CSEN 404

Introduction to Networks

Hisham Othman
Basma Mohamed Afifi
Nadeen Hamza

** Slides are attributed to J. F. Kurose

Lecture Outline
 Transport-layer services
 Multiplexing and
demultiplexing of
applications traffic on a
machine
 Connectionless transport:
UDP
 Principles of reliable data
transfer
 Connection-oriented
transport: TCP
 segment structure
 reliable data transfer
 flow control
 connection management
 Principles of congestion
control
Transport services and protocols
application
 provide logical communication transport
network
between app processes running on data link
physical
different hosts
 transport protocols run in end
systems
 send side: breaks app
messages into segments, passes
to network layer
 receive side: reassembles application
transport
segments into messages, passes network
data link
to app layer physical

 more than one transport protocol

available to apps
 Internet: TCP and UDP
Internet transport-layer protocols
 reliable, in-order delivery application
transport
(TCP) network
data link
physical
 connection setup network
data link
network
physical
 congestion control data link
physical

 flow control
 unreliable, unordered network
data link
physicalnetwork
delivery: UDP data link
physical
 no-frills extension of “best- network
data link
application
effort” IP physical network transport
data link network
 services not available: physical data link
physical

 delay guarantees
 bandwidth guarantees
Best Effort vs. Reliable
 Both TCP and UDP use the IP protocol for data delivery
 IP is a best-effort delivery protocol
 No guarantees for segment delivery
 No guarantees for ordered delivery
 No guarantees for data integrity
 Both TCP & UDP extend host-to-host delivery to process-to-
process delivery through multiplexing & demultiplexing
 Both TCP and UDP provide data integrity through checksums
 TCP extends IP delivery service and makes it reliable
 Uses flow control, handshaking (acknowledgement), sequencing, and
timing for reliable and ordered segment delivery
 Also uses congestion control for balanced use of network resources
 UDP doesn’t care about reliable delivery
Multiplexing/demultiplexing
Examples of Transport-layer Protocols
used by Applications

Application Application-Layer Protocol Underlying Transport

Protocol
E-mail SMTP TCP
Remote terminal access Telnet TCP
Web HTTP TCP
File transfer FTP TCP
Streaming multimedia Typically proprietary UDP or TCP
Internet telephony Typically proprietary UDP or TCP
Network management SNMP Typically UDP
Routing protocol RIP Typically UDP
Name translation DNS Typically UDP
UDP: User Datagram Protocol
UDP: User Datagram Protocol [RFC 768]
 “no frills,” “bare bones” Internet
transport protocol Why is there a UDP?
 “best effort” service, UDP  no connection establishment
segments may be: (which means less delay)
 lost  simple: no connection state at
 delivered out of order to
sender, receiver
app  small segment header
 connectionless:  no congestion control: UDP can
 no handshaking between blast away as fast as desired
UDP sender, receiver
 each UDP segment handled
independently of others
UDP: more

 often used for streaming 32 bits

multimedia apps source port # dest port #
Length, in
 loss tolerant
bytes of UDP length checksum
 rate sensitive segment,
 other UDP uses including
header
 DNS
 SNMP (Simple Network
Application
Management Protocol)
data
 How to achieve reliable transfer (message)
over UDP: add reliability at
application layer
 application-specific error
recovery! UDP segment format
TCP: Transmission Control Protocol
 Reliable Data Transfer

Application
Layer

rdt_send() deliver_data()

Transport
Layer
Reliable Channel
udt_send() rdt_rcv()

Network
Layer

Unreliable Channel
Provided Service Service Implementation
TCP: Overview
 TCP: Features
 point-to-point:  Receiver feedback
 one sender, one receiver  acknowledgements
 reliable, in-order byte  full duplex data:
stream:  bi-directional data flow in
 no “message boundaries” same connection
 connection-oriented:  MSS: maximum segment size
 handshaking (exchange of  flow controlled:
control msgs) initializes sender  sender will not overwhelm
& receiver state before data receiver
exchange
 pipelined:
application application
socket
door
writes data reads data
socket  TCP congestion and flow
door
TCP
send buffer
TCP
receive buffer
control set window size
send & receive buffers
segment

TCP Connection Management

Three-way handshake:
Step 1: client host sends TCP SYN
segment to server
 specifies initial seq. #

 no data

Step 2: server host receives SYN,

replies with SYNACK segment
 server allocates buffers
 specifies server initial seq. #
Step 3: client receives SYNACK,
replies with ACK segment, which
may contain data
TCP Connection Management (cont.)

Closing a connection: client server

client closes socket: close

clientSocket.close();

Step 1: client end system sends TCP

FIN control segment to server close

Step 2: server receives FIN, replies

with ACK. Closes connection,

timed wait
sends FIN

closed
TCP Connection Management (cont.)

Step 3: client receives FIN, replies client server

with ACK. closing
 Enters “timed wait” - will
respond with ACK to
received FINs
closing
Step 4: server, receives ACK.
Connection closed

timed wait closed

closed
TCP segment structure

32 bits
URG: urgent data
source port # dest port # counting
(generally not used)
by bytes
sequence number of data
ACK: ACK # (not segments!)
valid acknowledgement number
head not Receive window
PSH: push data now UA P R S F
len used
(generally not used) # bytes
checksum Urg data pointer rcvr willing
to accept
RST, SYN, FIN: Options (variable length)
connection estab
(setup, teardown
commands)
application
data
Internet (variable length)
checksum
(as in UDP)
Receiver Feedback – ACKs
Pipelined protocols

Pipelining: sender allows multiple, “in-flight”, yet-to-be-

acknowledged pkts
 range of sequence numbers must be used
 buffering at sender and/or receiver

 TCP is a pipelined protocol

 Pipelining: Increased Utilization
Ex: L = 8000 bits, R = 1 Gbps, RTT = 30 ms. Find the sender utilization of the link, Usender.

3 ∗ 𝐿/𝑅 𝐿/𝑅
𝑈𝑠𝑒𝑛𝑑𝑒𝑟 = 𝑈𝑠𝑒𝑛𝑑𝑒𝑟 =
𝑅𝑇𝑇 + 𝐿/𝑅 𝑅𝑇𝑇 + 𝐿/𝑅
𝑈𝑠𝑒𝑛𝑑𝑒𝑟 = 0.00081 𝑈𝑠𝑒𝑛𝑑𝑒𝑟 = 0.00027
TCP sender events:
data rcvd from app: timeout:
 create segment with seq #  retransmit segment that

 seq # is byte-stream caused timeout

number of first data byte  restart timer
in segment ACK rcvd:
 start timer if not already  if acknowledges previously
running (timer is set in unACK’ed segments
terms of oldest unACK’ed  update what bytes known to
segment) be ACK’ed
 start timer if there are
outstanding segments
 TCP: Cumulative ACK scenario

Host A Host B
timeout

X
loss

time
Cumulative ACK scenario
TCP: retransmission scenarios

Host A Host B Host A Host B

Seq=92 timeout
timeout

X
loss

Seq=92 timeout

premature timeout
time time
lost ACK scenario
 TCP sender events:
Segment 1 Segment 2 Segment 3 Segment 4 Segment 5

… 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 …

Host A Host B

Cumulative ACK
timeout

X buffer
loss
Cumulative ACK

Restart timer do not buffer

timeout

Cumulative ACK
X
timeout

Duplicate ACK
time
TCP Round Trip Time, Timeout

Q: how to set TCP Q: how to estimate RTT?

timeout value?  SampleRTT: measured
 longer than RTT time from segment
 but RTT varies transmission until ACK receipt
 too short: premature  ignore retransmissions
timeout, unnecessary  SampleRTT will vary, want
retransmissions estimated RTT “smoother”
 too long: slow  average several recent
reaction to segment measurements, not just
loss current SampleRTT
 Summary
1-27

 Transport-layer services
 Multiplexing and demultiplexing

 Connectionless transport: UDP

 Best-effort data transfer
 Segment structure

 Connection-oriented transport: TCP

 Segment structure
 Reliable data transfer
 Any Question?
1-28