap tree, 184 huntington’s postulates, 7
defined, 177 min-heap tree, 184—185 negative logic, 8
properties of, 177 time complexity, 186 positive logic, 8
steps to solve a problem, 177 Binary number system, 3 Boolean expression, 8
Algorithmic languages, 106 Binary search Boolean function, 15
All-pair shortest path, 205 pseudocode for, 187 implementation using multiplexers, 15
Alphabets of a language, 238 time complexities, 187 Boolean logic, 6—10
Ambiguities, removal of, 254 Binary search tree, 126 Bubble sort algorithm
Ambiguous grammar, 254, 293 traversal techniques, 126—127 pseudocode for, 187
Amdahl’s law, 58 Binary semaphores, 338 time complexities, 187
AND gate, 6 Binary trees, 125—126
Arithmetic logic unit (ALU) almost complete, 126
arithmetic micro-operations, 61 array representation of, 126 C
logic micro-operations, 61—62 children, 125 Cache memory, 73—74
Arithmetic micro-operations, 61 complete, 125—126 cache mapping techniques, 74—75
Array edge, 125 cache size, 74
address calculation in an, 110—111 height, 125 direct mapping, 74
declaring, 110 leaf node, 125 full associative mapping, 74—75
Index.indd 631 4/9/2015 10:31:22 AM

632 INDEX
mapping function, 74 LAN (local area network), 492—493 detection of, 346
set associative mapping, 75 network layer protocols, 501—505 prevention of, 341—342
write-back updating policies, 74 network security, 510—512 receovery from, 346—347
write-through policy, 74 presentation layer, 508 resource allocation graph, 340—341
Calling function, 120 routing algorithms, 498—499 Decimal number system, 3
Cascading Style Sheets (CSS), 541 session layer, 507 Decoder, 14
Central processing unit (CPU) sockets, 507 logic diagram of, 14
data paths, 64 topology, 492 Defect rate, 464
design, 64—65 Transmission Control Protocol Defect removal efficiency (DRE), 464
general register organization, 62 (TCP), 506—507 Degree of multiprogramming, 330
hardwired control unit, 64 transport layer, 505 Demultiplexer, 16
instruction execution, 63—64 User Datagram Protocol (UDP), 506 Derivation tree, 253
micro-programmed control unit, Computer organization, 57 D flip—flops, 17—18
64—65 Concatenation, 238 truth table of, 18
RISC vs. CISC processors, 65 Concurrency control protocols, 423—424 Digital circuits
single accumulator organization, 62 Consensus theorem, 7 classification of, 11
stack organization, 63 Constructive cost model (COCOMO), combinational circuits, 11—16
Chomsky normal form (CNF), 256 463—464 sequential circuits, 16—26
Class, 107 Context-free grammar, 252 Digital electronics, 3
Clear accumulator (CLA), 59 Context-free grammar simplification, Digital logic families, 26—27
Clear operation, 62 254—256 pros and cons of logic families, 27
Client—server computing, 544—546 Control flow statements Dijkstra’s algorithm, 200—201
functions of a client, 544 break statement, 110 Dining philosopher’s problem, 336
functions of a server, 544—545 conditional statements, 108—110 Direct memory access (DMA), 66—67
merits and demerits, 546 continue statement, 110 burst or block transfer mode, 66
N-tier architecture, 546 goto statement, 110 channel, 66
server types, 544 return statement, 110 controller, 66
three-tier architecture, 545—546 Control hazards, 69 cycle stealing mode, 66
topologies for, 545 Conversions of number system, 4 data transfer mode, 66
two-tier architecture, 545 decimal number system to any interconnection with memory, CPU
Clocked RS flip—flops, 17 number system, 4 and I/O devices, 67
logic diagram of, 17 number system to decimal number modes of operation, 66
Code optimization phase system, 4 Dispatcher, 330
primary source of, 307—308 Counters, 22—26 Do … while loop, 109
types and levels of, 307 Counting semaphores, 338—339 Duality theorem, 7
Column major order, 111 Counting sort Dynamic programming, 203—204
Combinational circuits, 11—16 pseudocode for, 193 applications of, 204—2057
decoder, 14 time complexities, 192
demultiplexer, 16 CPU scheduling algorithms, 331—335
encoder, 14 Critical section problem, 337 E
full adder, 12 Cryptography, 510—511
Early binding, 118—119
full subtractor, 13
Else—if condition statement, 109
half adder, 11—12
Encapsulation, 107
half subtractor, 12
D Encoder, 14
look-ahead carry adder, 13—14
Database design Enumerated data types, 117
multiplexer, 14
attribute closure, 411 Ethernet LAN technology, 492—493
parallel adder, 13
decomposition process, 412—414 Even palindrome, 238
Compiler, 287
integrity constraints, 410—411 Expression conversion, 120
construction tools, 290
keys, 411—412 infix to postfix conversion, 120
grouping of phases, 290
process of normalization, 411 infix to prefix conversion, 121—122
phases of, 288—289
Database management system External interrupts, 65
Complement accumulator (CMA), 59
Computer architecture, 57 (DBMS), 403—404
quantitative principles to design architecture of, 405
high-performance processor, 58 components of, 404—405 F
registers and their functions, 57—58 Data models, 405 Fibonacci series, 204
Computer networks ER model, 407—410 File processing approach, 403—404
application layer, 508—509 relational model, 406—407 File structures
congestion and its causes, 505—506 Data path structures of CPU, 64 B trees, 416
connections, 491 single bus structure, 64 B+ tree structure, 416—417
devices used for internetworking, 509 three bus structure, 64 indexing attributes, 416
digital signature, 511 two bus structure, 64 sequential files, 416
firewalls, 511—512 Data types in programming language, File system
International Standards 107—108 allocation methods, 359—362
Organization/Open Systems Deadlocks, 339 directory structures, 359
Interconnection (Reference Model avoidance of, 342—343 Finite automata, 237—238
1984) (ISO/OSI) model, 494—498 characteristics of, 340 conversion in, 246—249
Index.indd 632 4/9/2015 10:31:22 AM

INDEX 633
deterministic finite automata (DFA), simple graph, 194—195 double precision, 5

241—243, 246—247, 250 spanning tree of, 133—135 single precision, 5
grammar, 239—240 subgraph, 127 If condition statement, 109
model characteristics, 238 traversal techniques for, 127—132 If—else condition statement, 109
non-deterministic finite automata tree, 127 Index address mode, 60
(NFA), 241—243 undirected, 127 Information system, 453—454
technical terms, 238—239 Graph traversal components of, 454
Finite state machine (FSM) breath-first traversal (BFT), decision support system (DSS), 455
mealy machines, 245 201—202 expert system, 455
moore machines, 246 depth-first traversal (DFT), 202—203 integrated, 455
Flip—flops, 16 Greedy algorithm, 195 management information system
applications of, 20—26 applications of, 196 (MIS), 455
clocked RS, 17 important terms, 196 office information system (OIS), 454
conversion equations, 19—20 Greibach normal form (GNF), 256 transaction processing system (TPS),
D, 17—18 Grouping of cells for simplification, 9 454—455
excitation table of, 19 Inheritance, 107
JK, 18 Insertion sort
obtained by K-map simplification, 17 pseudocode for, 188
RS, 16—17
H recurrence relation for, 188
T, 19 Half adder, 11—12 time complexities, 188
Floyd’s algorithm, 205 Half subtractor, 12 Insert operation, 62
Fork() function, 347—348 Hardware interrupts, 65 Instruction execution
For loop, 109 Hardwired control unit, 64 decoding instruction, 63
Fractional knapsack, 196 Hashing executing instruction, 63—64
pseudocode for, 196 collisions, 182—184 fetching instruction, 63
time complexities, 196 double, 183—184 operand fetching, 64
Full adder, 12 functions, 182 Instruction pipelining, 67—69
logic diagram of, 12 hash table, 182 hazards in, 67—69
Full subtractor, 13 linear probing, 183 speed up of a pipelined system over
logic diagram of a, 13 open addressing, 182—184 a non-pipelined system, 67
Function, 111 quadratic probing, 183 Instruction set architecture (ISA), 57
actual and formal arguments, 112 Heap sort Intermediate code generation, 305—307
advantages, 112 algorithm, 188—189 Internal interrupts, 65
call, 112 build heap function, 189 Interpreter, 287
definition of, 112 heapify function, 189 Inter-process communication, 336
parts of, 111—112 pseudocode for, 189 Interrupt initiated I/O, 65
pre-defined, 112 recurrence relation, 189 I/O interface
prototype, 111—112 time complexities, 189 direct memory access (DMA), 66—67
user-defined, 112 High-level languages, 106 interrupts, 65—66
Function point analysis (FPA), 462 HTML, 538—541 programmed, 65
attributes, 538—539 I/O systems
backgrounds, 540 disk scheduling algorithms, 366—368
character entities, 539 disk structure, 363—364
G elements, 539
Global variables, 117—118 formatting, 539
Grammar, 239—240 forms, 541 J
chomsky hierarchy, 239—240 frames, 541 J2EE platform, 546—547
regular, 244 images, 540 application server, 547
regular expressions, 243—244 links, 540 architecture, 547
types, 240 lists, 540 EJB container, 547
Graphs marquees, 541 services, 546—547
adjacent vertices, 127 phrase tags, 540 JK flip—flops, 18
AVL tree of, 135—136 tables, 540—541 logic diagram, 18
balanced binary tree, 127 tags, 538 master—slave, 18
binary heap, 137 Huffman coding, 197 truth table of, 18
complete graph, 195 code tree according to, 198 Johnson counter, 26
connected, 127 principle of, 197
cycle, 127 pseudocode for, 197
degree of a vertex, 127 time complexities, 197 K
directed, 127 Huntington’s postulates for Boolean Karnaugh-map (K-map), 9, 25
Kn graph, 195 algebra, 7 four-variable, 10
multigraph, 194 POS expression using, 9
null graph, 194 SOP expression using, 9
path, 127 I three-variable, 9
regular graph, 195 Identity element, 238 two-variable, 9
representations of, 195 IEEE standard for floating point 0/1 knapsack, 204—205
shortest paths, 132—133 numbers, 5 Kruskal’s algorithm, 199—200
Index.indd 633 4/9/2015 10:31:22 AM

634 INDEX
L recurrence relation for, 191 multitasking, 328

Language complement, 239 time complexities, 191 real-time (RTOS), 329
Language intersections, 239 Metrics scheduler for, 330
Languages, closure and decidability effort and schedule (duration) OR gate, 6
properties of, 258—259 estimation, 463—465 Overloading, 107
Language subtractions, 239 halstead, 465
Language union, 239 process, 461
Language XOR, 239 product, 461
Largest common subsequence, 205 project, 461 P
Late binding, 119 size-oriented, 462 Pairs, 9
Leftmost derivation (LMD) tree, 253 Micro-programmed control unit, Parallel adder, 13
Length of a string, 238 64—65 block diagram of a, 13
Lexical analyzer horizontal, 65 Parallel-in to parallel-out
functions of, 290 vertical, 65 (PIPO), 22
implementation of, 290—291 Micro-program sequencer, 65 Parallel-in to serial-out (PISO), 21
LIFO (last in, first out) mechanism, 62 Minimization of expresson, 8 Parameter-passing techniques
Linear search Minimum spanning tree call-by-reference, 115—116
pseudocode for, 186 Kruskal’s algorithm, 199—200 call-by-value, 115
time complexities, 186—187 Prim’s algorithm, 198—199 Parser, 291
Line of code (LOC), 462 Minterms, 8 algorithm of predictive, 294
Linked list MOD-10 asynchronous counter, bottom-up, 297—300
basic operations, 123—124 23—24 canonical LR, 300—301
implementation of, 124—125 logic diagram for a, 24 context-free grammar (CFG)
Local variables, 117 MOD-5 synchronous counter, 24 for, 292
Logic gate, 6 logic diagram for a, 26 derivation tree or parse tree, 292
symbols, 6 Multiplexer, 14 leftmost derivation (LMD), 292
Logic micro-operations, 61—62 block diagram of, 14 operator precedence, 301
Look-ahead carry adder, 13—14 boolean function implementation rightmost derivation tree
logic diagram of, 14 using, 15 (RMD), 293
Loops, 109 Myhill—Nerode theorem, 245 top-down, 293—294
Parse tree, 253
Parsing, 120, 252
Parsing table, 295—296
M N P-complexity class, 205—206
Machine instructions, 58—59 NAND gate, 6 Peterson’s algorithm, 337—338
data manipulation instruction, 58 N-bit parallel adder, 13 Pipelining hazards, 67—69
data transfer instruction, 58 Negative logic, 8 data hazards, 68
one-address instruction, 59 Non-maskable interrupts, 65 structural hazards, 68
program control instruction, 58 Non-procedural languages, 106 Pointers, 113
three-address instruction, 58 Non-vectored interrupts, 65 arithmetic, 113—114
two-address instruction, 59 NOR gate, 6 dangling, 114—115
zero-address instruction, 59 NP-complete class, 206 Polymorphism, 107
Maskable interrupts, 65 NP-hard class, 206 Positive logic, 8
Mask operation, 62 Number system, 3—6 Power of a string, 238
Master—slave JK flip—flops, 18 binary, 3 Pre-defined function, 112
Maxterms, 8 complement of a number, 4—5 Prefix of a string, 239
Memory conversions of, 4 Prime implicant, 8
cache, 73—74 decimal, 3 Prim’s algorithm, 198—199
main, 72 IEEE standard for floating point Problem-oriented languages, 106
m-bit main, 74 numbers, 5 Procedural languages, 106
secondary, 73 representation of negative Procedural programming
Memory management numbers, 5 languages, 106
loading and linking, 349—350 Process, 329
page replacement algorithms, attributes of, 329
355—358 states, 329—330
paging, 350—354 O synchronization, 335—339
partition allocation policies, Object-oriented languages, 106 time periods of, 330—331
348—349 Object-oriented programming Process models
partitioning methods, 348 (OOP), 106—107 component-based model, 460—461
segmentation technique, 354 Octet, 9 incremental model, 459—460
thrashing, 358 Odd palindrome, 238 prototype model, 458—459
virtual memory, 355 Operating system (OS), 327 rapid application development
Merge sort batch, 328 (RAD) model, 459
applications of, 191 multiprocessor, 328 spiral model, 460
pseudocode for, 190—191 multiprogramming, 328 waterfall model, 458
Index.indd 634 4/9/2015 10:31:22 AM

INDEX 635
Producer—consumer problem, 336 Rightmost derivation (RMD) tree, 253 Sorting, 186
Product-of-sums (POS) form, 8 Ring counter, 26 Spinning or busy waiting, 337
using K-map, 9 Ripple carry adder, 13 Stack, 119
Programmed I/O, 65 Risk analysis applications of, 120
Programming languages, 105—106 identification of risk, 466 POP operation on, 119—120
fifth generation /5GL/(natural risk projection or risk estimation, 466 PUSH operation on, 119
language programming), 106 risk strategies, 466 Stack pointer (SP), 63
first generation/1GL/(machine types of risk, 466 Standard context-free language,
level programming), 106 Row major order, 111 252—253
fourth generation /4GL/(non- RS flip—flops, 16—17 Storage classes, 118
procedural programming), 106 clocked, 17 String of a language, 238
second generation /2GL/(assembly logic diagram of, 17 Structured programming, 107
level programming), 106 truth table of, 17 Structured query languages (SQL), 414
third generation /3GL/(high-level Runtime environment commands, 414—416
programming), 106 activation record and activations Structures, 117
Pumping lemma, 244—245 trees, 304 Substring, 238
Pushdown automata (PDA) lexical vs dynamic scoping, 305 Suffix of a string, 239
deterministic (NPDA), 251 procedure call return model, 304—305 Sum-of-products (SOP) form, 8
model of, 250—251 storage organization, 303—304 using K-map, 9
non-deterministic (NPDA), 251 symbol table, 305 Switch condition statement, 109
transition function of, 251 Synchronous counter, 22
Synchronous interrupts, 66
Syntax-directed translation (SDT)
S applications of, 302
Q Schedule, 417—419 attributes supported by, 302
Quads, 9 based on recoverability, 419 characteristics, 302
Queue, 122 based on serializability, 420—423 types of, 302
applications of, 123 Scoping, 117
basic operations, 122—123 Scripting languages, 106
types of, 123 Searching, 186
Quick sort Security of computers, 368 T
pseudocode for, 192 from accidental data loss, 369 T flip—flops, 19
randomized version of, 193 anti-virus approaches, 371 truth table, 19
recurrence relation for, 192 attacks from outside the system, Threads
time complexities, 192 370—371 benefits of, 347
attacks from within the system, 370 types of, 347—348
cryptography, 369 Three-variable K-map, 9
from threats, 369 Token bus and ring, 493
R Selection sort Transactions, 417
Race-around problem, 18 pseudocode for, 187 Trees
Race condition, 336—337 time complexities, 188 applications, 126
Radix complement, 4 Selective clear operation, 62 binary, 125—126
subtraction using, 6 Selective complement operation, 62 properties of, 125
Radix minus one complement, 4 Selective set operation, 62 Truth table, 12—14
subtraction using, 4—5 Semaphores, 338 Turing machine (TM), 256—258
Radix or base, 4 Sequential circuits, 16—26 Two-variable K-map, 9
Read after write (RAW), 68 Serial-in/serial-out shift register, 20—21
Readers—writers problem, 336 Serial-in to parallel-out (SIPO), 21
Recurrence relations Shift left register, 21 U
iteration method, 180 Shift register counter, 26
Unions, 117
master theorem, 181 Shift right register, 20
User-defined function, 112
recursion tree, 181 Sockets, 507
Recursion, 112 Software
advantages, 113 architecture, 472
function, 112—113 characteristics of, 456 V
Recursive and recursive enumerable coding, 472 Vectored interrupts, 65
languages, 258 design stage, 470—472 Virtual memory, 355
Recursive function, 120 engineering, 457
Registers, 20 interrupts, 65
Regular sets/languages, properties testing, 472—477
of, 249 Software Development Life Cycle (SDLC) W
Relative address mode, 60 elicitation techniques, 469—470 Web technology
Return statement, 112 external interfaces, 468 Cascading Style Sheets (CSS), 541
Reversal of a string, 238 requirement, 467—468 Document Object Model (DOM), 543
Reversing a list, 120 requirement engineering, 468—469 flash, 543
Index.indd 635 4/9/2015 10:31:22 AM

636 INDEX
milestones in development X vocabularies, 542

of, 537—538 2X4 decoder, 14 XNOR gate, 6
silverlight, 543 XHTML, 543 XOR gate, 6
user-interface languages, 543 XML, 541—543 XUL, 543
World Wide Web (WWW), 537 advantages of, 542
While loop, 109 documents, 543
Write after read (WAR), 69 tag rules for, 542
Write after write (WAW), 69
Index.indd 636 4/9/2015 10:31:22 AM

Anil Kumar Verma, Gaurav Sharma, Kuldeep Singh - WILEY ACING THE GATE COMPUTER SCIENCE AND INFORMATION TECHNOLOGY (2015, Wiley India) PDF

Uploaded by

Document Information

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Anil Kumar Verma, Gaurav Sharma, Kuldeep Singh - WILEY ACING THE GATE COMPUTER SCIENCE AND INFORMATION TECHNOLOGY (2015, Wiley India) PDF

Uploaded by

Copyright:

Available Formats

WILEY ACING THE GATE

COMPUTER SCIENCE AND

Anil Kumar Verma

Other Wiley Editorial Offices:

First Edition: 2015

FM.indd 5 4/9/2015 10:14:41 AM

FM.indd 6 4/9/2015 10:14:42 AM

Gaurav Sharma is currently working as an Assistant Professor in the School of Computing

FM.indd 7 4/9/2015 10:14:42 AM

Applying for the Examination

FM.indd 9 4/9/2015 10:14:42 AM

Computerized Mode of Examination

FM.indd 10 4/9/2015 10:14:42 AM

ATTRIBUTES FOR SUCCESS IN EXAMINATION

Preparing for the Examination

Knowing Your Subject (Knowledge of Material)

Preparation of a Study Plan

Refining Problem-Solving Skills

FM.indd 11 4/9/2015 10:14:42 AM

Taking the Examination

Effective Preparation and Time Management

Maximizing your Exam-Taking Efficiency

FM.indd 12 4/9/2015 10:14:42 AM

Acing the GATE ix

FM.indd 13 4/9/2015 10:14:42 AM

2 Computer Organization and Architecture 57

3 Programming and Data Structures 105

FM.indd 14 4/9/2015 10:14:42 AM

3.2.9 Parameter Passing 115

FM.indd 15 4/9/2015 10:14:42 AM

4.5 Searching and Sorting 186

5 Theory of Computation 237

FM.indd 16 4/9/2015 10:14:42 AM

5.2.4 Noam Chomsky Grammar Classification 239

6 Compiler Design 287

FM.indd 17 4/9/2015 10:14:42 AM

6.2.2 Interpreter 288

7 Operating System 327

FM.indd 18 4/9/2015 10:14:42 AM

7.5 Process Synchronization 335

FM.indd 19 4/9/2015 10:14:42 AM

9 Information Systems and Software Engineering 453

FM.indd 20 4/9/2015 10:14:43 AM

9.5 Measurement of Metrics 461

10 Computer Networks 491

FM.indd 21 4/9/2015 10:14:43 AM

Important Formulas 512

11 Web Technologies 537

FM.indd 22 4/9/2015 10:14:43 AM

Important Formulas 547

Solved GATE (CS) 2014 Set 1 555

FM.indd 23 4/9/2015 10:14:43 AM

LAST SIX YEARS' GATE ANALYSIS

Concepts on which questions were asked in the previous six years

Chapter 1.indd 1 4/9/2015 9:40:08 AM

1.1 INTRODUCTION Number system

Number system is an age old method to represent BCD XS3 code

Chapter 1.indd 3 4/9/2015 9:40:08 AM

Table 1.1 | Types of number systems

Another example is of (101.11)2 = (5.75)10 = (5.6)8 =

2 14 0 2 × 0.3 = 0.6 0 1.2.2.1 Subtraction Using R-1’s Complement

Chapter 1.indd 4 4/9/2015 9:40:09 AM

If there is no carry, then find the R-1’s complement

Acing the GATE ix

2 Computer Organization and Architecture 57

3 Programming and Data Structures 105

3.2.9 Parameter Passing 115

4.5 Searching and Sorting 186

5 Theory of Computation 237

5.2.4 Noam Chomsky Grammar Classification 239

6 Compiler Design 287

6.2.2 Interpreter 288

7 Operating System 327

7.5 Process Synchronization 335

9 Information Systems and Software Engineering 453

9.5 Measurement of Metrics 461

10 Computer Networks 491

Important Formulas 512

11 Web Technologies 537

Important Formulas 547

Solved GATE (CS) 2014 Set 1 555