You are on page 1of 11

INTRODUCTION: ‘Queuing theory’ is the mathematical study of waitin g lines, or queues.

The theory enables mathematical analysis of several related processes, including arriving at the (back of the) queue, waiting in the queue ( essentially a storage process), and being served at the front of the queue. The theory permits the derivation and calculation of several performance measures in cluding the average waiting time in the queue or the system, the expected number waiting or receiving service, and the probability of encountering the system in certain states, such as empty, full, having an available server or having to wa it a certain time to be served. Queuing theory has applications in diverse fie lds, including telecommunications, traffic engineering, computing and the design of factories, shops, offices and hospitals. DEFINITION: What is queuing theory? Queuing Theory is a collection of mathematical models of various que uing systems. It is used extensively to analyze production and service processes e xhibiting random variability in market demand (arrival times) and service times. Can you tell why queues form? Queues or waiting lines arise when the demand for a service facility e xceeds the capacity of that facility, that is, the customers do not get service immediately upon request but must wait, or the service facilities stand idle and wait for customers. Some customers wait when the total number of customers requiring serv ice exceeds the number of service facilities, some service facilities stand idle when the total number of service facilities exceeds the number of customers req uiring service. Waiting lines, or queues are a common occurrence both in everyd ay life and in variety of business and industrial situations. Most waiting line problems are centered about the question of finding the ideal level of services that a firm should provide. For example: • Supermarkets must decide how many cash register check out positions should be op ened. • Gasoline stations must decide how many pumps should be opened and how many atten dants should be on duty. • Manufacturing plants must determine the optimal number of mechanics to have on d uty in each shift to repair machines that break down. • Banks must decide how many teller windows to keep open to serve customers during various hours of the day. Some more examples of waiting lines are given below Situation Arriving Customers Service Facility Passage of customers through a supermarket checkout Shoppers Checkout counters Flow of automobile traffic through a road network work OVERVIEW: The word queue comes, via French, from the Latin cauda, meaning ta il. The spelling "queuing" over "queuing" is typically encountered in the academ ic research field. In fact, one of the flagship journals of the profession is na Automobiles Road net

med "Queuing Systems". Queuing theory is generally considered a branch of operations resear ch because the results are often used when making business decisions about the r esources needed to provide service. It is applicable in a wide variety of situat ions that may be encountered in business, commerce, industry, healthcare, public service and engineering. Applications are frequently encountered in customer se rvice situations as well as transport and telecommunication. Queueing theory is directly applicable to intelligent transportation systems, call centers, PABXs, networks, telecommunications, server queuing, mainframe computer of telecommunic ations terminals, advanced telecommunications systems, and traffic flow. Notation for describing the characteristics of a queuing model was first suggested by David G. Kendall in 1953. Kendall s notation introduced an A/ B/C queuing notation that can be found in all standard modern works on queuing t heory, for example, Times. The A/B/C notation designates a queuing system having A as interrac ial time distribution, B as service time distribution, and C as number of server s. For example, "G/D/1" would indicate a General (may be anything) arrival proce ss, a Deterministic (constant time) service process and a single server. More de tails on this notation are given in the article about queuing models. HISTORY OF QUEUING THEORY: Queuing Theory had its beginning in the research work of a Dani sh engineer named A. K. Erlang. In 1909 Erlang experimented with fluctuating dem and in telephone traffic. Eight years later he published a report addressing the delays in automatic dialing equipment. At the end of World War II, Erlang’s early work was extended to more general problems and to business applications of wait ing lines. BASIC COMPONENTS OF A QUEUING SYSTEM: Firstly there are some basic components in every queuing system: INPUT SOURCE OF QUEUE: An input source is characterized by • Size of the calling population • Pattern of arrivals at the system • Behaviour of the arrivals Customers requiring service are generated at different times by an input source, commonly known as population. The rate at which customers arri ve at the service facility is determined by the arrival process. Size of the calling population: The size represents the total number of potential customers who will require service. According to source : The source of customers can be finite or infinite. For example, all people of a city or state (and others) could be the potential customers at a supermarket. The number of people being very large, it can be taken to be infi nite. Whereas there are many situations in business and industrial conditions wh ere we cannot consider the population to be infinite—it is finite. According to numbers: The customers may arrive for service individually or in groups. Single a rrivals are illustrated by patients visiting a doctor, students reaching at a li brary counter etc. On the other hand, families visiting restaurants, ships disch arging cargo at a dock are examples of bulk, or batch arrivals. According to time: Customers arrive in the system at a service facility accordin g to some known schedule (for example one patient every 15 minutes or a candidat e for interview every half hour) or else they arrive randomly. Arrivals are cons idered random when they are independent of one another and their occurrence cann ot be predicted exactly. The queuing models wherein customers’ arrival times are k nown with certainity are categorized as deterministic models. (insofar as this c haracteristic is concerned) and are easier to handle. On the other hand, a subst antial majority of the queuing models are based on the premise that the customer s enter the system stochastically, at random points in time.


Pattern of arrivals at the system— The arrival process (or pattern) of customers to the service sy stem is classified into two categories: static and dynamic. These two are furthe r classified based on the nature of arrival rate and the control that can be exe rcised on the arrival process. In static arrival process, the control depends on the nature of arrival rate (random or constant). Random arrivals are either at a constant r ate or varying with time. Thus to analyze the queuing system, it is necessary to attempt to describe the probability distribution of arrivals. From such distrib utions we obtain average time between successive arrivals, also called inter-arr ival time (time between two consecutive arrivals), and the average arrival rate (i.e. number of customers arriving per unit of time at the service system). The dynamic arrival process is controlled by both service fa cility and customers. The service facility adjusts its capacity to match changes in the demand intensity, by either varying the staffing levels at different tim ings of service, varying service charges (such as telephone call charges at diff erent hours of the day or week) at different timings, or allowing entry with app ointments. Frequently in queuing problems, the number of arrivals per un it of time can be estimated by a probability distribution known as the Poisson d istribution, as it adequately supports many real world situations. Behavior of arrivals— Another thing to consider in the queuing structure is the beha vior or attitude of the customers entering the queuing system. On this basis, the customers may be classified as being (a) patient, or (b) impatient. If a customer, on arriving at the service system stays in the system until served, no matter how much he has to wait for service is called a p atient customer. Machines arrived at the maintenance shop in a plant are examples of patient customers. Whereas the customer, who waits for a certain time in the queue and leaves the service system without getting service due to certain reas ons such as a long queue in front of him is called an impatient customer.

Now, Let us see some interesting observations of human behavior in queues : • Balking – Some customers even before joining the queue get discouraged by seeing t he number of customers already in service system or estimating the excessive wai ting time for desired service, decide to return for service at a later time. In queuing theory this is known as balking. • Reneging - customers after joining the queue, wait for sometime and leave the se rvice system due to intolerable delay, so they renege. For example, a customer who has just arrived at a grocery stor e and finds that the salesmen are busy in serving the customers already in the s ystem, will either wait for service till his patience is exhausted or estimates that his waiting time may be excessive and so leaves immediately to seek service elsewhere. • Jockeying - Customers who switch from one queue to another hoping to receive ser vice more quickly are said to be jockeying

SERVICE SYSTEM : The service is provided by a service facility (or facilitie s). This may be a person (a bank teller, a barber, a machine (elevator, gasoline pump), or a space (airport runway, parking lot, hospital bed), to mention just a few. A service facility may include one person or several people operating as a team. There are two aspects of a service system—(a) the configuration of the service sys tem and (b) the speed of the service.

a) Configuration of the service system: The customers’ entry into the service system depends upon the queue conditions. If at the time of customers’ arrival, the server is idle, th en the customer is served immediately. Otherwise the customer is asked to join t he queue, which can have several configurations. By configuration of the service system we mean how the service facilities exist. Service systems are usually cl assified in terms of their number of channels, or numbers of servers.

Single Server – Single Queue -- The models that involve one queue – one service stat ion facility are called single server models where customer waits till the servi ce point is ready to take him for servicing. Students arriving at a library coun ter is an example of a single server facility.

Single Server – Single Queue Model

Single Server – Several Queues – In this type of facility there are several queues a nd the customer may join any one of these but there is only one service channel.

Single Server – Several Queue Model Several (Parallel) Servers – Single Queue – In this type of model there is more than one server and each server provides the same type of facility. The customers wa it in a single queue until one of the service channels is ready to take them in for servicing.

Several, Parallel Servers – Single Queue Model Several Servers – Several Queues – This type of model consists of several servers wh ere each of the servers has a different queue. Different cash counters in an ele ctricity office where the customers can make payment in respect of their electri city bills provide an example of this type of model.

Several, Parallel Servers – Several Queues Mo del Service facilities in a series – In this, a customer enters the first station and gets a portion of service and then moves on to the next station, gets some servi ce and then again moves on to the next station. …. and so on, and finally leaves t he system, having received the complete service. For example, machining of a cer tain steel item may consist of cutting, turning, knurling, drilling, grinding, a nd packaging operations, each of which is performed by a single server in a seri es.

Multiple Servers in a Series b) Speed of Service: In a queuing system, the speed with which service is provided can be expressed in either of two ways—as service rate and as service time. • The service rate describes the number of customers serviced during a particular time period. • The service time indicates the amount of time needed to service a customer. • Service rates and times are reciprocal of each other and either of them is suffi cient to indicate the capacity of the facility. Thus if a cashier can attend, on an average 5 customers in an hour, the service rate would be expressed as 5 customers/hour and service time w ould be equal to 12 minutes/customer. Generally, we consider the service time only. If these service times are known exactly, the problem can be han dled easily. But, as generally happens. if these are different and not known wit

h certainty, we have to consider the distribution of the service times in order to analyze the queuing system. Generally, the queuing models are based on the as sumption that service times are exponentially distributed about some average ser vice time. QUEUE CONFIGURATION : The queuing process refers to the number of queues, and t heir respective lengths. The number of queues depend upon the layout of a servic e system. Thus there may be a single queue or multiple queues. Length (or size) of the queue depends upon the operational situation such as • physical space, • legal restrictions, and • attitude of the customers. In certain cases, a service system is unable to accommodate more t han the required number of customers at a time. No further customers are allowed to enter until space becomes available to accommodate new customers. Such type of situations are referred to as finite (or limited) source queue. Examples of finite source queues are cinema halls, restaurants, etc. On the other hand, if a service system is able to accommodate any number of cust omers at a time, then it is referred to as infinite (or unlimited) source. queue . For example, in a sales department, here the customer orders are received, there is no restriction on the number of orders that can come in, so that a queue of any size can form. In many other situations, when arriving customers experience long queue(s) in fr ont of a service facility, they often do not enter the service system even thoug h additional waiting space is available. The queue length in such cases depends upon the attitude of the customers. For example, when a motorist finds that there are many vehicles wai ting at the petrol station, in most of the cases he does not stop at this statio n and seeks service elsewhere. QUEUE DISCIPLINE : In the queue structure, the important thing to know is the q ueue discipline. The queue discipline is the order or manner in which customers from the queue are selected for service. There are a number of ways in which customers in the queue are served. Some of t hese are: (a) Static queue disciplines are based on the individual customer s status i n the queue. Few of such disciplines are: If the customers are served in the order of their arrival, then this is known as the first-come, first-served (FCFS) service discipline. Prepaid taxi queue at airports where a taxi is engaged on a first-come, first-served ba sis is an example of this discipline. Last-come-first-served (LCFS)-- Sometimes, the customers are se rviced in the reverse order of their entry so that the ones who join the last ar e served first. For example, assume that letters to be typed, or order forms to be processed accumulate in a pile, each new addition being put on the top of the m. The typist or the clerk might process these letters or orders by taking each new task from the top of the pile. Thus, a just arriving task would be the next to be serviced provided that no fresh task arrives before it is picked up. Simil arly, the people who join an elevator last are the first ones to leave it. (b) Dynamic queue disciplines are based on the individual customer attributes in the queue. Few of such disciplines are: Service in Random Order (SIRO)-- Under this rule customers are selected for serv


ice at random, irrespective of their arrivals in the service system. In this eve ry customer in the queue is equally likely to be selected. The time of arrival o f the customers is, therefore, of no relevance in such a case. Priority Service-- Under this rule customers are grouped in priority classes on the basis of some attributes such as service time or urgency or according to som e identifiable characteristic, and FCFS rule is used within each class to provid e service. Treatment of VIPs in preference to other patients in a hospital is an example of priority service. For the queuing models that we shall consider, the assumption would be that the customers are serviced on the first-come-first-served basis. QUEUING NETWORKS: Networks of queues are systems which contain an arbitrary, but fi nite, number m of queues. Customers, sometimes of different classes, travel thro ugh the network and are served at the nodes. The state of a network can be descr ibed by a vector , where ki is the number of customers at queue i. In open netw orks, customers can join and leave the system, whereas in closed networks the to tal number of customers within the system remains fixed. The first significant result in the area was Jackson networks, for which an efficient product form equilibrium distribution exists. QUEUING MODELS: In queuing theory, a queuing model is used to approximat e a real queuing situation or system, so the queuing behaviour can be analysed m athematically. Queuing models allow a number of useful steady state performance measures to be determined, including: • the average number in the queue, or the system, • the average time spent in the queue, or the system, • the statistical distribution of those numbers or times, • the probability the queue is full, or empty, and • the probability of finding the system in a particular state. These performance measures are important as issues or problems caused by queuing situations are often related to customer dissatisfaction with service or may be the root cause of economic losses in a business. Analysis of the relevant queuing models allows the cause of queuing issues to be identified and the impact of proposed changes to be assessed. • Notation: Queuing models can be represented using Kendall s notation: A/B/S/K/N/D where: • A is the interracial time distribution • B is the service time distribution • S is the number of servers • K is the system capacity • N is the calling population • D is the service discipline assumed Many times the last members are omitted, so the notation becomes A/B/S and it is assumed that K = , N = and D = FIFO. Some standard notation for distributions (A or B) are: • M for a Markovian (poisson, exponential) distribution • Eκ for an Erlang distribution with κ phases • D for degenerate (or deterministic) distribution (constant) • G for general distribution (arbitrary) • PH for a phase-type distribution Models: Construction and analysis: Queuing models are generally constructed to represent the ste ady state of a queuing system, that is, the typical, long run or average state o f the system. As a consequence, these are stochastic models that represent the p


robability that a queuing system will be found in a particular configuration or state. A general procedure for constructing and analysing such queuing models is: 1. Identify the parameters of the system, such as the arrival rate, service time, queue capacity, and perhaps draw a diagram of the system. 2. Identify the system states. (A state will generally represent the intege r number of customers, people, jobs, calls, messages, etc. in the system and may or may not be limited.) 3. Draw a state transition diagram that represents the possible system stat es and identify the rates to enter and leave each state. This diagram is a repre sentation of a Markov chain. 4. Because the state transition diagram represents the steady state situati on between state there is a balanced flow between states so the probabilities of being in adjacent states can be related mathematically in terms of the arrival and service rates and state probabilities. 5. Express all the state probabilities in terms of the empty state probabil ity, using the inter-state transition relationships. 6. Determine the empty state probability by using the fact that all state p robabilities always sum to 1. Whereas specific problems that have small finite state models can often be analysed numerically, analysis of more general models, using calculus, yields useful formulae that can be applied to whole classes of problems. Single-server queue: Single-server queues are, perhaps, the most commonly enco untered queuing situation in real life. One encounters a queue with a single ser ver in many situations, including business (e.g. sales clerk), industry (e.g. a production line), transport (e.g. a queues that the customer can select from.) C onsequently, being able to model and analyse a single server queue s behaviour i s a particularly useful thing to do. Poisson arrivals and service: M/M/1/ / represents a single server that has unlimited queue c apacity and infinite calling population, both arrivals and service are Poisson ( or random) processes, meaning the statistical distribution of both the inter-arr ival times and the service times follow the exponential distribution. Because of the mathematical nature of the exponential distribution, a number of quite simp le relationships are able to be derived for several performance measures based o n knowing the arrival rate and service rate. This is fortunate because an M/M/1 queuing model can be used to a pproximate many queuing situations. Poisson arrivals and general service: M/G/1/ / represents a single server that has unlimited qu eue capacity and infinite calling population, while the arrival is still Poisson process, meaning the statistical distribution of the inter-arrival times still follow the exponential distribution, the distribution of the service time does n ot. The distribution of the service time may follow any general statistical dist ribution, not just exponential. Relationships are still able to be derived for a (limited) number of performance measures if one knows the arrival rate and the mean and variance of the service rate. However the derivations are generally mor e complex and difficult. A number of special cases of M/G/1 provide specific solutions that give broad insights into the best model to choose for specific queuing situ ations because they permit the comparison of those solutions to the performance of an M/M/1 model. Multiple-servers queue: Multiple (identical)-servers queue situations are frequently encoun tered in telecommunications or a customer service environment. When modelling th ese situations care is needed to ensure that it is a multiple servers queue, not a network of single server queues, because results may differ depending on how the queuing model behaves. One observational insight provided by comparing queuing models is t


hat a single queue with multiple servers performs better than each server having their own queue and that a single large pool of servers performs better than tw o or more smaller pools, even though there are the same total number of servers in the system. One simple example to prove the above fact is as follows: Consider a system having 8 input lines, single queue and 8 servers.The output line has a capacity of 64 kbit/s. Considering the arrival rate at each input as 2 packets/ s. So, the total arrival rate is 16 packets/s. With an average of 2000 bits per packet, the service rate is 64 kbit/s/2000b = 32 packets/s. Hence, the average r esponse time of the system is 1/(μ − λ) = 1/(32 − 16) = 0.0625 sec. Now, consider a seco nd system with 8 queues, one for each server. Each of the 8 output lines has a c apacity of 8 kbit/s. The calculation yields the response time as 1/(μ − λ) = 1/(4 − 2) = 0.5 sec. And the average waiting time in the queue in the first case is ρ/(1 − ρ)μ = 0. 03125, while in the second case is 0.25. Infinitely many servers: While never exactly encountered in reality, an infinite-servers ( e.g. M/M/ ) model is a convenient theoretical model for situations that involve storage or delay, such as parking lots, warehouses and even atomic transitions. In these models there is no queue, as such, instead each arriving customer recei ves service. When viewed from the outside, the model appears to delay or store e ach customer for some time. ROLE OF POISSON PROCESS, EXPONENTIAL DISTRIBUTIONS: A useful queuing model represents a real-life system with s ufficient accuracy and is analytically tractable. A queuing model based on the P oisson process and its companion exponential probability distribution often meet s these two requirements. A Poisson process models random events (such as a cust omer arrival, a request for action from a web server, or the completion of the a ctions requested of a web server) as emanating from a memoryless process. That i s, the length of the time interval from the current time to the occurrence of th e next event does not depend upon the time of occurrence of the last event. In t he Poisson probability distribution, the observer records the number of events t hat occur in a time interval of fixed length. In the (negative) exponential prob ability distribution, the observer records the length of the time interval betwe en consecutive events. In both, the underlying physical process is memoryless. Models based on the Poisson process often respond to i nputs from the environment in a manner that mimics the response of the system be ing modeled to those same inputs. The analytically tractable models that result yield both information about the system being modeled and the form of their solu tion. Even a queuing model based on the Poisson process that does a relatively p oor job of mimicking detailed system performance can be useful. The fact that su ch models often give "worst-case" scenario evaluations appeals to system designe rs who prefer to include a safety factor in their designs. Also, the form of the solution of models based on the Poisson process often provides insight into the form of the solution to a queuing problem whose detailed behavior is poorly mim icked. As a result, queuing models are frequently modeled as Poisson processes t hrough the use of the exponential distribution. APPLICATION TO TELEPHONY: The public switched telephone network (PSTN) is designed to accom modate the offered traffic intensity with only a small loss. The performance of loss systems is quantified by their grade of service, driven by the assumption t hat if sufficient capacity is not available, the call is refused and lost. Alter natively, overflow systems make use of alternative routes to divert calls via di fferent paths — even these systems have a finite traffic carrying capacity. However, the use of queuing in PSTNs allows the systems to queue t heir customers requests until free resources become available. This means that if traffic intensity levels exceed available capacity, customer s calls are not lost; customers instead wait until they can be served. This method is used in qu euing customers for the next available operator. A queuing discipline determines the manner in which the exchange han



dles calls from customers. It defines the way they will be served, the order in which they are served, and the way in which resources are divided among the cust omers. Here are details of four queuing disciplines: First in first out This principle states that customers are served one at a time and that the custo mer that has been waiting the longest is served first. Last in first out This principle also serves customers one at a time, however the customer with th e shortest waiting time will be served first. Also known as a stack. Processor sharing Customers are served equally. Network capacity is shared between customers and t hey all effectively experience the same delay. Priority Customers with high priority are served first. Queuing is handled by control processes within exchanges, which can be modelled using state equations. Queuing systems use a particular form of state equations known as a Markov chain that models the system in each state. I ncoming traffic to these systems is modelled via a Poisson distribution and is s ubject to Erlang’s queuing theory assumptions viz. • Pure-chance traffic – Call arrivals and departures are random and independent even ts. • Statistical equilibrium – Probabilities within the system do not change. • Full availability – All incoming traffic can be routed to any other customer withi n the network. • Congestion is cleared as soon as servers are free. Classic queuing theory involves complex calculations to determi ne waiting time, service time, server utilization and other metrics that are use d to measure queuing performance. LIMITATIONS OF QUEUING THEORY: The assumptions of classical queuing theory may be too restrict ive to be able to model real-world situations exactly. The complexity of product ion lines with product-specific characteristics cannot be handled with those mod els. Therefore specialized tools have been developed to simulate, analyze, visua lize and optimize time dynamic queuing line behavior. For example; the mathematical models often assume infinite num bers of customers, infinite queue capacity, or no bounds on inter-arrival or ser vice times, when it is quite apparent that these bounds must exist in reality. O ften, although the bounds do exist, they can be safely ignored because the diffe rences between the real-world and theory is not statistically significant, as th e probability that such boundary situations might occur is remote compared to th e expected normal situation. Furthermore, several studies show the robustness of queuing models outside their assumptions. In other cases the theoretical soluti on may either prove intractable or insufficiently informative to be useful. Alternative means of analysis have thus been devised in order to pr ovide some insight into problems that do not fall under the scope of queuing the ory, although they are often scenario-specific because they generally consist of computer simulations or analysis of experimental data. See network traffic simu lation. REFERENCES: 1. 2. TU Berlin: Technische Universität Berlin 3. Lawrence W. Dowdy, Virgilio A.F. Almeida, Daniel A. Menasce (Thursday Ja nery 15, 2004). "Performance by Design: Computer Capacity Planning By Example". pp. 480. 4. Schlechter, Kira (Monday March 02, 2009). "Hershey Medical Center to ope n redesigned emergency room". The Patriot-News. /index.ssf/2009/03/hershey_med_to_open_redesigned.html 5. Mayhew, Les; Smith, David (December 2006). Using queuing theory to analy se completion times in accident and emergency departments in the light of the Go

vernment 4-hour target. Cass Business School. ISBN 978-1-905752-06-5. http://www Retrieved 2008-05-20. 6. Tijms, H.C, Algorithmic Analysis of Queues", Chapter 9 in A First Course in Stochastic Models, Wiley, Chichester, 2003 7. 8. Flood, J.E. Telecommunications Switching, Traffic and Networks, Chapter 4: Telecommunications Traffic, New York: Prentice-Hall, 1998. 9. Bose S.J., Chapter 1 - An Introduction to Queueing Systems, Kluwer/Plenu m Publishers, 2002. 10. Penttinen A., Chapter 8 – Queueing Systems, Lecture Notes: S-38.145 - Intr oduction to Teletraffic Theory. 11. F. P. Kelly Networks of Queues with Customers of Different Types Journal of Applied Probability, Vol. 12, No. 3 (Sep., 1975), pp. 542-554 12. Wikipedia 13. Quantitative Technique in Management by N.D.Vora. 14. Operation Research by J.K.Sharma 15. Operation Research by K.Sarup 16. 17. 18.