COMPUTER NETWORKS

CS610
Lecture-1
Hammad Khalid Khan
Course Answers the Basic Question:

How do COMPUTER NETWORKS and

INTERNET Operate?
 Textbook
 Title
Computer Networks and Internets, with Internet
Applications, 4th edition

 Author Douglas E. Comer

 Publisher Pearson Education
 Book website
http://www.netbook.cs.purdue.edu
 Reference Books

 “Computer Networks” by Andrew S. Tanenbaum, Prentice

Hall Publishers

 “Networking: A Top-Down Approach Featuring the Internet”

by James F. Kurose and Keith W. Ross
 Course Composition
 Four Main Parts

– Using and Building Internet Applications

(Chap. 1 - 3)
– Data Transmission (Chap. 4 - 6)
– Packet Transmission (Chap. 7-16)
– Interntworking (Chap. 17-27)
– Network Applications (Chap. 28-41)
 Part 1: Data Transmission

 Transmission Media

 Local Asynchronous Communication

(RS 232)

 Long Distance Communication

(Modems and Carriers)
 Part 2: Packet Transmission

 Packets and Frames

 Local Area Networks (LANs)

– Hardware Addressing
– Ethernet, WLAN and Bluetooth
– Extending LANs: Fiber Modems, Repeaters, Bridges,
Switches
 Part 2: Packet Transmission (Cont’d)

 WAN Technologies and Routing

 Network Ownership and Service Paradigm

 Protocols and Layering

 Part 3: Interntworking

 Concepts, Architecture and Protocols

 IP Addresses

 Binding Protocol Addresses

 Part 3: Interntworking (Cont’d)

 IP Datagrams and Datagram Forwarding

 IP Encapsulation, Fragmentation and Reassembly

 TCP/IP
 Part 4: Network Applications

 Client-Server Interaction

 The Socket Interface

 Example Applications
 Grade Distribution
 Assignments 15%

 Midterm Examination 35%

 GMDB 5%

 Final Examination 45%

COURSE INTRODUCTION
CHAPTER No. 1
INTRODUCTION
 Network and Internet

 Network
System for connecting computers using a single
transmission technology

 Internet
Set of networks connected by routers that are configured to
pass traffic among any computers attached to any network
in the set
 Classification of Networks

 By SIZE
 Local Area Network (LAN) versus
Wide Area Network (WAN)
 By Connectivity
 Point-to-Point versus Broadcast
 By Communication Medium
 By Mobility
 Fixed versus Mobile
 Networks in Daily Life

 Now, networks are an important part of everyday activities

 Business
 Home
 Government
 Education
 Mastering The Complexity

 Concentrate on understanding the concepts

 Learning the Networking Terminology

(summarized in a Glossary in Appendix 1)
 Complexity in Network Systems
 Many different technologies exist

 No single underlying theory or model

 Models are either so simplistic or so complex

 No simple or uniform terminology

 Review Lecture 1

 Course Introduction
 Network and Internet
 Classification of Networks
 Networks in Daily Life
 Complexity in Computer Networking
 Mastering the Complexity
COMPUTER NETWORKS

CS610
Lecture-2
Hammad Khalid Khan
 CHAPTER No. 2
MOTIVATION and TOOLS
 Resource Sharing
 Sharing a peripheral device such as a printer or a disk

 Sharing a large-scale computational power, because

– Computers were extremely expensive

– The government budgets were not sufficient to provide
computers for all scientists and engineers
 Growth of the Internet

 U.S Department of Defence Advanced Research Projects

Agency (ARPA) started ARPANET project

 The early internet emerged in 1970s

 Has grown drastically since then

Growth of the Internet
Growth of the Internet
 Probing The Internet

 “PING” program
 Probing The Internet

 “PING” program
 Probing The Internet

 Problems with the ping

– “Ping” cannot tell the reason of the problem

– Some networks or computers reject the ping packets
– “Traceroute”
 Probing The Internet
 Tracing A Route
 Summary
 Resource Sharing

 Growth of the Internet

– Linear Scale
– Log Scale

 Tools for Probing the Internet

– PING
– TRACE ROUTE
COMPUTER NETWORKS

CS610
Lecture-3
Hammad Khalid Khan
 Types of Multiplexing

 Frequency Division Multiplexing (FDM)

 Wave Division Multiplexing (WDM0)
 Time Division Multiplexing (TDM)
 Review Lecture 2

 Resource Sharing

 Growth of the Internet

– Linear Scale
– Log Scale

 Tools for Probing the Internet

– PING
– TRACE ROUTE
 CHAPTER No. 4
TRANSMISSION MEDIA
 Basic Idea
 Encode data as energy and transmit energy

 Decode energy at destination back into data

 Energy can be electrical, light, radio, sound, ...

 Each form of energy has different properties and

requirements for transmission
 Transmission media
 Transmitted energy is carried through some sort of medium

 Transmitter encodes data as energy and transmits energy

through medium
– Requires special hardware for data encoding
– Requires hardware connection to transmission medium
 Transmission media
 Media can be copper, glass, air, ...
 Copper Wires
 Widely used because of low resistance to electric current,
inexpensive and easy to install
 Twisted pair wires
– Two insulated copper wires twisted together to reduce
Interference
– Examples: Telephone wiring (one pair).LAN wiring UTP
cable (4 pairs of copper wire)
 Types of Media
 GUIDED Media:
 Uses physical path
– Examples: copper wires, optical fiber, coaxial cables
etc.
 UNGUIDED Media:
 Uses no physical path
– Example: Radio waves
 Coaxial Cable
 Single wire surrounded by an insulation and a metal
shield against interference
 Example use: Cable TV wiring and early LAN wirings
 Optical Fibers

 Uses light to transport data

 Advantages of Optical Fibers
 Neither causes nor susceptible to electrical interference

 Can carry pulse much farther than copper

 Light can encode more information than electric

 One single fiber is sufficient to carry data

 Radio

 Used for public radio and TV broadcast

 Data transmission is said to operate at Radio Frequency

(RF)

 Transmission range depends on antenna size

 Satellites

 Satellites for long distance

transmission

 Satellite transponders receive,

amplify, and transmit the RF
signal back to the ground
 Geosynchronous Satellites

 Placed in an orbit (at 35,785 km or 22,223 miles above the

earth) that is exactly syncronized with the rotation of the
earth

 From the ground, it appears to stay at the same position at

all times
 Low Earth Orbit Satellites

 Rotates faster than the rotation of the earth at 200 to

400 miles above the earth

 They do not appear to remain stationary

 Low Earth Orbit Satellite Arrays is a set of satellites

communicating with each other and with the earth
surface
 Microwave

 Electromagnetic radiation beyond the frequency range

used for radio and TV

 It can be aimed in a single direction

 Can carry more information than low frequency RF

transmissions
 Infrared

 Remote controls used with TV etc uses IR

 Inexpensive wireless communication

 Limited to a small area

 Convenient for portable computers

 CHAPTER No. 5
LOCAL ASYNCHRONOUS
COMMUNICATION
(RS-232)
 Introduction
 Computers use binary digits (0s and 1s) (bits) to
represent data

 Computers communicate by sending bits through a

transmission medium (i.e. cable, air).

 This chapter describes how electric current is used to

send digital information across Short Distances
 The Need For Async Communication

 In asynchronous communication the sender and

receiver do not need to coordinate (synchronize)
before the transmission begins

 In an asynchronous system, the receiving hardware

must be ready to accept and interpret the signal
whenever it arrives
 Using Electric Current to Send Bits
 A small electric current is used to encode data
 Example: Negative voltage to represent a 0 bit, positive
voltage to represent a 1 bit.
 Standards For Communication
 Standard specifies both the timing of the signals and
the electrical details of voltage and current

 RS232-C (by EIA) is one of the most widely used

standard to transfer characters between a computer
and a device such as modem, keyboard etc.

 RS-232 defines a serial (bits are sent one after

another) asynchronous communication
 RS232-C Specifications

 Cable length < 50 feet (20mt)

 -15v (bit 1) , +15v (bit 0)

 One character => 7 or 8 data bits

 No delay between character bits

 When idle, puts -15v (bit 1) on the wire

 Each character start with a start bit(0), and end with a stop
bit(1)
 Baud Rate, Framing, And Errors
 Baud rate:
– The number of changes in the signal per second

 For simple RS-232, Baud rate = Number of bits/sec

 If the voltage measurements do not all agree or if the stop

bit does not occur exactly at the same time expected,
framing errors occur
 Full-Duplex Asynchronous Comm

 Electrical circuits require at least two wires, one of

them to carry the data signal, the other is used as
the return path, called ground

 Simultaneous transfers in two directions are

known as full-duplex transmission
Full-Duplex Asynchronous Comm
 CHAPTER No. 6
LONG DISTANCE COMMUNICATION
(Carriers, Modulation and Modems)
 Long Distance Communication

 Electric current becomes weaker as it travels over

copper

 Signal loss occurs when RS232 is attempted to use to

communicate to a remote site

 A continuous, oscillating signal, called a Carrier,

propagates farther than other signals
 Modulation
 To send data, a transmitter slightly modifies the original
information by using a carrier

 Such modifications are called “Modulation”

 The receiver
– monitors the incoming carrier,
– detects modulation,
– reconstructs the original data, and
– discards the carrier
 Types of Modulation

 Three of the modulation schemes are:

– Amplitude Modulation (AM)

– Frequency Modulation (FM)

– Phase Modulation (PM)

 Modem
 A Modulator is a hardware circuit that
– accepts a sequence of data bits
– applies modulation to a carrier wave according to the
bits
 A Demodulator is a hardware circuit that
– accepts a modulated carrier wave and
– recreates the sequence of data bits
 A Modem is a hardware that combines both modulator and
demodulator in a single device
Modem
 Multiplexing
 Q: How can two or more signals can be transmitted
simultaneously over a single wire ?

 A: By using different carrier frequencies

 Career Frequencies And
Multiplexing
 Frequency division multiplexing (FDM)

– Using multiple carrier frequencies to allow

independent signals, to travel through a medium
Wavelength Division Multiplexing (WDM)

 FDM concept applied to the optical transmission

systems (optical fibers) is known as Wave Division
Multiplexing
 Time Division Multiplexing (TDM)

 Sources sharing a medium “take turns” to send data

 Synchronous TDM gives each source an opportunity to

send data

 Statistical TDM does not give chance to a source that

does not have data to send
COMPUTER NETWORKS

CS610
Lecture-4
Hammad Khalid Khan
 CHAPTER No. 7
PACKETS, FRAMES
and
ERROR DETECTION
 Introduction

 Previous chapters describe how bits are transmitted

across a physical network

 This chapter introduces the concept of packets

 The Concept Of Packets
 The problems:
– A sender and a receiver need to coordinate to detect
transmission errors

– Individual connections between each pair of computers

is not possible
 The Concept Of Packets

 The solution:

– Shared network connections among many workstations.

 Reasons For Using Packets

 Resource Sharing

 To Detect The Errors

Use of Packets
 Problems with the Sharing

 Demand high – Many computers need to use the shared

networks
 Some applications have large data transfers – Holds the
network for a long time
 Some applications cannot wait
 Need mechanism for fairness
 Solution for Fairness
 Divide the data into small blocks/chunks called “PACKETS”

 Computers take turns to send one packet at a time over

the shared connection.

 Because each packet is small, no computer experiences a

long delay
Example
 Example
 WITHOUT PACKETS

– 5MB file transferred across network with 56Kbps

capacity will require 12 minutes:

5x106 bytes * 8 bits/ byte

3
= 11.9 minutes
60 secs/ minute * 56x10 bits/ second
– All other computers will be forced to wait 12 minutes
before initiating other transfers
 Example
 WITH PACKETS
– If file is broken into packets, other computers must only
wait until packet (not entire file) has been sent
– Suppose file is broken into 1000 byte packets

1000 bytes * 8 bits/ byte

3
=.143 seconds
56x 10 bits / second
– Now each packet takes less than .2 seconds to transmit
 Advantage
 Other computer must only wait .143 seconds before
beginning to transmit

 Note:
– If both files are both 5MB long, each now takes 24
minutes to transmit

– BUT if second file is only 10KB long, it will be transmitted

in only 2.8 seconds, while 5MB file still takes roughly 12
minutes
 Packets and TDM
 Dividing data into small packets allows Time-Division
Multiplexing

 Each packet leaves the source and is switched onto the

shared communication channel through a Multiplexer

 At the destination, the packet is switched through a

Demultiplexer to the destination
Packets and TDM
 Packets and Frames

 Packet is ``generic'' term that refers to a small block of

data
 Each hardware technology uses different packet
format
 Frame or hardware frame denotes a packet of a
specific format on a specific hardware technology
 Frame Format

 Need to define a standard format for data to indicate the

beginning and end of the frame

 Header and trailer used to “frame” the data

Frame Format
 Disadvantages of Framing
 Incurs extra overhead - soh and eot take time to transmit,
but carry no data

Advantages
 Accommodates transmission problems:
– Missing eot indicates sending computer crashed
– Missing soh indicates receiving computer missed
beginning of message
– Bad frame is discarded
 Data Stuffing

 Sometimes, the special characters (i.e. soh or eot) may

appear in the data

 Problem: soh and eot as part of data will be

misinterpreted as framing data

 Solution: Data stuffing

 Summary
 The Concept of Packets

 Why use Packets?

 Packets and Frames

 Data Stuffing
Frame Delimater
COMPUTER NETWORKS

CS610
Lecture-5
Hammad Khalid Khan
 Review Lecture 4

 The Concept of Packets

 Why use Packets?

 Packets and Frames

 Byte Stuffing
 Byte Stuffing
 Sometimes, the special characters (i.e. soh or eot) may
appear in the data

 Problem: soh and eot as part of data will be

misinterpreted as framing data

 Solution: Byte stuffing

 Byte Stuffing (cont’d)

 Byte stuffing translates each reserved byte into two

unreserved bytes

 For example, can use esc as prefix, followed by x for soh, y

for eot and z for esc
 Byte Stuffing (cont’d)

 The sender replaces each occurance of special characters

with a combination of two characters (i.e.esc x, esc y, esc z)

 The receiver then replaces each occurence of esc x, esc y,

esc z, by the corresponding single character
Byte Stuffing (cont’d)
Byte Stuffing (cont’d)
 Transmission Errors
 Interference or power surges may destroy data during
transmission
– Bits may be lost
– Bit values may be changed

 Frames include additional information to detect / correct

errors
– Inserted by the sender
– Checked by the receiver
 Error Detection & Correction

 Error detection - send additional information so that

incorrect data can be detected and rejected

 Error correction - send additional information so that

incorrect data can be corrected and accepted
 Parity Checking
 Parity refers to the number of bits set to 1 in the data
item

– Even parity - An even number of bits are 1

– Odd parity - An odd number of bits are 1
 Parity Checking
 A Parity Bit is an extra bit transmitted with a data item,
chose to give the resulting bits even or odd parity
– Even parity - data: 10010001, parity bit 1

– Odd parity - data: 10010111,

– parity bit 0
 Parity and Error Detection
 If noise or other interference introduces an error, one of the
bits in the data will be changed from a 1 to a 0 or from a 0 to
a1

 Parity of resulting bits will be wrong

– Original data and parity: 10010001+1 (even parity)
– Incorrect data: 10110001+1 (odd parity)
 Limitations of Parity Checking
 Parity can only detect errors that change an odd number of
bits
– Original data and parity: 10010001+1 (even parity)
– Incorrect data: 10110011+1 (even parity!)

 Parity usually used to catch one-bit errors

 Alternative Error Detection Schemes

 Alternative error detection mechanisms have been

introduced
 Differences among mechanisms:
– The size of the additional information (transmission
overhead)
– Computational complexity of the algorithm
(computational overhead)
– The number of bit error that can be detected (how well
errors are detected)
 Checksum
 Checksum is calculated before transmission
– Treats data as a sequence of integers
– Computes their arithmetic sum
– Carry bits are added into the final sum

 Calculated checksum is sent along with data

 Receiver then performs the same calculation

Checksum
 Checksum
 Can be 8-, 16- or 32-bit integers
 Easy to do - uses only addition
 May not detect all errors
 Cyclic Redundancy Check (CRC)

 Problem : How can a network system detect more errors

without increasing the amount of information in each
packet?

 Answer: Cyclic Redundancy Check (CRC)

 Cyclic Redundancy Check (CRC)
 Consider data in message as coefficients of a polynomial

 Divide that coefficient set by a known polynomial

 Transmit remainder as CRC

– Good error detection properties
– Easy to implement in hardware
 Hardware Components

 Uses two hardware components:

– Shift Register
– Exclusive or (xor) unit
 Hardware Components

 Exclusive OR (XOR) unit

 Hardware Components

 Shift Register: Performs two operations:

– Initialize : Sets all bits to zero
– Shift: Moves all bits to the left one position
 SUMMARY
 Byte Stuffing
 Transmission Errors
 Error Detection
– Parity Check
– Checksum
– CRC
COMPUTER NETWORKS

CS610
Lecture-6
Hammad Khalid Khan
 Review of lecture 5

 Byte Stuffing
 Transmission Errors
 Error Detection
– Parity Check
– Checksum
– CRC
 Chapter 7 (cont’d) Combining Building
Blocks
 Example 16-bit CRC hardware that uses three shift
registers and three exclusive or (xor) units
Combining Building Blocks (cont’d)
 Combining Building Blocks (cont’d)
-Input data is all 1s
-CRC shown after 15,16,17 bits shifted
-Feedback introduces 0s in CRC
DIRECT POINT-TO-POINT
COMMUNICATION
 Types of Errors
▪ CRCs can detect the following errors better than
checksums

▪ Vertical errors: Occur due to a hardware failure e.g.

Damaged second bit of every character

▪ Burst errors: Changing a small set of bits near a

specific location, caused by lightening, electric motor
start, etc.
 Frame Format And Error Detection

▪ Modified frame format including CRC

▪ Error in frame typically causes receiver to discard frame

▪ Is Byte stuffing also required for CRC itself?

Frame Format And Error Detection
 CHAPTER No. 8
LAN TECHNOLOGIES
and
NETWORK TOPOLOGIES
 Introduction
 Most networks are Local and are designed to share
resources among multiple computers

 Hardware technologies used for Local networks allow

multiple devices to connect with a shared network

 The computers must take turns using the shared medium

 Direct Point-to-Point Communication

 Each communication channel connects exactly two

computers
 Form a MESH or a POINT-to-POINT network
 Direct Point-to-Point Communication
 Advantages:
– The connection type of individual connections can be
different

– Individual connection can choose a different frame

format, error detection mechanisms, etc

– Easy to enforce Security and Privacy

 Direct Point-to-Point Communication
 Disadvantages:
– The number of connections grow more rapidly than the
number of computers

• For ‘N’ computers:

Connections = N² - N
2

– Most connections use the same physical path

– Expensive
 Direct Point-to-Point Communication
 Disadvantages:
– Adding a new computer to the network requires
‘N-1’ new connections
 Star Topology in Practice
 The Result is:
 Summary

 CRC Hardware Components

 Frame Format and CRC

 LAN technology and Network topology (Chapter 8)

 Direct Point-to-Point Communication

COMPUTER NETWORKS

CS610
Lecture-7
Hammad Khalid Khan
 Review Lecture 6

 CRC Hardware Components

 Frame Format and Error Detection

 LAN technology and Network Topology (Chapter 8)

 Direct Point-to-Point Communication

 Shared Communication Channels

 LANs developed in late 1960s and early 1970s

 Key idea
– Reduce number of connections by sharing connections
among many computers
 Shared Communication Channels

 Each LAN consists of a single shared medium

 The computers take turns using the medium

 Sharing a single medium over long distances is inefficient,

due to the long delays
 Growth of LAN Technologies
 LAN technologies reduce cost by reducing number of
connections

 But ... attached computers compete for use of shared

connection

 Local communication almost exclusively LAN

 Long distance almost exclusively point-to-point

 Significance Of LANs And Locality Of
Reference
▪ LANs are most popular form of computer networks

▪ LAN technologies are inexpensive

▪ The demand for LANs is related to the “Locality of

Reference” principle
 Locality Of Reference Principle

 Principle of Locality of Reference helps predict computer

communication patterns:

– Spatial (or physical) locality of reference

• Computers are likely to communicate with other
computers that are located nearby

– Temporal locality of reference

• Computers are likely to communicate with the same
computers repeatedly
 Locality Of Reference Principle

 Thus
– LANs are effective because of spatial locality of
reference,
and
– Temporal locality of reference may give insight into
which computers should be on a LAN
 LAN Topologies
 Networks may be classified by shape

 Three most popular:

– Star
– Ring
– Bus
 Star Topology
 All computers are attached to a central point
 Center of star is sometimes called a “Hub”
 Logical versus Physical Star
 Star Topology in Practice

 Previous diagram is idealized

 Ring Topology

 Computers connected in a closed loop

 First passes data to second, second passes data to third,

and so on

 Logical versus Physical Ring

Ring Topology
 Bus Topology

 All computers are attached to a single long cable

 Any computer can send data to any other computer

 Coordination required to decide which computer uses the

line at what time
Bus Topology
 Reason for Multiple Topologies
 Each topology has advantages and disadvantages

– In a Ring, it is easy to coordinate access, however

entire network is disabled if a cable cut occurs
– In a Star, only one computer is affected, when a cable
cut occurs
– In a Bus, needs fewer wires than a star, however
entire network is disabled when a cable cut occurs
 Example Bus Network: Ethernet
 Widely used LAN technology

– Invented at Xerox PARC (Palo Alto Research Center)

in 1970s
– Defined in a standard by Xerox, Intel and Digital
Equipment Corporation - DIX standard
– Standard now managed by IEEE (802.3) defines
formats, voltages, cable lengths, ...
 Example Bus Network: Ethernet
 Uses Bus Topology
– Single coax cable - the Ether
– Multiple computers connect

 One Ethernet cable is sometimes called a Segment

– Limited to 500 meters in length
– Minimum separation between connections is 3 meters
 Ethernet Speeds

 Originally 3Mbps

 Current standard is 10Mbps

 Fast Ethernet operates at 100Mbps

 Giga-Bit Ethernet
 Encoding used in Ethernet
 Manchester Encoding: Uses signal changes to encode data

– A change from positive voltage to 0 encodes a 0 bit

– A change from 0 voltage to positive voltage encodes a 1 bit
Encoding used in Ethernet
 Ethernet Operation

 One computer transmits at a time

 Signal propagates from transmitter in both directions along

length of segment
Ethernet Operation
 Carrier Sense Multiple Access Networks
(CSMA)

 No central control management when computers transmit

on Ether

 Ethernet employs CSMA to coordinate transmission

among multiple attached computers
 Summary
 Shared Communication Channel
 Locality of Reference Principle
 LAN Topologies
– Star
– Ring
– Bus
 Ethernet
COMPUTER NETWORKS

CS610
Lecture-8
Hammad Khalid Khan
 Review Lecture 7
 Shared Communication Channel
 Locality of Reference Principle
 LAN Topologies
– Star
– Ring
– Bus
 Ethernet
 Carrier Sense Multiple Access
 A coordination scheme that defines how to take turns
using a shared cable

 A computer listens to the cable (carrier sense)

– If the cable is idle then starts transmitting
– If the cable is in use, then waits

 If Simultaneous transmissions occur, the frames

interfere with each other called a Collision
 Collision Detection - CD

 Signals from two computers will interfere with each other

 Overlapping frames is called a Collision

– No harm to hardware
– Data from both frames is Garbled
 Ethernet CD

 Ethernet interfaces include hardware to detect

transmission
– Monitor outgoing signal
– Garbled signal is interpreted as a collision
 After collision is detected, computer stops transmitting
 So, Ethernet uses CSMA/CD to coordinate transmissions
 Recovery from Collision
 Computer that detects a collision sends special signal to
force all other interfaces to detect collision
 Computer then waits for ether to be idle before
transmitting
– If both computers wait same length of time, frames
will collide again
– Standard specifies maximum delay, and both
computers choose random delay less than maximum
 Recovery from Collision
 After waiting, computers use carrier sense to avoid
subsequent collision

– Computer with shorter delay will go first

– Other computers may transmit later
 Exponential Backoff
 Even with random delays, collisions may occur

 Especially likely with busy segments

 Computers double delay with each subsequent collision

 Reduces likelihood of sequence of collisions

 802.11 Wireless LANs And CSMA/CA

 Use radio signals at 2.4 GHz (IEEE 802.11)

 Data rate at 11Mbps
 Older devices use radio signals at 900MHz
 Data rate of 2Mbps
 Bluetooth specifies a wireless LAN for short distances
 Shared medium - radio instead of coax
 Limited Connectivity with Wireless

 In contrast with wired LAN, not all participants may be

able to reach each other
– Low signal strength
– Propagation blocked by walls, etc.

 Can't depend on CD; not all participants may hear

Limited Connectivity with Wireless
 CSMA/CA
 Wireless uses Collision Avoidance rather than
Collision Detection
– Transmitting computer sends very short message
to receiver
– Receiver responds with short message reserving
slot for transmitter

 Response from receiver is broadcast, so all potential

transmitters receive reservation
 Collisions
 Receiver may receive simultaneous requests
– Results in collision at receiver
– Both requests are lost
– Neither transmitter receives reservation; both use
backoff and retry

 Receiver may receive closely spaced requests

– Selects one
– Selected transmitter sends message
– Transmitter not selected uses backoff and retries
 Local Talk

 LAN technology invented by Apple that uses bus

topology
 Interface included with all Macintosh computers
 Relatively low speed - 230.4Kbps
 Low cost (``free'' with a Macintosh); easy to install and
connect
 Uses CSMA/CA
 Token Ring
 Many LAN technologies that use ring topology use token
passing for synchronized access to the ring

 Ring itself is treated as a single, shared communication

medium

 Bits pass from transmitter, past other computers and are

copied by destination
 Token Ring
 Hardware must be designed to pass token even if attached
computer is powered down
 Using the token

 When a computer wants to transmit, it waits for the token

 After transmission, computer transmits token on ring


 Next computer ready to transmit receives token and then
transmits
 Token and Synchronization

 Because there is only one token, only one computer will

transmit at a time

– Token is short, reserved frame that cannot appear in

data
– Hardware must regenerate token if lost
 Token and Synchronization

 Token gives computer permission to send one frame

– If all ready to transmit, enforces ``Round-Robin''

access
– If none ready to transmit, token circulates around
ring
 IBM Token Ring

 Very widely used

 Originally 4mbps, now 16Mbps

 Uses special connector cable between computer and ring

interface
 FDDI
 Fiber Distributed Data Interconnect (FDDI) is another ring
technology

– Uses fiber optics between stations

– Transmits data at 100Mbps

 Uses pairs of fibers to form two concentric rings

 FDDI and Reliability

 FDDI uses counter-rotating rings in which data flows

in opposite directions

 In case of fiber or station failure, remaining stations
loop back and reroute data through spare ring

 All stations automatically configure loop back by

monitoring data ring
FDDI and Reliability
 ATM – Star Network
 Asynchronous Transfer Mode technology consists of
electronic packet switches to which computers can
connect
 ATM switches form hub into which computers connect in
a star topology
 Computers get point-to-point connections - data from
transmitter is routed directly through hub switches to
destination
– If none ready to transmit, token circulates around ring
– If all ready to transmit, enforces ``Round-Robin''
access
ATM – Star Network
 ATM Details

 Transmits data at over 100Mbps

 Uses fiber optics to connect computer to switch

 Each connection includes two fibers

ATM Details
ATM Switches
 Summary

 CSMA/CD
 Wireless LANs and CSMA/CA
 Token Ring
 FDDI
 ATM
COMPUTER NETWORKS

CS610
Lecture-9
Hammad Khalid Khan
 Review Lecture 8

 CSMA/CD
 Wireless LANs and CSMA/CA
 Token Ring
 FDDI
 ATM
 Chapter 9

Hardware Addressing
and
Frame Type Identification
 Introduction
 Need to devise technique for delivering message through
LAN medium to single, specific destination computer

 Sending computer uses a Hardware Address to identify the
intended destination of a frame

 Sending computer also identifies Type of data carried in the
frame
 Specifying a Destination
 Data sent across a shared network reaches all attached
stations - for all LAN topologies

 Interface hardware detects delivery of frame and extracts

frame from medium

 But ... most applications want data to be delivered to One

specific application on another computer - not All computers
 Hardware Addressing

 Most network technologies have a hardware addressing

scheme that identifies stations on the network

 Each station is assigned a numeric Hardware Address or

Physical Address
 Hardware Addressing

 Sender includes hardware address in each transmitted

frame

 Only station identified in frame receives copy of frame

 Most LAN technologies include sender's hardware address

in frame, too
LAN Hardware
and
Packet Filtering
 LAN Interface
 LAN interface handles all details of frame transmission and
reception
 Adds hardware addresses, error detection codes, etc. to
outgoing frames

 May use DMA to copy frame data directly from main
memory

 Obeys access rules (e.g., CSMA/CD) when transmitting
 LAN Interface
– Checks error detection codes on incoming frames

– May use DMA to copy data directly into main memory

– Checks destination address on incoming frames

• Station's address, a copy of the frame is passed to the
attached computer

– Frames not addressed to the local computer are ignored

and don't affect the local computer in any way
 Format of Hardware Addresses

 Numeric value

 Size selected for specific network technology

 Length is one to six bytes

 Assigning Hardware Addresses

 Hardware addresses must be unique on a LAN

 How can those address be assigned and who is responsible

for uniqueness?
 Assigning Hardware Addresses

Static Hardwar e manufacturer assigns Manufacturer must ensure

permanent physical address to every interface has a
each network interface . unique address
Configurable Address can be set by end System administrators must
user, either manually (e.g., coordinate to avoid conflict
switches or jumpers on the
inte rface ) or electronically
(e.g., through software
Dynamic Interfac e automatically assigns Automatic scheme must be
physical address each time it is reliable to prevent conflicts
powered up
Broadcasting
 Problems with Broadcasting
 For every broadcast frame on the network, each computer
uses computational resources:
– Places contents into memory
– Interrupts the CPU
– Allows system software to make a decision whether to
discard or use the frame
 Problems with Broadcasting
– If a pair of computers, use broadcasting, instead of
sending them directly, all other computers waste CPU
time while discarding the frames.
 Multicasting
 Solution: The restricted form of broadcasting: Multicasting:
– It works like broadcasting
– However it does not forward frames automatically to the
CPU
– The interface hardware is programmed in advance to
accept certain frames that have multicast address as the
destination address
 Summary

 Hardware Addressing
 Specifying a Destination
 Packet Filtering
 Assigning Hardware Addresses
 Broadcasting
 Multicasting
COMPUTER NETWORKS

CS610
Lecture-10
Hammad Khalid Khan
 Review Lecture 9

 Hardware Addressing
 Specifying a Destination
 Packet Filtering
 Assigning Hardware Addresses
 Broadcasting
 Multicasting
 Problems with Broadcasting
 For every broadcast frame on the network, each computer
uses computational resources:
– Places contents into memory.
– Interrupts the CPU.
– Allows system software to make a decision whether to
discard or use the frame.
 Problems with Broadcasting
– If a pair of computers, use broadcasting, instead of
sending them directly, all other computers waste CPU
time while discarding the frames.
Multicasting
 Multicast Addressing
 If an application program wishes to receive certain frames,
then
– it programs the interface hardware to accept an
additional set of addresses.
 The interface hardware then begins accepting:
– those frames (multicast frames),
– broadcast frames and
– frames that are destined to the station itself.
 Multicast Addressing
 For example: Computers running an audio application

– can receive audio frames, if the interfaces are

programmed to receive them.
– The other computers that are not running that audio
application will not waste resources.
 Identifying Packet Contents

 Destination must get some clue about how to interpret frame

data.

 Can use:
– Explicit Frame Type - Identifying value included with frame
describes type of included data.
– Implicit Frame Type - Receiver must infer type from frame
data.
 Headers and Frame Formats
 LAN technology standards define frame format for each
technology.

 All contemporary standards use the following general format:

– Frame header has address and other identifying
information.
– Information typically in fields with fixed size and location.
– Data area may vary in size.
Headers and Frame Formats
Ethernet Frame Format
 Ethernet Frame Format
Field Purpose
Preamble Receiver synchronization

Dest. addr. Identifies intended receiver

Source addr. Hardware address of sender

Frame type Type of data carried in frame

Data Frame payload

CRC 32-bit CRC code

 Ethernet Fields

 Preamble and CRC often not shown

 Destination address of all 1s is the broadcast address.

 Special values are reserved for frame type field.

Ethernet Fields
 Frames without Type Fields
 Some LAN technologies do not include a type field.
 Sender and receiver can agree on interpretation:
– Agree on a single data format and use only that format.
– Limits LAN to one type of data.
– All computers on LAN must use one format.
– Agree to encode the data format in the first few bytes of the
data field.
 Encoding the Data Type
 Illustration of using data area to encode data type:
 Encoding the Data Type

 To ensure interoperability, format of encoding area must be

universally agreed upon.

 Format typically set by standards body.

 IEEE 802.2 LLC

 To ensure IEEE 802.2 standard includes Logical Link Control

(LLC) Sub Network Attachment Point (SNAP) header.

 SNAP/LLC format widely used; e.g., by Ethernet.

IEEE 802.2 LLC
 IEEE 802.2 LLC

 LLC portion indicates SNAP field to follow.

 OUI (Organizationally Unique Identifier) identifies Ethernet

specification organization.

 TYPE field interpreted as in Ethernet (in this case, IP).
 Unknown Types
 For either encoding format, some computers may not be
prepared to accept frames of some types like:

– Protocol type not installed

– Newly defined type

 Receiving computer examines type field and discards any

frames with unknown type.
 Network Analyzers

 A network analyzer or network monitor or Network Sniffer is

used to examine the performance of or debug a network.

 Can report statistics such as capacity utilization, distribution of

frame size, collision rate or token circulation time.
 Operation of Network Analyzer
 Basic idea is a computer with a network interface that
receives all frames.
 Sometimes called Promiscuous Mode.
 Many desktop computers have interface that can be
configured for promiscuous mode.

– Combined with software, computer can examine any frame
on LAN.
– Communication across a LAN is not guaranteed to be
private!
 Operation of Network Analyzer

 Computer receives and displays (but does not respond to)

frames on the LAN.
 Filtering Incoming Frames
 Analyzer can be configured to filter and process frames.
– Count frames of a specific type or size.
– Display only frames from or to specific computers.
– In general, can be configured to match value of any field
and capture only those frames meeting the filter
specification.
 Summary

 Multicasting
 Frame Type Identification
 Ethernet Frame Format
 Frame Without Type Fields
 Network Analyzer
COMPUTER NETWORKS

CS610
Lecture-11
Hammad Khalid Khan
 Review Lecture 10

Multicasting
Frame Type Identification
Ethernet Frame Format
Frame Without Type Fields
Network Analyzer
 Chapter 10

LAN WIRING, PHYSICAL TOPOLOGY, and

INTERFACE HARDWARE
 Introduction
 Interface cards:
– Why a separate card?
– How to connect the interface to the computer?
– What is a Transceiver?

 LAN Wiring Schemes

 Logical and Physical Topology

 Speeds of LANs and Computers

 LAN data transmission speeds are typically Fast relative to

CPU speeds.

 LAN speeds are defined independent of any specific

processor speeds.
– Allows for mix of attached systems.
– New computers can be attached without affecting LAN
speeds.
 Network Interface Hardware
 CPU can't process data at network speeds.

 Computer systems use special purpose hardware for network

connection.
– Typically a separate card in the backplane.
– Network adapter card or network interface card (NIC).

 Connector at back of computer then accepts cable to physical

network.
I/O Interfaces
Network Connector
Network Interface Card (NIC)
Network Interface Card (NIC)
 NICs and Network Hardware
 NIC is built for one kind of physical network.
– Ethernet interface can't be used with token ring.
– ATM interface can't be used with FDDI.

 Some NICs can be used with different, similar hardware.

– Thick, thin and 10Base-T Ethernet.
– 10Mbps and 100Mbps Ethernet.
 NIC and CPU Processing
 NIC contains sufficient hardware to process data independent
of system CPU.
– Some NICs contain separate microprocessor.
– Includes analog circuitry, interface to system bus, buffering
and processing.

 Looks like any other I/O device to system CPU.

– System CPU forms message request.
– Sends instructions to NIC to transmit data.
– Receives interrupt on arrival of incoming data.
 Connection between NIC and Physical
Network
 Two alternatives:
– NIC contains all circuitry and connects directly to network
medium.
– Cable from NIC connects to additional circuitry that then
attaches to the network medium.

 Thin Ethernet vs. 10Base-T:

– Both are Ethernet; network technology not limited to one
style of connection.
 Thick Ethernet Wiring
 Uses thick coax cable.
 AUI cable (or transceiver or drop cable connects from NIC to
transceiver.
 AUI cable carries digital signal from NIC to transceiver.
 Transceiver generates analog signal on coax.
 Wires in AUI cable carry digital signals, power and other
control signals.
 Thick Ethernet also requires Termination to avoid signal
reflectance.
Thick Ethernet Example
Thick Ethernet
(10 base 5)
Thin Ethernet
(10 base 2)
 Connection Multiplexing

 In some circumstances, transceivers may be inconvenient;

e.g., workstations in a lab.

 Connection Multiplexer connects multiple computers to a

single transceiver.
– Each computer's AUI cable connects to connection
multiplexor.
– One AUI from multiplexor to Ethernet coax.
Connection Multiplexing
 Thin Ethernet Wiring

 Uses thin coax that is cheaper and easier to install than thick
Ethernet coax.

 Transceiver electronics built into NIC; NIC connects directly to

network medium.

 Coax cable uses BNC connector.

 Thin Ethernet Wiring

 Coax runs directly to back of each connected computer.

 T-Connector attaches directly to NIC.

Thin Ethernet Wiring
Thin Ethernet
Thin Ethernet
 Thin Ethernet

 Useful when many computers are located close to each other.

 May be unreliable - any disconnection disrupts entire net.

Thin Ethernet
 10Base-T
 Commonly called 10Base-T, Twisted Pair or TP Ethernet.

 Replaces AUI cable with twisted pair cable.

 Replaces thick coax with Hub.

10Base-T
10Base-T
10Base-T
 Hubs

 Extension of connection multiplexing concept.

 Sometimes called “Ethernet-in-a-Box”.

 Effectively a very short Ethernet with very long AUI cables.

 Can be connected into larger Ethernets.

Hubs
Hubs
 Protocol Software and Ethernet Wiring
 All wiring technologies use identical Ethernet specification.
– Same frame format.
– Same CSMA/CD algorithms.
 Can mix different technologies in one Ethernet.
 NICs can provide all three connection technologies.
 Protocol software can't differentiate among wiring
technologies.
Protocol Software and Ethernet Wiring
 Comparison of Wiring Schemes
 Separate transceiver allows computer to be powered off or
disconnected from network without disrupting other
communication.

 Transceiver may be located in an inconvenient place.

 Finding malfunctioning transceiver can be hard.

 Comparison of Wiring Schemes
 Thin coax takes minimum of cable.

 Disconnecting one computer (or one loose connection) can disrupt entire network.

 Hub wiring centralizes electronics and connections, making management easier.

 Bottom line - 10Base-T most popular because of cost.

Comparison of Wiring Schemes
Topologies and Network Technologies

 10Base-T network topology is a bus; wiring topology is a star.

 Token ring network topology is a ring; wiring topology is a star.

 Remember to distinguish between Logical and Physical Topologies.

 Filtering Incoming Frames
 Analyzer can be configured to filter and process frames.
– Count frames of a specific type or size.
– Display only frames from or to specific computers.
– In general, can be configured to match value of any field and capture only those
frames meeting the filter specification.
 Summary
 Network Interface Hardware
 Connection Between NIC and Physical Network
 Thick Ethernet ( 10 BASE 5 )
 Thin Ethernet ( 10 BASE 2 )
 10BASE-T
 Comparison Of Wiring Schemes
COMPUTER NETWORKS

CS610
Lecture-12
Hammad Khalid Khan
 Review Lecture 11

 Network Interface Hardware

 Connection Between NIC and Physical Network
 Thick Ethernet ( 10 BASE 5 )
 Thin Ethernet ( 10 BASE 2 )
 10BASE-T
 Comparison Of Wiring Schemes
 10Base-T
 Formally known as 10BaseT

 Does not use coaxial cable, uses twisted pair wiring with RJ-
45 connectors

 Each computer has a dedicated connection to a central

electronic device called an Ethernet Hub
10Base-T
10Base-T
 Hubs

 Extension of connection multiplexing concept


 Sometimes called “Ethernet-in-a-Box”

 Electronic components in a Hub simulate a shared physical

cable, making the entire system operate like a conventional
Ethernet
Hubs
Hubs
 Advantages and Disadvantages of Wiring
Schemes
 Each of the three wiring schemes has advantages and
disadvantages:
– Reliability issues:
• Wiring that uses a transceiver for each connection does
not affect the entire network if a transceiver cable is
disconnected
• A cable cut occuring in a Hub wiring only affects one
computer.
 Advantages and Disadvantages of Wiring
Schemes
– Cost issues:
• Twisted pair Ethernet is the cheapest wiring that makes
it so popular.
• Thicknet is the most costly wiring which is no longer
used
 Example
– Eight offices wired with a) Thick Ethernet, b)Thin Ethernet,
c) Twisted Pair Ethernet
Comparison of Wiring Schemes
 The Topology Paradox
 Question: Is Ethernet a bus topology, or does the topology
depend on the wiring?

 Answer: Both. Twisted pair Ethernet forms a classic star

topology, however functions like a bus

 10BaseT Ethernet is often called a star-shaped bus

 The Topology Paradox

 Two different types of topologies:

– Logical Topology: defined by the specific network

technology

– Physical Topology: depends on the wiring scheme

 Network Interface Cards and Wiring
Schemes

 To allow changing the wiring

without changing the interface
hardware, NICs support multiple
wiring schemes
Network Interface Cards and Wiring
Schemes
 10/100 Network Interfaces and
Autonegotiation
 10 Base-T version of Twisted pair Ethernet operated at 10
Mbps
 100 Base-T Twisted pair Ethernet operates at 100 Mbps
 100 Base-T technology is backward compatible and allows
the participants to negotiate a speed when connection is
established
 Process is known as Autonegotiation
10/100 Ethernet NIC
 Categories of Wires

 Cable used for wiring should match:

– The intended data rate
– The distance between the devices
– The amount of electromagnetic noise
– Anticipated future needs
– COST
Categories of Wires
 Summary

 Wiring Schemes
– 10 Base-T
 Advantages and Disadvantages of Wiring Schemes
 The Topology Paradox
 NICs and Wiring Scheme
 Categories of Wires
COMPUTER NETWORKS

CS610
Lecture-13
Hammad Khalid Khan
 Review Lecture 12

 Wiring Schemes
– 10 Base-T
 Advantages and Disadvantages of Wiring Schemes
 The Topology Paradox
 NICs and Wiring Scheme
 Categories of Wires
 Wiring Schemes and Other Network
Technologies
 Multiple wiring schemes are not limited to Ethernet
technology. Almost all other network technologies use
different wiring schemes

 Localtalk uses Hubs (Physical Star) to simulate a bus

topology

 IBMs Token Ring also uses Hubs (Physically a Star

Topology) to simulate a logical ring network
Wiring Schemes and Other Network
Technologies
 Chapter No 11

Extending LANs: Fiber Modems, Repeaters, Bridges,

and Switches
 Introduction

 LAN technologies are designed to operate within the same

building

 However most companies or institutions have offices located

far apart from each other
 Distance Limitation And LAN Design

 The maximum cable length of a LAN is fixed, because:

– The electrical signal level gets weaker as it travels
– The delays must be short to allow access mechanisms
(CSMA/CD, token passing), work properly
Distance Limitation And LAN Design
 LAN Extensions
 Several techniques extend diameter of LAN medium

 Most techniques use additional hardware

 LAN signals relayed between LAN segments


 Resulting mixed technology stays within original engineering
constraints while spanning greater distance
Fiber Optic Extensions
 Fiber Optic Extensions
 The Fiber Modems:
– Convert digital data into pulses of light
– Transmit over the optical fiber
– Receive light and convert into digital data
 Because
– Delays on optical fiber is low and
– Bandwidth is high
– This mechanism will successfully extend the LAN across
several kilometers
Fiber Optic Extensions
Fiber Optic Extensions
 Repeaters
 Fact: Electrical signals gets weaker while travelling over
copper

 A repeater (a hardware device)

– Amplify the weakenning signals received from one
segment

– And retransmit them on to another segment

Repeaters
 Repeaters

 One repeater doubles, two repeaters triples the maximum

cable lenght limitation

 Computers attached to different segments communicate as if

they are connected to the same cable
 Repeaters
 Question: Can we increase the maximum cable lenght as
many times as we wish by just adding repeaters?
 Answer: No
– Every repeater introduces a delay
– Access mechanisms such as CSMA/CD does not work
with long delays
– Ethernet standard specifies that any two stations cannot be
seperated by more than four repeaters
 Repeaters
 Using a vertical
segment:
– Only two
repeaters
seperate any two
stations in this
scheme
 Repeaters
 Disadvantages:
– Repeaters do not recognize frame formats, they just
amplify and retransmit the electrical signals.
– If a collision or error occurs in one segment,
– Repeaters amplify and retransmits also the error onto
the other segments
An Ethernet Repeater (Front)
An Ethernet Repeater (Back)
 Bridges
 A Bridge is a hardware device also used to connect two LAN
segments to extend a LAN
 Unlike a repeater, a bridge:
– Uses two NICs, to connect two segments
– Listens to the all traffic
– Recognizes frame format
– Forwards only correct, complete frames
– Discards the collided and errored frames
An Ethernet Bridge
 Frame Filtering
 A typical bridge has two NICs, a CPU, a memory and a ROM

 It only runs the code stored in its ROM

 The most important task a bridge performs is frame filtering

 Frame Filtering
 If both the source and destination are on the same segment, it
does not forward the frame to the other segment

 A frame is forwarded to the other segment, if the frame is

destined to the that segment

 Broadcast and multicast frames are also forwarded

 Frame Filtering

 Question: How does a bridge know on which segment a

destination computer is attached?

 Answer: It keeps a list for each segment that consists of

physical addresses of the computers attached to that segment
 Frame Filtering
 Question: How does a bridge build up address lists?

 Answer: Most bridges are self learning bridges

– As soon as a frame arrives to a bridge
– It extracts the source address from its header and
– Automatically adds it in the list for that segment
Frame Filtering
 Summary
 Distance Limitation on LANs

 Fiber Optic Extensions

 Repeaters

 Bridges
COMPUTER NETWORKS

CS610
Lecture-14
Hammad Khalid Khan
 Review Lecture 13

 Distance Limitation on LANs

 Fiber Optic Extensions

 Repeaters

 Bridges
Bridges
 Frame Filtering
 Question: How does a bridge build up address lists?

 Answer: Most bridges are self learning bridges

– As soon as a frame arrives to a bridge
– It extracts the source address from its header and
– Automatically adds it in the list for that segment
Frame Filtering
 Startup And Steady State
 When a bridge first boots, the address lists are empty (in
startup state)

– The bridge forwards frames to the other segment if it can

not find its destination address in its lists

 After sometime, when the bridge has received at least one

frame from every computer, it has the lists built (steady
state), and

– It forwards frames as far as it is necessary

 Planning a Bridged Network
 In a steady state, a bridge allows simultaneous use of each
segment

 When designing a LAN, bridges can be installed to divide the

LAN into segments to improve performance

 For example: Frequently contacting computers can be
attached to the same segment

 The frame traffic on one segment does not affect the other
segment
 Bridging Between Buildings

 If two buildings are located far from each other, a bridge, a

pair of fiber modems and an optical fiber can be used to
connect two LANs
 Bridging Between Buildings

 Advantages:
– Cost: An optical fiber and fiber modems are sufficient to
connect many computers located on seperate buildings

– Maintanence: No need to change the wiring between

the buildings when installing and removing a computer

– Performance: The traffic on each building does not

affect the other
 Bridging Across Longer Distances

 Question: Is it always possible to connect two sites with

optical fiber?
 Answer: No
– The distance may be too long
– It is usually not allowed to lay an optical fiber if the land
does not belong to you
 Bridging Across Longer Distances

 There are two common methods to connect two distant sites:

– Leased serial line connection

– Less expensive

– Leased satellite channel

– Can span arbitrarily long distance
 Bridging Across Longer Distances

 Leased Satellite Connection:

 Unlike optical fibers, satellite connections are low-bandwidth
to save cost

 Because the frames arrive from local network much faster

than they can be sent across a satellite link,
– The bridge must use buffering(saving a copy of frame into
memory until it can be sent)
– The bridge may run out of memory
– The communication software usually waits for a response
after sending a few frames
 A Cycle of Bridges

 A bridged network can connect many segments

 One bridge is needed to connect each segment to the rest

of the bridged network
A Cycle of Bridges
 A Cycle of Bridges

 If the bridges on a large network form a cycle, then

broadcasting frames introduce a problem

 Copies of a broadcast frame continuosly flows around the

cycle, each computer receiving an infinite number of
copies
A Cycle of Bridges
Bridges
 Summary

 Frame Filtering

 Planning a Bridged Network

 A Cycle of Bridges

 Distributed Spanning Tree

COMPUTER NETWORKS

CS610
Lecture-15
Hammad Khalid Khan
 Review Lecture 14
 Start up and Stand by Modes of a Bridge

 Planning a Bridge

 Bridging between Buildings

 Bridging over Long Distance

 Cycle of Bridges and DST

 Switching

 A switched LAN consists of a single electronic device that

transfers frames among the connected computers

 A hub with multiple ports simulates a single shared medium

 However a switch simulates a bridged LAN with one

computer per segment
Switching
 Switching

 If a hub is used to connect many computers on a LAN, then

only two computers can communicate at a given time

 However if a switch is used, the communication between two

computers will not affect the others, pairs of computers can
communicate at the same time
 Combining Switches and Hubs

 To reduce costs, computers can be connected and

distributed to a number of hubs, then the hubs can be
connected to a switch

 Hubs simulate single shared LAN segments

 Switch simulates a bridged LAN connecting segments
 Bridging And Switching With Other
Technologies

 Hubs, bridges and switches are not limited to Ethernet

logical bus topology

 They are available also for other networking technologies

such as Token ring, FDDI etc.

 FDDI hub
 Token ring hub
Chapter 13

WAN Technologies
and
Routing
 Introduction

 LANs can be extended using techniques in previous chapter

 Can not be extended arbitrarily far or to handle arbitrarily

many computers
– Distance limitations even with extensions
– Broadcast a problem

 Need other technologies for larger networks

 Characterization of Networks
 Local Area Network (LAN)
– Single building

 Metropolitan Area Network (MAN)
– Single city

 Wide Area network (WAN)

– Country
– Continent
WAN Technology and Routing
 Packet Switches
 To span long distances or many computers, network must
replace shared medium with packet switches

– Each switch moves an entire packet from one connection
to another

– A small computer with network interfaces, memory and

program dedicated to packet switching function
Packet Switches
 Connections to Packet Switches

 Packets switches may connect to computers and to other

packet switches

 Typically high speed connections to other packets switches,

lower speed to computers

 Technology details depend on desired speed

 Packet Switches as Building Blocks
 Packet switches can be linked together to form WANs

 WANs need not be symmetric or have regular connections


 Each switch may connect to one or more other switches and
one or more computers
Packet Switches as Building Blocks
 Store and Forward
 Data delivery from one computer to another is accomplished
through store-and-forward technology
– Packet switch stores incoming packet
– ... and forwards the packet to another switch or computer
–
 Packet switch has internal memory
– Can hold packet if outgoing connection is busy
– Packets for each connection held on queue
 Physical Addressing in a WAN
 Similar to LAN
– Data transmitted in packets (equivalent to frames)
– Each packet has format with header
– Packet header includes destination and source addresses

 Many WANs use hierarchical addressing for efficiency

– One part of address identifies destination switch
– Other part of address identifies port on switch
Physical Addressing in a WAN
 Next-Hop Forwarding
 Packet switch must choose outgoing connection for
forwarding

– If destination is local computer, packet switch delivers
computer port

– If destination is attached another switch, this packet

switch forwards to next hop through connection to another
switch

 Choice based on destination address in packet

 Choosing Next Hop

 Packet switch doesn't keep complete information about all

possible destination

 Just keeps next hop

 So, for each packet, packet switch looks up destination in
table and forwards through connection to next hop
Choosing Next Hop
 Summary
 Wide Area Networks

 Packet Switches

 Physical Addressing in a WAN

 Store and Forward

 Next Hop Forwarding

COMPUTER NETWORKS

CS610
Lecture-16
Hammad Khalid Khan
 Review Lecture 15
 Wide Area Networks

 Packet Switches

 Physical Addressing in a WAN

 Store and Forward

 Next Hop Forwarding

 Choosing Next Hop

 Packet switch doesn't keep complete information about all

possible destination

 Just keeps next hop

 So, for each packet, packet switch looks up destination in

table and forwards through connection to next hop
Choosing Next Hop
 Source Independence
 Next hop to destination does not depend on source of packet,
called Source Independence

 Allows fast, efficient routing

 Packet switch need not have complete information, just next

hop
– Reduces total information
– Increases dynamic robustness - network can continue to
function even if topology changes without notifying entire
network
 Hierarchical Addressing and
Routing

The Process of forwarding the packets from one hop to

the next is known as Routing
Hierarchical Addressing and
Routing
 Hierarchical Addressing and
Routing
 In particular, all destinations on same switch have same next
hop
 Thus, routing table can be collapsed:
 Routing in a WAN
 More computers == more traffic
 Can add capacity to WAN by adding more links and packet
switches
 Packet switches need not have computers attached
 Interior switch - No attached computers
 Exterior switch - Attached computers
 Routing in a WAN
 Both Interior and Exterior switches:
– Forward Packets
– Need Routing Tables

 Must have:
– Universal Routing - Next hop for each possible
destination
– Optimal Routes - Next hop in table must be on shortest
path to destination

 Modeling a WAN

 Use a graph:
– Nodes model switches
– Edges model direct connections between switches

 Captures essence of network, ignoring attached computers

Modeling a WAN
 Route Computation with a Graph

 Can represent routing table with edges:

 Graph algorithms can be applied to find routes

 Redundant Routing Information
 Notice duplication of information in routing table for node 1:

 Switch 1 has only one outgoing connection; all traffic must

traverse that connection
 Default Routes

 Can collapse routing table entries with a Default Route

 If destination does not have an explicit routing table entry,

use the default route:
Default Routes
 Summary
 Source Independence

 Hierarchical Addressing and Routing

 Routing in a WAN

 Default Routes
COMPUTER NETWORKS

CS610
Lecture-17
Hammad Khalid Khan
 Review Lecture 16
 Source Independence

 Hierarchical Addressing and Routing

 Routing in a WAN

 Modeling a WAN

 Route Computation and Default Routes

 Building Routing Tables
 How to enter information into routing tables:
– Manual entry
– Software

 How to compute routing table information:

– Static routing - At boot time
– Dynamic routing - Allow automatic updates by a program
 Building Routing Tables

 Static Routing
– Simple
– Low Network Overhead
– Inflexible

 Dynamic Routing
– Can work around network failures automatically
 Computing Shortest Path in a Graph
 Assume graph representation of network at each node

 Use Djikstra's algorithm to compute shortest path from each

node to every other node

 Extract next-hop information from resulting path information

 Insert next-hop information into routing tables

 Weighted Graph
 Djikstra's algorithm can accommodate weights on edges in
graph

 Shortest path is then the path with lowest total weight (sum
of weights of all edges)

 Shortest path not necessarily fewest edges (or hops)

Weighted Graph
 Distance Metrics
 Weights on graph edges reflect "cost" of traversing edge
– Time
– Dollars
– Hop count (weight == 1)

 Resulting shortest path may not have fewest hops

 Distributed Route Computation

 Each packet switch computes its routing table locally and

sends messages to the neighbors

 Updates information periodically

 Network adapts if a link or a packet switch fails

 Packet switches modifies tables to avoid failed hardware
 Distance Vector Routing
 Local information is next-hop routing table and distance from
each switch

 Switches periodically broadcast topology information i.e.

(destination, distance)

 Other switches update routing table based on received

information
 Vector-Distance Algorithm

 In more detail:
Wait for next update message
Iterate through entries in message
If entry has shorter path to destination:
Insert source as next hop to destination
Record distance as distance from next hop to destination
PLUS distance from this switch to next hop
Link-State Routing (SPF)
 Comparison
 Distance Vector Routing
– Very simple to implement
– Packet Switch updates its own routing table first

– Used in RIP

 Link-State Algorithm
– Much more complex
– Switches perform independent computations
– Used in OSPF
 Examples-WAN Technology

 ARPANET
– Began in 1960s
– Funded by Advanced Research Projects Agency, an
organization of the US Defense Department
– Incubator for many of current ideas, algorithms and
internet technologies
 Examples-WAN Technology

 X.25
– Early standard for connection-oriented networking
– From ITU, which was originally CCITT
– Predates computer connections, used for
terminal/timesharing connection
 Examples-WAN Technology

 Frame Relay
– Telco service for delivering blocks of data
– Connection-based service; must contract with telco for
circuit between two endpoints
– Typically 56Kbps or 1.5Mbps; can run to 100Mbps
 Examples-WAN Technology

 SMDS - Switched Multi-megabit Data Service

– Also a Telco service
– Connectionless service; any SMDS station can send a
frame to any other station on the same SMDS "cloud"
– Typically 1.5-100Mbps
 Examples-WAN Technology

 ATM - Asynchronous Transfer Mode

– Designed as single technology for voice, video, data, ...
– Low jitter (variance in delivery time) and high capacity
– Uses fixed size, small cells - 48 octets data, 5 octets
header
– Can connect multiple ATM switches into a network
 Chapter 14

Connection Oriented Networking

&
ATM
 Summary (CH. 13)
 Packet Switch

 Next-Hope Forwarding

 Source Independence

 Hierarchical Addressing
 Introduction

 LANs and WANs can both connect multiple computers, but

they have different base technologies and meet different
goals

 ATM is a single technology that is designed to meet the

goals of both LANs and WANs

 ATM uses the concept of Connection-Oriented Networking

 Summary
 Routing Table Computation

 Shortest Path Computation in a Graph

 Dijkstra’s Algorithm

 Distributed Route Computation

 Distance Vector Routing
 Link State Routing

 Example WAN Technologies

COMPUTER NETWORKS

CS610
Lecture-18
Hammad Khalid Khan
 Review Lecture 17
 Routing Table Computation

 Shortest Path Computation in a Graph

– Dijkstra’s Algorithm

 Distributed Route Computation

– Distance Vector Routing
– Link State Routing

 Example WAN Technologies

 Chapter 14

CONNECTION ORIENTED NETWORKING

and
ATM
 Introduction

 LANs and WANs can both connect multiple computers, but

they have different base technologies and meet different
goals

 ATM is a single technology that is designed to meet the

goals of both LANs and WANs

 ATM uses the concept of Connection-Oriented Networking

 Asynchronous Transfer Mode

 Telephone companies (Telcos) introduced ATM to meet

several goals:
– Universal service – for all subscribers
– Support for all uses – voice, video, and data
– A single, unified infrastructure (no separate LANs &
WANs)
– Guaranteed service, when appropriate
– Support for low-cost devices
 Jitter
 Jitter is the term used for variance in transmission delays

 Jitter is significant for voice, video, and data

 Jitter in LANs can occur when a packet is delayed because

the network is busy
 Packet Sizes
 Large packets result in less overhead because a smaller
fraction of the packet is used for header information

 Optimum networks use 4KB packets or larger

 Large packets can’t easily be used for voice

– 8 bit samples (at 125 usec per sample) would require half
a second to fill a 4KB packet.
– Echo cancellation can only be used with low transmission
delays
 ATM Cells
 To meet its goals, ATM uses small, fixed-sized packets called
cells

 Each cell has 53 octets

 VCI/VPI fields identify the cell’s destination

 PRIO tell if cell can be discarded

 ATM Cells
 CRC checks the header’s bits only!

 ATM header is about 10 percent of the cell

 Ethernet can have overhead of only 1 %

 Engineers sometime call ATM overhead the Cell Tax

ATM Cells
 Connection-Oriented Service
 The connection-oriented service paradigm for networking is
similar to the manner in which telephones are used:
– A caller dials the number of the destination
– The telephone at the destination signals the arrival of a
Connection Request
– If the called person doesn’t answer, the caller gives up
after waiting for a Timeout
– If the called person does answer, then the connection is
established, and data is exchanged
 Connection-Oriented Service
 In data communications, a binary connection identifier is
given to each of the two parties to the connection to enable
identification of the connection
 Virtual Channels (or Circuits)
 Connections in ATM are called virtual channels (VC), or
virtual circuits (a term preferred by some)

 These are called virtual, since connections are formed in

ATM by storing values in memory locations (tables) in ATM
switches, as opposed to making actual electrical connections
 Virtual Channels (or Circuits)
 The VC is identified by a 24-bit value formed from:
– The VPI, or virtual path indicator (8 bits), which identifies a
particular path through the network; &
– The VCI, or virtual channel indicator (16 bits), which
identifies the channel in the virtual path is being used by
the connection

 Most frequently, the 24-bit pair is treated as just a single

connection identifier by computers
 Summary
 Connection-Oriented Networking and ATM
 Asynchronous Transfer Mode
 ATM Design and Cells
 Connection-Oriented Services
 Virtual Channels
COMPUTER NETWORKS

CS610
Lecture-19
Hammad Khalid Khan
 Review Lecture 18
 A single Global Technology

 ATM Design and Cells

 Connection Oriented Service

 VPI/VCI
 Labels and Label Switching

 An ATM network is built from interconnected ATM switches

 Attachment points on switches, or ports, can be connected to

computers or other ATM switches
 Labels and Label Switching

 As cells arrive at an ATM switch, their VPI/VCI is modified

using a forwarding table which gives the new VPI/VCI for the
next leg of the cell’s trip

 The forwarding table is essentially indexed by the incoming

cell’s VPI/VCI, and the contents yield the new VPI/VCI
 Label Rewriting
 The replacement of the incoming cell’s VPI/VCI with a
probably different VPI/VCI is called Rewriting

 ATM is thus called a Label Rewriting, or Label Switching

system

 Thus two computers with a connection through an ATM

network will likely have different VPI/VCI values for each end
of the connection!
Label Switching
Example
 Example
 The sending computer, S, uses a VPI/VCI of 3 and sends
a cell

 Switch A receives the cell and looks up 3, then rewrites the

VPI/VCI as 4, and sends the cell out of its port 2

 Switch B receives the cell and looks up 4, then rewrites

the VPI/VCI as 1, and sends the cell out of its port 2
 Example
 Switch C receives the cell and looks up 1, then rewrites
the VPI/VCI as 6, and sends the cell out of its port 2

 The receiving computer, R, receives the cell with a

VPI/VCI of 6, which is the value it is using for the
connection

 Forwarding tables in each switch must be coordinated to

define meaningful “paths” through the network
Permanent Virtual Circuits
 Permanent Virtual Circuits
 Provisioning requires two steps:

– Determine a complete path (that is, identify the switches

that will be used)

– Choose appropriate VCI/VPI for each step in the path,

and configure each adjacent pair of switches (easy, since
each switch rewrites the VCI/VPI)
 Switched Virtual Circuits
 Most networks offer dynamic connections which last for a
relatively short time

 To handle this, ATM can dynamically establish a Switched

Virtual Circuit (SVC), allow it last as long as necessary, and
then terminate it

 The terminology comes from the telcos, where switching

system normally refers to call switching
 Establishing an SVC
 Computer sends a connection request to the switch to which it
is attached

 Software in the switch finds a network path to the destination

and sends along the connection request

 Each pair of switches in the path communicate to choose a

VPI/VCI for their tables
 Establishing an SVC
 Once the connection is accepted by the destination, a
message is sent back to the originating computer to indicate
the SVC is ready

 If any switch, or the destination computer, does not agree to

setting up the VC, an error message is sent back and the SVC
is not established
 Signaling
 The term signaling is used to describe communication about
the network, as opposed to communication that just uses the
network

 A computer uses signaling, with reserved VCI/VPI values, to

communicate with a switch to establish a connection or send
other network control messages
 Signaling

 These connection requests and network control messages

are also sent in cells, and since the VPI/VCI used in these
cells are reserved, there is no confusion between data and
control cells
 Quality of Service (QoS)
 Networks are more frequently being designed to allow
specification of the quality of service required by users

 For example, a typical voice telephone call might give a QoS

parameter for througput of 64Kbps and delay less than 500
msec. A user sending video might require throughput of
2Mbps
 Providing Desired QoS
 QoS desires are specified at connection setup time, and are
never altered for the connection duration

 Switches along the path must reserve resources to

guarantee the QoS

 If a switch cannot provide the resources, it rejects the

connection request and an appropriate error message is
returned
 ATM QoS Specifications

 Constant Bit Rate (CBR): Used for uncompressed audio and

video, since these have predefined maximum data rates

 Variable Bit Rate (VBR): Used for compressed audio and

video, where the data rate depends on the level of
compression that can be achieved
 ATM QoS Specifications

 Available Bit Rate (ABR) is used for typical data applications

(where the data rate may be unknown and bursty), and
allows use of whatever bandwidth is available at a given time
 Summary
 Label switching and Label Rewriting

 Permanent Virtual Circuits

 Switched Virtual Circuits

 Quality Of Service
COMPUTER NETWORKS

CS610
Lecture-20
Hammad Khalid Khan
 Review Lecture 19
 Label Switching

 An Example ATM Network

 Permanent Virtual Circuits (PVC’s)

 Switched Virtual Circuits (SVC’s)

 Quality of Service (QoS)

 Cells vs Packets
 ATM designers chose cells over packets because:
– Cells aren’t variable length, and memory management for
them is simpler. Handling variable length packets leads
to memory fragmentation

– Variable length packets require hardware to

accommodate the largest possible packet, and thus to
detect the end of the packet. With cells, bits can just be
counted as they arrive
 Cells vs Packets

– The length of time required to send a variable length

packet is variable, and requires complicated interrupt
schemes to detect completion of transmission

– QoS can’t be guaranteed with variable length packets as

easily as it can with fixed length cells
 ATM Speed
 ATM designers also chose cells to meet the need for speed,
since it was designed to handle arbitrarily large numbers of
users, each of which could be willing to pay for high
throughput

 ATM is designed to work on fiber (but can be used with

twisted pair). A typical port on an ATM switch operates at
OC-3 speed (155 Mbps) or higher
 ATM Critique

 ATM is far more expensive than typical LAN hardware

 Connection setup time may be excessive for short
communications
 Cell tax consumes 10 % of network capacity
 QoS requirements might be unknown, leading to applications
picking values that are too high or too low
 ATM Critique

 Broadcast is inefficient, and has to be simulated by sending

the same message to each computer separately

 ATM, as a single universal networking technology, has only

minimal provision for interoperation with other technologies
 Chapter 15

Network Characteristics: Ownership, Service Paradigm

and Performance
 Network Ownership

 Private Network
– Owned by single organization or company

 Public Network
– Owned by common carrier e.g. Phone Company)
 Private Network

 Often LAN Technology

 Multiple LANs in a building or campus linked together

 Sometimes called Intranet

 Private Network Architecture

 Operates autonomously from other networks (e.g. Internet)

 Usually includes one or few closely managed external

connections

 May restrict access at connections

 Managing Private Networks

 Organization buys own equipment

 Hires staff to design, implement, maintain and upgrade

network

 Responsible for all network management

 Extending Private Networks

 Large organizations may have multiple buildings or

campuses

 Can only install cables on own property

 May contract for leased lines from common carrier

 Public Networks
 Operated by common carrier
 May be telephone company or other organization that builds
network out of leased lines
 Multiple organizations subscribe and connect
 Data transits public network to other organizations
 Advantages & Disadvantages
Private Public
 Advantages  Advantages

– The owner has complete – Flexibility

control over both the  Disadvantages
technical decision and
polices
 Disadvantages – No decision making
power equipment or
policies
– Expensive to install and
maintain
 Virtual Private Network
 VPN (Virtual Private Network) combines features of private
and public networks
– Limited to single organization
– Uses public network for connectivity

 Connections, sometimes called Tunnels, connect sites

– Each site sees tunnel as point-to-point link
– No access for other users of public network
 Service Paradigm
 Connection-Oriented
– Similar to telephone system; endpoints establish and
maintain a connection as long as they have data to
exchange

 Connectionless
– Similar to postal system; endpoint puts data to send into a
packet and hands to network for delivery
Public and Private Networks
 Summary

 Cells vs. Packets

 Label switching vs. Routing
 ATM Critique
 Network Characteristics (Chapter 15)
 Network Ownership
 Private and Public Networks
COMPUTER NETWORKS

CS610
Lecture-21
Hammad Khalid Khan
 Review Lecture 20
 Cells vs Packets

 Label Switching vs Routing

 ATM Critique

 Network Ownership
– Private vs Public Networks
 Virtual Private Network
 VPN (Virtual Private Network) combines features of private
and public networks.
– Limited to single organization
– Uses public network for connectivity

 Connections, sometimes called Tunnels, connect sites

– Each site sees tunnel as point-to-point link
– No access for other users of public network
 Guaranteeing Absolute Privacy
 In addition to restricting packets, VPN systems use
encryption to guarantee absolute privacy.

 Even if an outside does manage to obtain a copy of the

packet, the outside will be unable to interpret the contents.
 Service Paradigm
 At the lowest level, most networks transfer individual packets
of data and the network requires each packet to follow an
exact format dictates by the hardware.

 Other offer additional facilities the hide may of the details of

the packets.
– Interface Paradigms Or Service Paradigms
 Service Paradigm
 Connection-Oriented
– Similar to telephone system; endpoints establish and
maintain a connection as long as they have data to
exchange.

 Connectionless
– Similar to postal system; endpoint puts data to send into a
packet and hands to network for delivery.
 Connection-Oriented Service
 One endpoint requests connection from network.

 Other endpoint agrees to connection.

 Computers exchange data through connection.

 One endpoint request network to break connection when

transmission is complete.
 Continuous and Bursty Traffic
 Networks handling voice or video are engineered to accept
and deliver continuous data at fixed rate.

 Others are designed to handle burst traffic typical of computer

networks.

 Connection does not disappear when no data is sent.

 Simplex & Full Duplex Connections
 Some connection-oriented technologies provide full duplex
while other allow on simplex connection.

– To communicate using a simplex design a pair of

computers must establish two connection; one from
Computer A to Computer B and another from Computer B
to A.
 Connection Duration & Persistence
 Connections can be made on-demand or set up
permanently.
– Switched connection or switched virtual circuit.
– Permanent connection or provisioned virtual circuit.

 Permanent connections
– Originally hard-wired
– Now configured at system unit time.
 Connection Duration & Persistence

 Switched Connections
– Computer maintains permanent connection to network
– Network makes connection on demand

 Internal components are switches; network is a switched

data network.
 Service Guarantees
 Some connection-oriented networks provide guarantees
about the service that computer will receive.
 They may guarantee a throughput rate or maximum packet
loss rate.
 For example, ATM provides statistical guarantee about
performance.
 Stream or Message Interface
 Some connection-oriented networks provide stream
interfaces
– No boundaries are recorded; that receive may receive a
single block of 60 characters even though the send sends
three blocks of 20 characters.

 Others provide a message interface

– Delivers data in the same size chunks that the sender
transmitted.
 Connectionless Service
 No connection necessary

 Source of data adds destination information to data and

delivers to the network.

 Network delivers each data item individually.

 Interior and Exterior Service Paradigms

 A network providing one service paradigm to the attached

computers can use an entirely different service paradigm
internally.

 For Example: ARPANET

– Connection-Oriented Internally
– Connection-less externally
 Comparison
 Connection-Oriented
– Accounting easier.
– Application can learn of network problems immediately.

 Connectionless
– Less overhead.
– Easier to implement network.
Examples of Service Paradigms
 Addresses & Connection Identifiers
 Address is a complete, unique identifier.

 Connectionless delivery requires address on each packet.

 Connection-Oriented delivery can use a shorthand that

identifies the connection rather than the destination.
 Addresses & Connection Identifiers
 ATM - 160-bit address, 24-bit connection identifier
– Connection identifier includes:
• 8-bit virtual path identifier (VPI)
• 16-bit virtual circuit identifier (VCI)
– Connection identifier local to each computer.
– May be different in different parts of the ATM switch.
 Summary
 Virtual Private Networks

 Service Paradigm

 Connection-Oriented Service

 Connection-less Service
COMPUTER NETWORKS
CS610
Lecture-21
Hammad Khalid Khan
 Review Lecture 21
 Virtual Private Networks (VPNs)

 Guaranteeing Absolute Privacy

 Service Paradigm
– Connected-Oriented Service
– Connectionless Service

 Address and Connection Identifiers

Network Performance Characteristics

Delay
Throughput
 Delay
 Important quantitative property of networks.

 Delay is a measure of how long it takes for a bit of data to

travel across the network from one computer to the other.

 Measured in seconds or fractions of seconds (milliseconds).

 Maximum and Average Delay.

 Types of Delay
 Propagation delay - time to travel across medium.
 Switching delay - time for network. component (hub, bridge,
packet switch) to forward data.
 Access delay - time to get control of medium (CSMA/CD,
token).
 Queuing delay - time enqueued in packet switches.
 Throughput

 Throughput is a measure of the rate at which data can be

sent through the network.

 The throughput capability of the underlying hardware is

called bandwidth.
 Throughput
 Because each frame contain headers the effective
throughput is less than the hardware bandwidth.

 Networking professional often use the term speed as a

synonym for throughput.
 Relationship between Delay and
Throughput

 If a packet switch had a queue

of packets waiting when a new
packet arrives the new packet
D0
will be placed on the ail of the D=
queue and will need to wait (1-U)
while the switch forward the D = Effective Delay
previous packets.
D0 = Delay when the network
is idle
U = Value between 1 and 0
denoting the current
utilization
 Relationship between Delay and
Throughput
 Throughput and delay are not completely independent.
 As traffic in a computer network increase, delays increase;
a network that operates at close to 100% of its throughput
capacity experiences severe delay.
 Delay-Throughput Product

 Computed as delay time multiplied by effective throughput.

 Measures amount of data that can be present in the
network.
 In fast networks with long delay times, sending computer
can generate large amounts of data before destination
receives first bit.
 Jitter
 The amount of delay that a network introduces.

 A network with zero jitter takes exactly the same amount of

time to transfer each packet.

 A network with high jitter takes much longer to deliver some

packets than others.
 Summary

 Delay
 Throughput
 Relationship between Delay and Throughput
 Delay-Throughput Product
 Jitter