CS-260

ADVANCED NETWORKING
Table of Contents

CHAPTER 1 : FUNDAMENTALS OF NETWORKING I...................................1-1

1.1 Introduction................................................................................................................... 1-2

1.2 Functional Components of a Computer Network...................................................1-2
1.3 Major Tansmission Function.......................................................................................1-3
1.4 Signal Power.................................................................................................................... 1-4
1.5 Signalling Rate................................................................................................................ 1-4
1.6 Bit Rate Vs. Baud........................................................................................................... 1-6
1.7 Serial Transmission....................................................................................................... 1-6
1.8 Parallel Transmission.................................................................................................... 1-7
1.9 Serial Vs. Parallel Transmission.................................................................................1-7
1.10 Simplex, Half-Duplex and Full-Duplex Transmission............................................1-8
1.11 Asynchronous and Synchronous Tansmission..........................................................1-8
1.12 Asnchronous Vs Synchronous Tansmission..............................................................1-9
1.12.1 Asychronous Character Format.....................................................................1-9
1.12.2 Synchronous Message Format......................................................................1-10
1.13 Interfacing.................................................................................................................... 1-11
1.14 Review Questions......................................................................................................... 1-13

CHAPTER 2 : FUNDAMENTALS OF NETWORKING II..................................2-1

2.1 Modulation...................................................................................................................... 2-2

2.2 Digital Signals and Analog Systems.........................................................................2-4
2.2.1 Amplitude Shift Keying (ASK)......................................................................2-4
2.2.2 Frequency Shift Keying (FSK).......................................................................2-5
2.2.3 Plase Shift Keying (PSK).................................................................................2-5
2.3 The Role of a Modem in Data Communication........................................................2-6
2.4 Communications Resource Sharing............................................................................2-6
2.4.1 Multiplexing....................................................................................................... 2-7
2.4.2 Polling.................................................................................................................. 2-9
2.5 Contention..................................................................................................................... 2-11
2.6 Review Questions........................................................................................................ 2-12

CHAPTER 3 : TRANSMISSION MEDIA AND SWITCHING TECHNIQUES...........3-1

3.1 Transmission Media...................................................................................................... 3-2

3.1.1 Twisted Pair....................................................................................................... 3-2
3.1.2 Coaxial Cable...................................................................................................... 3-3
3.1.3 Wave Guides...................................................................................................... 3-4
 3.1.4 Optical Fibre..................................................................................................... 3-4
3.1.5 Terrestrial Microwave....................................................................................3-6
3.1.6 Satellite Microwave.........................................................................................3-7
3.2 Switching Techniques.................................................................................................. 3-8
3.2.1 Circuit Switching.............................................................................................. 3-9
3.2.2 Message Switching.........................................................................................3-10
3.2.3 Packet Switching............................................................................................. 3-11
3.3 Review Questions........................................................................................................ 3-14

CHAPTER 4 : NETWORK SECURITY......................................................4-1

4.1 Introduction................................................................................................................... 4-2

4.2 Network Faults.............................................................................................................. 4-2
4.2.1 Attenuation........................................................................................................ 4-2
4.2.2 Delay Distortion................................................................................................ 4-3
4.2.3 Noise.................................................................................................................... 4-3
4.3 Error Detection and Correction Techniques..........................................................4-3
4.3.1 Even / Odd Parity............................................................................................. 4-4
4.3.2 Longitudinal Redundancy Check (LRC) and Vertical Redundancy Check
(VRC).................................................................................................................... 4-5
4.4 Assuring Error-free data Communication...............................................................4-6
4.5 The Automatic Repeat / Request (ARQ) System.................................................4-6
4.6 Flow Control................................................................................................................... 4-7
4.6.1 ARQ Sliding Window Protocol.......................................................................4-7
4.7 Human Interference.................................................................................................... 4-8
4.7.1 Authentication.................................................................................................. 4-8
4.7.2 Encryption.......................................................................................................... 4-8
4.7.3 Data Compression........................................................................................... 4-10
4.8 Review Questions........................................................................................................ 4-10

CHAPTER 5 : OSI MODEL AND TCP / IP................................................5-1

5.1 Open Systems Interconnection.................................................................................5-2

5.2 Layered Architecture Concept..................................................................................5-3
5.3 Layers, Services and Functions.................................................................................5-4
5.4 Entities, Protocol and Interfaces.............................................................................5-5
5.5 OSI Reference Model................................................................................................. 5-6
5.5.1 Physical Layer.................................................................................................... 5-7
5.5.2 Data Link Layer................................................................................................. 5-8
5.5.3 Network Layer.................................................................................................. 5-9
 5.5.4 Transport Layer............................................................................................... 5-11
5.5.5 Session Layer............................................................................................................... 5-12
CHAPTER 1 : FUNDAMENTALS OF
NETWORKING I
Chapter Objectives

Upon completion of this chapter, you will be able to:

⇒ describe what is data communication;

⇒ describe the functional components of a data communication network;

⇒ describe major transmission functions;

⇒ differentiate between serial and parallel transmission;

⇒ differentiate between synchronous and asynchronous transmission;

⇒ describe bit rate and baud rate;

⇒ describe simplex, half duplex and full duplex transmission;

⇒ describe Interface standard RS-232C.

1.1 Introduction
A computer network is defined as an interconnection of network components
that allow movement of information between the users. The users may be
the end-users or service providers.

The types of information that are carried by a computer network has

significantly changed over the years. In the 1970's this information was
mostly voice with very small amount of data. This has changed in present
decade and will keep on changing in the years to come. Now the information
includes voice, text, data, video and images. Each of these types of
information places different requirements on the network. The speed at
which this information needs to be moved varies from 100 bits/sec (for
telemetry types of application) to 100 m bits/sec (for high resolution TV
application).

1.2 Functional Components of a Computer Network

Network components can be any piece of hardware or software that helps in

the movement of information. The figure below shows some of the possible
candidates of network components.

1-6
CS260 CHAPTER 1: FUNDAMENTALS OF NETWORKING I

Customer Network
premises components
components

Telephones Transmission
Data terminals Switching
Workstations Signalling
LAN Network Management

Figure 1-1
♦

Services provided by a data communication network are provided by two

major categories of components; customer premises component and the
network components. The customer premises components include all the
terminals and other communication equipment that reside in customer
premises. The most common one is, of course, the telephone set. It also
includes PC's and work stations etc. On the other hand network components
will provide connectivity functions so that the terminals can communicate
effectively. Network components include transmission links, switching nodes,
signalling links etc. Network components provide four major services and
they are:

♦ Transmission
♦ Switching

♦ Signalling
♦ Network Management

In this particular subject we will be concentrating on customer premises

equipment and the technology behind them. Some of the underlying
technologies behind transmission will be discussed in this and subsequent
chapters for a better understanding.

1.3 Major Transmission Function

The major transmission functions include the following :

♦ Transmitting / Receiving

• The basic function of transmission is to transmit information

over a media to the receiving side. An example will be a telephone set
which includes a transmitter and receiver.
♦ Modulation / Demodulation

• Modulation converts a signal from one form to another form,

suitable for transmission. Demodulation restores the signal back to
its original form. An example will be the use of modem to convert
digital signal to analog and vice-versa.

1-7
CS260 CHAPTER 1: FUNDAMENTALS OF NETWORKING I

Transmitting Intermediatede Receiving

terminal Medium vice Medium terminal

Transmitting Signal Receiving

Modulation Amplification Demodulation
Coding Synchronisation Decoding
Multiplexing Demultiplexing

Figure 1-2
♦ Coding / Decoding

• Coding is a technique that defines quantized signals into a

predetermined binary sequence for transmission on a digital system.
At the receiving end the signal is decoded. This technique is used for
better synchronisation and error detection.

♦ Multiplexing / Demultiplexing
• Multiplexing is a technique that enables a number of
communication channels to be combined and transmitted over a
common broadband channel. At the receiving end, demultiplexing of
the broadband channel separates and recovers the original channels.
The primary purpose of multiplexing is to make efficient use of
transmission facilities bandwidth capability to achieve a low
transmission cost.

♦ Signal amplification or regeneration

• As signals travel through the transmission media, the
strength of the signal decreases with distance because of losses.
These signals must be amplified (analog) or regenerated (digital) in
order to have acceptable levels at the receiving end.
♦ Synchronisation

• Synchronisation is required in a network for accurate

frequency and stable time signals to co-ordinate numerous analog and
digital transmission. In a digital system, synchronisation is especially
important.

1.4 Signal Power

Signal power is a very important parameter in a transmission system. When
signal travels through the transmission media, the strength of a signal
attenuates with distance. It is important to know the signal power at various
points along the transmission path. Each signal possesses a unique power
which is conventionally expressed in terms of watts, mw, uw, nw and pw. But
there are times that signal power at various points of a transmission network
are too small to be expressed in conventional units. The unit dBm becomes
handy.
♦ Expression of dBm

♦ from mw to dBb

1-8
CS260 CHAPTER 1: FUNDAMENTALS OF NETWORKING I

• P (dBm) = 10 log P(mw)

♦ from dBm to mw
• P (mw) = 10 P(dBm)/10

Example:
Express the transmitted and the received signal powers in dBm.
Students should be aware of the numerical distinction between the
conventional power unit and dBm.

Tanjong Pagar Dhoby Ghaut

1 watt 1 mw

Solution:
Transmitted power = 1 watt = 10 3

= 10log103 = 30dBm
Received power = 1 mw = 100 mw

= 10 log 100 = 0dBm

1.5 Signalling Rate

The rate of transmission in data communication systems is expressed in

terms of number of bits sent per second or the number of signalling
elements sent per second. Depending on the application, both measures are
important.

The signalling rate is expressed as the number of signalling elements per

second or baud. A 1600 baud transmission then, is one that sends 1600
signalling elements per second.

Amplitude

Time
1 second
Signaling rate = 6 baud

Figure 1-3

The definition of a signalling element will depend on the transmission scheme

used. In the example shown above, pulses are defined as signalling elements.
Different signalling elements within a transmission might have different

1-9
CS260 CHAPTER 1: FUNDAMENTALS OF NETWORKING I

duration. Typically baud is calculated based upon the signalling element of

the shortest length. Thus if the shortest signalling element last y seconds,
the band is 1/y.

1.6 Bit Rate Vs Baud

Baud and bit rates are not always the same. This is because of the reason
that a signalling element can carry more than a single bit of data. The
relationship between baud and bit rate is:

bit rate(in bits/sec) = (signalling rate in band) X (number of bits per signal
element).
The bit rate is given as the number of bits per second (bps). If each
signalling element carry one bit of data then bps = baud.
Example:

Voltage Level Information content

0 volts 00

2 volts 01

4 volts 10

6 volts 11

Then 01 11 00 01 10 11 00 will be transmitted as:

Amplitude

6
4
2
0

Time

.833 milliseconds

Signalling Rate = 1/(.833 X 10 -3) = 1,200 baud

1 - 10
CS260 CHAPTER 1: FUNDAMENTALS OF NETWORKING I

Bit Rate = 1,200 X 2 = 2400 bits per second

1.7 Serial Transmission

In serial transmission, data bits of a character are transmitted serially one

bit after another. Typically the least significant bit of a character is
transmitted first. Serial transmission is slow as the bits are sent one at a
time on a single path. Consequently to transmit a character which consists of
8 bits, 8 bit-times is required for transmission.

Signal 1 0 0 1 0 0 1 0 1 0 1 00 1 0 1 000 1 0 1 00 1

LSB LSB LSB

Figure 1-4
1.8 Parallel Transmission
In parallel transmission a dedicated path is allocated to each bit position in
the character. Consequently 8 paths need to be available to transmit a 8 bit
character. In the example shown below the most significant bit is
transmitted on the top path where as the least significant bit is transmitted
on the bottom path.

Receiver Transmitter
msb

lsb
clock

Figure 1-5

For parallel transmission an explicit clock signal is required which is critical.

It is a requirements that all bits of a character are transmitted at the same
time so that they can be correctly received by the receiver. The rate of
transmission is controlled by the clock. The clock signal informs the receiver
when it should record or sample all of its inputs simultaneously.

1 - 11
CS260 CHAPTER 1: FUNDAMENTALS OF NETWORKING I

One major problem in parallel transmission is skew. The transmitter to send

all bits of a character at the same time uses the clock signal. Over large
distances (above 100 feet), the bits can lose the relative timing relationship.
The arrival of the bits from the same character at different times at the
receiver is called skew. Skew occurs because the propagation delay is not
identical for all transmission paths. Increase in distance will cause increase
in skew implying a proportional relationship. Significant skew will cause the
bits to be received wrongly at the receiving end which will mean receiving
wrong character. One way to reduce skew is to reduce the transmission
distance. Recommended transmission distance ranges from fifty to seventy
five feet.

1.9 Serial Vs Parallel Transmission

Typical usage of parallel transmission is between central processing unit and
its peripherals such as disk drives, printers and tape drivers. Parallel nature
of connection is

1 - 12
CS260 CHAPTER 1: FUNDAMENTALS OF NETWORKING I

consistent with machines internal architecture and it provides high speed

data transmission. One disadvantages of parallel transmission is that it is
expensive because of the number of transmission paths. It also requires
transmission and receiving equipment to have 8-bit internal architecture. The
cost increases with the distance of transmission.

On the other hand serial transmission is typically used for long distance,
computer to computer communication. Serial connections can cover large
distances at a comparatively low cost and are relatively easy to build. Serial
communication has a low data throughput compared to its parallel
counterpart. Furthermore, there is an additional overhead due to
serial/parallel conversion that needs to be performed.
1.10 Simplex, Half -Duplex and Full -Duplex Transmission

Simplex transmission means that the flow of information is only in one

direction. For example an input device (card reader or remote sensor)
attached to a computer could be configured in such a way so that it can only
transmit and can never receive. On the other hand, output devices (printer
or monitor) can be configured to receive data only. Normally Simplex kind of
transmission is not used generally because of lack of hand shaking.

Half duplex (HDX) transmission means that the connection is bidirectional,

but information can only flow in one direction at a time. Two- way radios are
example of the half-duplex system.

Full Duplex (FDX) transmission means that both stations may transmit to
each other at the same time. Full duplex transmission usually requires two
separate transmission paths. People talking to each other over the phone is
an instance of full duplex communication.

1.11 Asynchronous and Synchronous Transmission

Serial transmission can be accomplished using either Synchronous or

Asynchronous techniques. In asynchronous transmission each character in a
message is transmitted as an individual entity irrespective of when the
previous character was transmitted. At least two framing bits are associated
with each character; one framing bit indicates the beginning of the
character (start bit) and at least one framing bit indicates the end of the
character (stop bit). The ratio of framing bits to data bits is high. At least
ten bits need to be transmitted in order to transmit an 8-bit character
accounting for 80% utilisation. Asynchronous transmission is commonly used
between terminals and computers. This is because the terminal users are
relatively slow in entering data.
♦ Asynchronous:

1 - 13
CS260 CHAPTER 1: FUNDAMENTALS OF NETWORKING I

1001101 0 1 0100 1 1 0

Framing lsb msb Framing

Figure 1-6

1 - 14
CS260 CHAPTER 1: FUNDAMENTALS OF NETWORKING I

♦ Synchronous:

1001101010100110

lsb msb

Figure 1-7
Synchronous transmission means that all characters in a message are
transmitted one after another contiguously. Framing characters indicate the
beginning and end of the message block. The ratio of framing characters to
data characters is low. Consequently line efficiency in excess of 99 percent
can be maintained. Synchronous transmission is commonly used for computer
communication. Computers can take advantage of the high line efficiency
since they typically maintain a high transmission speed and send blocks of
characters at a time.
1.12 Asynchronous Vs Synchronous Transmission

Asynchronous transmission has a lower cost and is easy to implement.

Synchronisation between the transmitter and receiver is relatively simple as
individual characters are sent independently with it's own framing
information. Asynchronous transmission is less efficient and the best
utilisation that can be achieved is 80%.

Synchronous transmission on the other hand is used for interconnection of

high speed devices. Because of the high speed, a method for automatic error
detection and correction is required. Synchronous transmission requires
sophisticated timing method. In addition to determining the beginning and
end of block of characters, the receiver must also determine the duration of
each bit. Both steps are required to successfully receive a block. Since
synchronous transmission speeds may be several million bits per second,
either an extremely accurate clock must be present at all transmitters or
receivers, or else a signal transmission technique is required that provides
timing information. The latter is the preferred solution.

Due to required automatic error-recovery procedure, timing consideration

and complex signalling techniques, Synchronous transmission is more
expensive than its Asynchronous counterpart.

1.12.1 Asynchronous Character Format

The diagram below shows the format of an asynchronous character.
Observe that the line state may be either high (MARK) or low
(SPACE). Historically, the presence of a voltage on the line indicates
the MARK state, and the absence of a voltage indicates a SPACE
state.

1 - 15
CS260 CHAPTER 1: FUNDAMENTALS OF NETWORKING I

When the transmitter is not sending any signal, the line is idle. An
idle line is kept in the MARK state to distinguish lack of transmission
from a cut wire; a cut wire would lose power and present itself in the
SPACE state.

The asynchronous character, is preceded by a framing bit called the

START bit. The START bit is always in the SPACE status; thus, the
receiver knows that a character is arriving when it detects the
MARK-to-SPACE transition. This transition also synchronises the
receiver's clock.

The next "N" bit-times will be comprised off the data bits, where
there are "N" bits in each character. In practice, the most common
value for N is eight. Note that the LSB of the character is most
commonly sent first. If parity is used, it will be typically placed in
the most significant (MSB) position.

Start Bits Stop Interval Idle Line

Data Bits
MARK
SPACE

lsb
Figure 1-8

The character is followed by a STOP interval, which must last for at

least 1, 1.5 or 2 bit times. The STOP interval always returns the line
to the MARK state. This ensures that if two asynchronous
characters are sent contiguously, the MARK-to-SPACE transition will
still occur when the START bit presents itself. A long STOP interval
is required for those devices that need additional time to reset after
receiving a character. The transmitter and receiver must agree upon
a number of parameters in order to have successful communication.
These parameters include the number of bits per character, the
length of a bit time (i.e. the transmission speed), the number of
STOP bits, electrical signal levels, the order of bit transmission, and
the character code.

1.12.2 Synchronous Message Format

The following diagram shows the format of a generic synchronous
message. A synchronous message is typically referred to as a block
or a frame. A synchronous message is sent as a contiguous bit
stream, with no break between characters.

A synchronous message generally comprises the following fields:

1 - 16
CS260 CHAPTER 1: FUNDAMENTALS OF NETWORKING I

SYNC char Message Data Message Trailer

Header

Figure 1-9

1 - 17
CS260 CHAPTER 1: FUNDAMENTALS OF NETWORKING I

♦ Synchronous Characters
• These characters indicate to the receiver that a message is
about to be received. Further more, they define eight-bit blocks so
that the individual characters can be correctly interpreted

♦ Message Header
• Beginning-of-message framing, which may include the
sequence number of the block (or frame)
♦ Data

• Use information
♦ Message Trailer

• End-of-message framing, which may include error detection

information

♦ Successful synchronisation transmission will depend upon the

agreement of several parameters in asynchronous transmission. These
parameters include
• Length of a bit time (data rate)

• Signal levels
• SYNC characters

• Size and content of the message header and trailer

1.13 Interfacing

Digital devices generate digital signals and they are not capable of
transmitting this digital data to a longer distance. So typically digital devices
are not tied to the transmission facility directly. Computers and terminals
are termed as Data Terminal Equipment (DTE). The device which helps DTE
for data communication is known as Data circuit-terminating equipment (DCE)
and it sits between the DTE and the transmission media. Modem is one
example of DCE. On the transmission side, DCE receives data from DTE and
transmits data over the transmission media. On the receiver end, the DCE
receives data from transmission media and passes it to the DTE.

RS-232C

PC Modem Modem PC

DTE DCE DCE DTE

Figure 1-10

1 - 18
CS260 CHAPTER 1: FUNDAMENTALS OF NETWORKING I

In order to standardise the interface between DTE and DCE, the Electronics
Industries Association in the United States has come out with a interface
standard known as RS-232C. This RS-232 interface standard is described by
means of mechanical, electrical, functional and procedural specifications.
Mechanical specification deals with the physical connection of DTE and DCE
whereas the electrical specification deals with voltage levels and timing of
voltage changes. Functional specifications specify the functions that are
performed by different circuits. Procedural specification specifies the
protocol between DTE and DCE .

RS232-C

♦ RS232-C standard is specified in terms of the four specifications

mentioned above and they are:

• Mechanical Specification
- RS232-C is a 25 pin connector where the pins are organised into
two rows. Pins numbered 1-13 are in the top row where as the pins
numbered 14-25 are in the bottom row. All other mechanical
dimensions are specified in this specification.
• Electrical Specification

- In RS-232C standard, voltage greater than +4 volt is

- considered as logical 0 and voltage smaller than -3 volt is
considered as logical 1. The cable length between two RS-232C
connector can range a maximum of 15 meters and it can support a
data rate of 20 kbps.
• Functional Specification

- Although RS-232C is a 25 pins connector, only 9 of the pins are

always used in practice. Functional specification says that the
circuits connected to the pins can be grouped into four groups i.e.
data, control, ground and timing. Data groups include transmit and
receive signals which are connected to pin 2 and 3 respectively.
Whenever the PC wants to transmit data to the modem, it uses
the transmit line and it uses the receive line to get data from
modem. Control group include control signals like Request To Send
(RTS) and Clear To Send (CTS) lines which are connected to pin 4
and 5 respectively. PC uses RTS signal to inform the modem that
it wants to transmit data. The modem sends CTS signal to the PC
to inform that the modem is ready to receive data from the PC.
Timing group includes clock signals which are used for
synchronous transmission. Ground group is the set of circuits
used for grounding.
• Procedural Specification

1 - 19
CS260 CHAPTER 1: FUNDAMENTALS OF NETWORKING I

- The procedural specification for RS-232C specifies the sequence

of events that occur between DCE and DTE in order to transmit
(20) Data terminal ready
(8) Carrier Detect
(7) Signal Ground
Computer (6) Data Set Ready Modem
(5) Clear To Send
(4) Request To Send
(3) Receive
(2) Transmit
(1) Shield

data.
The sequence of events that occur between computer and modem during
communication are listed below:
1. When the computer is powered up, it sends a Data Terminal Ready
signal to the modem.
2. Whenever the modem powers up, it sends a Data Set Ready signal to
the PC.
♦ Transmit

3. Whenever the PC wants to transmits data, it sends Request To Send

signal to the modem.

4. In response, the modem transmit a carrier frequency on telephone

line and sends Clear To Send signal to the PC to indicate that it is ready
to receive data from the computer.
5. In response to DCE signal, the computer transmits data over the
transmit line.
♦ Receive

i) Whenever the modem detects a carrier frequency on the transmission line, it

means that the remote modem is transmitting. So, modem sends a carrier
detect signal to the computer.
ii) In response PC receives the incoming data over the Receive
line.

1.14 Review Questions

1. Describe in detail, the major functions in transmission.

2. What is parallel transmission? Under what condition will you choose
parallel transmission over serial transmission?
3. Describe simplex, half-duplex and full duplex transmission.

4. What is baud rate? What is the difference between baud and bit rate?

1 - 20
CS260 CHAPTER 1: FUNDAMENTALS OF NETWORKING I

5. Outline the difference between synchronous and asynchronous

transmission.
6. What will be the overhead of transmission if 7 bit ASCII character is
transmitted using asynchronous transmission. Assume 1 START, 1 STOP
and 1 parity bit.

7. Describe the hand shaking signals that are exchanged between a PC and a
modem during data transmission using RS-232C link.

8. What is RS-232C? Write down typical applications of RS-232C.

9. What are standards? Why do we need standard? List down a few of the
standard bodies and their functions.
10. What are the different functional components of a computer network?

1 - 21
CHAPTER 2: FUNDAMENTALS OF
NETWORKING II
Chapter Objectives
Upon completion of this chapter, you will be able to:

⇒ describe various modulation techniques;

⇒ describe the role of the modem in data communication;

⇒ describe communications resource sharing concepts such as multiplexing,

polling and contention.

2.1 Modulation

Modulation is a signal process which shifts the signal from one frequency
range to another so that the signal can be transmitted properly. The signal is
encoded into a carrier signal with frequency that is compatible with the
transmission medium being used.

Input Modulated
signal
X signal

carrier
signal

Figure 2-1

The input signal is called the modulating signal or baseband signal which can
be either analog or digital. The output signal is called the modulated signal.

♦ There are 3 types of modulation:

• Amplitude Modulation (AM)

- The amplitude of the carrier signal changes according to the

variation of the input signal.

Figure 2-2

2 - 22
CS260 CHAPTER 2: FUNDAMENTALS OF NETWORKING II

♦ Frequency Modulation(FM)

• The frequency of the carrier signal changes according to the

variation of the input signal

Figure 2-3

♦ Phase Modulation (PM)

• The phase of the carrier signal changes according to the

variation of the input signal.

Figure 2-4

2 - 23
CS260 CHAPTER 2: FUNDAMENTALS OF NETWORKING II

2.2 Digital Signals and Analog Systems

When digital signals are transmitted on an analog system, the discrete

elements are transmitted as continuous signals. Normally digital data is
transmitted over analog network using this technique. This type of data
typically requires the use of Modem (Modulator/Demodulator). The modem
modulates the digital data into an equivalent analog form at the transmission
end and at the receiving end it reconstructs the digital data from the analog
signal received.

In most of the applications, data is represented as a digital signal by the

DTE (Data Terminal Equipment e.g. PC). This digital signal is then supplied to
the modem for modulation or received from the modem as a demodulated
signal.

♦ A few of the modulation techniques are:

• Amplitude Shift Keying(ASK)

• Frequency Shift Keying(FSK)

• Phase Shift Keying(PSK)

2.2.1 Amplitude Shift Keying (ASK)

In amplitude shift keying, the digital input is converted to analog
signals where the amplitude of the signal represents the digital input.

Figure 2-5

In the diagram above two-level coding has been demonstrated where

the analog signal has two different amplitude levels representing
either a `1' or a `0'. In the figure above higher amplitude signal
represents a `1' and low amplitude represents a `0'. The frequency
and phase of the signal remains unchanged. ASK can be expanded to
carry additional bits by defining additional amplitude levels. Because
of the susceptibility to interference from noise, ASK techniques are
rarely used in modems.

2 - 24
CS260 CHAPTER 2: FUNDAMENTALS OF NETWORKING II

2.2.2 Frequency Shift Keying (FSK)

In frequency shift keying, the digital input is converted to analog

signal where the frequency of the analog signal represents the digital

Figure 2-6
input.

The above diagram shows two-level coding where the higher

frequency represents a `1' and lower frequency represents `0'. The
amplitude and phase of the analog signal remains unchanged. Similar
to ASK, FSK can be expanded to carry additional frequency levels.
FSK techniques are commonly used in inexpensive, low speed modems
(1200 bps and below).

2.2.3 Phase Shift Keying (PSK)

In phase shift keying, the amplitude and frequency of the signal
remains unchanged. At the beginning of a bit time, the phase of the

Figure 2-7
signal is changed.

PSK techniques is commonly used in modems.

2 - 25
CS260 CHAPTER 2: FUNDAMENTALS OF NETWORKING II

2.3 The Role of a Modem in Data Communication

The DTE (Data Terminal Equipment e.g. PC) generates or receives digital
signals. The modem transforms the digital signal into an equivalent analog
signal and send through the analog network (such as the analog Telephone
network). On the receiving end, the analog signal is received by the receiving
modem which transforms the signal into digital signals and delivers these to
the DTE.

The network supplies the user with a band-limited channel. For example the
filter in the analog telephone network supplies the user with a channel which
has a pass band of approximately 300Hz to 3300Hz. If we transmit high-
speed digital data through this channel, the output will get distorted because
of the removal of high frequency signal components. The modem however
converts the digital signal to an analog signal having only frequency
components that can pass through the passband of the channel. This allows
all the frequency components of the analog signal to pass through the analog
network undistorted.

2.4 Communications Resource Sharing

Resource Sharing in communications means allocation of a common high
capacity transmission facility to more than one user. Long distance
transmission facilities have a large bandwidth for communication and are
usually expensive. A single user generally does not require such a high
capacity for transmission and it also becomes quite expensive for a single
user to afford. For example, a twisted pair media has a bandwidth of 10MHz,
much more than the typical home computer can utilise. So we need to provide
a capability so that a number of low speed devices can share a single high-
speed channel. This technique is known as multiplexing which provides a cost
efficient way to allow K devices to share a single communications facility by
creating K channels over that facility. There are different multiplexing
schemes used in the analog and digital environments. We will briefly discuss
the analog techniques but the focus will be on digital techniques. We will also
take a look into broadcast network and two main schemes to achieve
resource sharing in those networks, i.e. polling and contention.

2 - 26
CS260 CHAPTER 2: FUNDAMENTALS OF NETWORKING II

2.4.1 Multiplexing

In multiplexing several low speed lines are combined into a high speed
line for long distance, high speed transmission. Generally there is a
point at which a long distance high-speed transmission facility is
cheaper than some number of lower speed lines whose combined
capacity is equivalent to the high speed link. So it is sensible to use
multiplexers in certain telecommunication networks.

Figure 2-8
[Link] Frequency Division Multiplexing (FDM)

In this multiplexing scheme, the total frequency of the

communication channel is divided among K users, where each user is
given a fixed portion of the frequency spectrum. In the figure below,
each of the four devices are assigned a free fixed frequency
channel. Some bandwidth between each channel is allocated to
provide a guard band. The guard band provides adequate separation
between the frequency allocation of each channel. TV and radio
stations share the frequency spectrum on the air using FDM.

[Link] Time Division Multiplexing

Figure 2-9

In digital environment, time on the channel is shared rather than the

frequency. This technique is known as Time Division
Multiplexing(TDM). In this scheme, a station gets access to all of the
channels bandwidth for some period of time. There are two types of
TDM scheme and they are:

2 - 27
CS260 CHAPTER 2: FUNDAMENTALS OF NETWORKING II

Synchronous TDM

In this type of TDM, every user is given a time slot on a scheduled

basis. In the example below, four stations are connected to a
multiplexer that samples each station in turn. Each station is
allocated a fixed time slot on a scheduled basis i.e. a station is
granted a time slice whether it has anything to transmit or not. The
composite line data will be at least equal to the sum of all of the
input lines.

Figure 2-10

Thus the composite link will have data rate of at least 4,800bps.
Synchronous TDM is a fair algorithm and easy to implement. How
ever this technique will be inefficient if the stations have low
utilisation due to the fact that many time slots will go unused. In the
case of bursty transmission, one station may have many transmission
queued up waiting for its time slot while all other devices are idle and
their time slots will go to waste.
Statistical Time Division Multiplexing

♦ In synchronous TDM, time slices may be wasted when a given station

has nothing to transmit during its time slice, thus making the scheme
inefficient. These unused time slices could be assigned to active stations
to increase the efficiency of line usage. Some header information must
now be included in the data stream to identify a particular station that
owns the time slice. This technique is called Statistical Time Division
multiplexing or statistical multiplexer and Statmux is the device used for
the purpose.

• Synchronous TDM
• Statistical TDM

2 - 28
CS260 CHAPTER 2: FUNDAMENTALS OF NETWORKING II

Idle time slots

T1 T2 T3 T1 T2 T3 T1 T2 T3 T1

Figure 2-12

T1 T2 T3 T1 T2 T3 T1 T2 T3 T1

Figure 2-11
The performance of a Statmux may be below that of an ordinary
time division multiplexer if all the attached stations use the line
frequently. However data transmission is typically bursty and so, in
the usual case, the statistical approach is superior.

Statmux assigns a time slot to a station only when the station has
something to transmit. Older statmuxes assigns a fixed time slot if a
time slot is needed where as newer statmuxes assign a variable
length time slot to user. The data rate of the composite link out
from a statmux may be less than the sum of all of the input data
rates. This is possible since each station will not usually be
transmitting at a rate that uses the entire available capacity of its
line to statmux. Consequently it is possible that twelve 1200bps
terminal could share a stamux that has a composite line rate of only
4800bps. Where as Synchronous TDM multiplexer will require a
composite link rate of 14,400bps. Statmuxes are more efficient than
synchronous TDM scheme although there are some overheads of
addressing. In synchronous TDM scheme, if there are K channels,
some station j owns every kth time slot. Consequently a time slot only
needs to contain data; the relative position of the slot in time implies
the station address. But this concept does not apply to Statmux.
Hence address must be explicitly included as part of the
transmission.

2.4.2 Polling

In broadcast type of network allocation of shared communication

facility is achieved by means of polling. In polling there is a central
controller which gives each of the stations control of the
communication facility according to the same strategy. However the

2 - 29
CS260 CHAPTER 2: FUNDAMENTALS OF NETWORKING II

bursty nature of most computer communications makes this

technique inappropriate. Most of the stations polled will have nothing
to transmit. Those stations who want to transmit will have a great
deal of data to transmit and be unhappy if they have to wait for the
entire polling cycle between transmission.

The design of a "local" broadcast network is an issue when more than

two machines share a link. This is called a multipoint, or multidrop,
environment. When multiple machines share the link, some algorithm
must exist so that the machine can all have orderly access to the
channel. There are two approaches to this - the balanced and
unbalanced approach.

2 - 30
CS260 CHAPTER 2: FUNDAMENTALS OF NETWORKING II

♦ Case 1

♦ Case 2

C C C C

Figure 13

Case 1 shows an unbalanced configuration, which is an example of

centralised polling. On an unbalanced link, one machine is designated
as the primary (P) station; all other stations are called secondary (S).
It is the primary’s responsibility to poll each secondary to determine
if the secondary wishes to transmit or is ready to receive. For the
case examined here, all communication on the link is between some
secondaries and the primary. Control of this system is centralised
since one station controls the access to the link for all other
stations. As an example, IBM's BISYNC is typically used in
unbalanced configuration.

Case 2 shows the balanced approach. On a balanced link, each station

is called a combined (C) station. The reason for using this term is
that all stations are peers and may communicate with each other,
thus having attributes in common with "primary" and "secondary"
devices. Control of this communication link is distributed since every
station has an equal role in determining which station gains access to
the line. All LANs use a balanced approach.

[Link] Distributed Polling: Token Passing

Token passing networks utilise two types of tokens. A busy token is
placed at the beginning of a data frame; user data immediately

2 - 31
CS260 CHAPTER 2: FUNDAMENTALS OF NETWORKING II

follows a busy token. When a station sees a free token, it is allowed

to change it to a busy token and transmit data. When the data
returns, the transmitter must reissue a free token. The key here is
that taps are active, rather than passive devices, and introduce a 1-
bit delay by removing and then regenerating a bit.

2 - 32
CS260 CHAPTER 2: FUNDAMENTALS OF NETWORKING II

The following diagram shows an example of changing a free token to a

busy token. In this example, when the station sees the incoming free
token, it can change it into a busy token merely by inverting the last
bit. IEE 802.4 token bus standard uses this polling mechanism.

Figure 2-14

2.5 Contention

Another approach for sharing communication facility in a broadcast

environment is Contention. In this scheme stations will transmit whenever
they want to. If a collision occurs, the stations have to back off and try
later again. Some mechanism must be provided to randomise the back off
period or else an infinite string of collisions will occur.

Most bus, or shared media networks control access via some form of
contention. When people talk to each other, they share a medium; namely the
air. When one person wants to speak, he or she speaks. If two or more people
want to speak, some algorithm is used so that only one does so. If more than
one person talks at the same time, "collisions" occur that destroy all of the
words. The algorithms that people use can not be easily adopted by
computer. People may defer to each other utilising politeness or seniority;
optionally, the loudest voice may win (utilising a rudeness protocol). In any
case, the people standing around talking to each other have more information
about their network than the stations on the bus.

2 - 33
CS260 CHAPTER 2: FUNDAMENTALS OF NETWORKING II

The following diagram illustrates a pure contention, a station starts to

transmit its frame as soon as it is ready. As shown, station 1 becomes ready
and transmits. Some time after 1 is finished, station 2 becomes ready and it
transmits. While station 2 is transmitting, station 3 becomes ready and it
too, starts to transmit. Since the end of station 2's frame collides with the
beginning of station 3's frame, both the frames are destroyed. The collision

Figure 2-15
scheme described here is called Pure Aloha, named for ALOHANET.

2.6 Review Questions

11. Why do we need to modulate a signal?

12. What is modulation? Describe different techniques used to modulate

an analog signal into a analog system.

13. What are the different modulation techniques used to modulate a

digital signal into an analog system?

14. What is the role of modem in data communication?

15. What is multiplexing? Describe different types of multiplexing
scheme.
16. When would you use Statmux in place of synchronous time division
multiplexing.
17. Describe how a communication facility is shared in broadcast type of
environment? Why do we need to share a communication facility?

2 - 34
CHAPTER 3 : TRANSMISSION MEDIA AND
SWITCHING TECHNIQUES

Chapter Objectives

Upon completion of this chapter, you will be able to:

⇒ describe the advantages and disadvantages of different transmission
media;

⇒ describe why a particular transmission media is used for a certain

application;

⇒ describe different types of switching networks;

⇒ describe how information flows through different nodes of a computer

network;

⇒ describe the advantages and disadvantages of different types of

switching networks.

3.1 Transmission Media

The nature of transmission media determines the quality and characteristics
of data transmission. For guided transmission media, the medium itself plays
an important role in determining the transmission limitation. However for
unguided media, the medium itself does not have a very big impact in
determining transmission characteristics. In this case the frequency band of
the signal produced by the transmitting device plays an important role. In
this chapter, we will take a look into different types of guided and unguided
media.

3.1.1 Twisted Pair

Twisted pair is composed of wood-pulp or plastic insulated wires

twisted together into pairs. In some cables, many twisted pairs are
stranded into a rope-like form called a binder group. Several binder
groups are, in turn, twisted together around a common axis to form
the cable core and a protective sheath is wrapped around the core.
Twisted pair is manufactured in a number of standard sizes and may
contain from six to 3,600 wire pairs. The electromagnetic
interference between the individual pairs are minimised by twisting
them together.

Twisted pair can be used to transmit both analog or digital signals.

Repeater distance for analog signals is 5-6 km where as for digital
signals, the repeater distance is 2-3 kms. Twisted pair is the least
expensive media for a number of reasons. It uses the minimum
amount of raw cable material, it is easy to make, and is readily
available. One of the problems of twisted pair is that it acts like an
antenna, it has high susceptibility to Electromagnetic Interference

3 - 35
CS260 CHAPTER 3: TRANSMISSION MEDIA AND SWITCHING TECHNIQUES

(EMI) and Radio Frequency Interference (RFI). Twisted pair has a

large bandwidth. A good estimate will be 250 KHz . At distances of
1.5 km, bit rates of 2.048 Mbps can be maintained; 56 kbps can be
maintained up to distances of 5 km. Unfortunately its error rate is
high and security of transmission is low. The main application of
twisted pair is in local loop (transmission link between user telephone
local exchange) and exchange areas.

3.1.2 Coaxial Cable

Coaxial cables contain a number of (typically from 4 to 24) coaxial
units called tubes. Each tube consists of an inner conductor and a
cylindrical outer conductor separated by insulating disks. In addition
to coaxial tubes, coaxial cable contains a small number of twisted pair
wire pairs and single pairs for maintenance and alarm functions.

Similar to twisted pair, coaxial cable can transmit both analog and
digital signals. Repeaters needs to be used every few kilometres for
a long distance analog signal transmission. In case of long distance
digital transmission, repeaters are required every kilometre or so.
Although coaxial cable provides shielding against EMI and RFI due to
the arrangement of its conductors, it still has low transmission
security. The shielding and cable structure adds to the cost of
material and manufacturing. The coaxial cable can yield data rates of
10 to 15 Mbps. Large diameter coaxial cable can yield larger data
rates. A good estimate for the bandwidth of a coaxial cable is 350
MHz .

Coaxial cable is enjoying increasing utilisation in a wide variety of

applications and it is perhaps the most versatile transmission media.
Coaxial cables are used primarily on intercity routes in the long-haul
network or in undersea cable systems. Other applications include
television transmission and distribution, Local area networks and
Short-run system links.

3 - 36
CS260 CHAPTER 3: TRANSMISSION MEDIA AND SWITCHING TECHNIQUES

Figure 3-1

3.1.3 Wave Guides

A wave guide is a rectangular or circular copper pipe that confines and guides
radio waves between two locations. It's main advantage is low attenuation at
microwave frequencies. The application of wave guide is limited because it
must be manufactured to extreme uniformity and extreme care is required
during installation to minimise sharp bends.

3.1.4 Optical Fibre

Due to the advantages of optical fibre technology such as wide bandwidth
interference immunity and so on, Fibre is obviously the transmission medium
of choice. The constantly growing need for large bandwidth services lead to
the advanced development of optical fibre system. Interference to optical

Figure 3-2
3 - 37
CS260 CHAPTER 3: TRANSMISSION MEDIA AND SWITCHING TECHNIQUES

fibre system is negligible, compared with other systems such as coaxial,

radio or satellite. A typical Optical Fibre link is shown below.

The signal source can be voice, data or video and the signal can be
analog or digital. In other words, an optical fibre system can be used
to transmit analog or digital signals. The driver is required because
optical devices, such as light source, require different voltages and
currents from the conventional electronic devices.

The structure of an optical fibre varies from one manufacturer to

Figure 3-3
another. But, the four-layer structure gradually became the industry
standard.

Functions of :
• Fibre core : signal transmission, that is, it serves as the
transmission medium.
• Fibre cladding :

- to guide the light so that it will travel within the fibre core
- to serve as a protection layer for the core.

• Protection layers
- Protect the fibre from extra forces.

- Protect from moisture

- Strengthen the fibre

- Prevent the penetration of outside material.

Optical fibre guides a beam of light (signal encoded) by means of
total internal reflection. Any transparent medium that has a higher
index of refraction than the surrounding medium can have total
internal reflection. Consequently optical fibre acts as a wave guide

3 - 38
CS260 CHAPTER 3: TRANSMISSION MEDIA AND SWITCHING TECHNIQUES

for frequencies. Depending on the size of the fibre, there are two
different kinds of optical fibre i.e. singlemode fibre and multimedia
fibre. For a multimedia fibre when light from a source enters the
plastic core, rays at shallow angles are reflected and propagated
along the fibre and other rays are absorbed by the surrounding
material. For multimode fibre, the maximum data rate that can be
achieved is limited. This is because multiple propagation paths exist
in a multimode transmission causing the signal elements to take
different paths and hence different time. This problem however
does not exist in a single mode fibre where the fibre core radius is
reduced to the order of a wavelength. Consequently only a single
angle or mode can pass, the axial ray. Since there is a single
transmission path, signal elements can not spread out in time like
multimode fibre. Singlemode fibre has superior performance than
multimode fibre and the data rate can reach up to 1.6 Gbps. Typically
standard multimode fibres have a core diameter of 50 um where as
single mode fibre has a core diameter of 5 um or less.

Figure 3-4

Signals travelling in optical fibre can be attenuated because of the

following reasons:
• The light coupling from the light source to the fibre is not
100%.
• The splicing of the fibre causes reflection of light and
misalignment of fibres.
♦ The signal will alternate due to the splicing.

• Glass absorbs light

• External / geometric effects of the following can cause
signal attenuation
• Absorption by impurity or OH ions

• Microbending from surface distortion

3 - 39
CS260 CHAPTER 3: TRANSMISSION MEDIA AND SWITCHING TECHNIQUES

• Diameter variation.

• Bubble penetrating the core.

There are three types of light source that is used for optical fibre
transmission and they are
i. Light emitting diode (LED)

ii. Laser diode (LD)

iii. Single mode or single frequency laser diode (LD)

These are semiconductor devices that emits light when a voltage is applied.
LD's can sustain larger data rates than LED.

3.1.5 Terrestrial Microwave

Microwave radio transmission is widely used as an alternative to
coaxial cable for long distance communication. Microwave antennas
are parabolic and the diameter is typically 10 ft. Transmitting
antenna focuses a narrow beam to achieve line-of-sight transmission
to the receiving antenna. Microwave antennas are mounted on towers
which have a substantial height. The tower height is required to
extend the range between antennas and to be able to transmit over
intervening obstacles. The higher the tower, the greater the range.

A substantial portion of the frequency spectrum is covered by

microwave. Transmission frequencies range from 2 to 40 GHz
resulting in higher potential data rate for transmission. Attenuation
in microwave transmission is caused by rainfall. Another source of
impairment for microwave is interference which is caused by overlap
transmission areas.
Terrestrial microwave is used for long-haul telecommunication
service. Compared to coaxial cable, microwave facility requires fewer
repeaters. Another application of microwave includes voice and
television transmission.

3.1.6 Satellite Microwave

Within the last 25 years, communication satellites have become the
dominant carrier of international communication. The first
commercial satellite INTESAT (or Early Bird) was launched on April
5, 1965. The International Telecommunication Satellite Organisation
(Intelsat) has grown at a rate of 20 percent per year since 1965.
Satellite communications are more effective than any other means of
communication:

♦ When the nodes or sites, which require services, are scattered

over a wide area

♦ When the nodes are in motion or portable. Besides radio system,

satellite communication becomes the only solution for adequate
services.

3 - 40
CS260 CHAPTER 3: TRANSMISSION MEDIA AND SWITCHING TECHNIQUES

♦ When a great many receivers are at one location, to keep the

cost of communication low. Satellite communications are the
ultimate means of service provisions.

The customer to a satellite communication can be voice frequency

signal, data, video or fax. Satellite communications can be viewed as a
microwave radio tower in the sky. A satellite is used to link to two or
more ground based microwave transmitter /receiver, known as earth
stations. The satellite receives transmission on one frequency band
called uplink, amplifies or repeats the signal, and transmits it on
another frequency called downlink. There are two common uses of

Figure 3-5
communication satellite and they are depicted in the figure below

3 - 41
CS260 CHAPTER 3: TRANSMISSION MEDIA AND SWITCHING TECHNIQUES

Figure 3-6
In the first figure, satellite provides a point-to-point link between
two distant
ground-based antennas whereas in the second figure the satellite
provides a transmission link between one ground-based transmitter
and a number of ground based receivers.

A satellite needs to be within the line-of-sight of its earth stations

at all times. This requires the satellite to remain stationary with
respect to its position over earth. In order to remain stationary, the
satellite must have a period of rotation equal to the earth's period
of rotation. This match occurs at a height of 35,784 km.

For satellite transmission, optimum frequency ranges from 1 to 10

GHz. Below 1GHz, there is significant noise from natural sources,
including solar, atmospheric and human-made interference from
various electric devices. Above 10 GHz the signal is severely
attenuated by atmospheric absorption. For most satellites providing
point to point communication, uplink frequency ranges from 5.925 to
6.425 GHz where as down link frequency ranges from 3.7 to 4.2 GHz.
This combination is referred to as C band or 4/6 GHz band. Among
the most important applications for satellites are Television
distribution, Long distance transmission and private business
network.

3.2 Switching Techniques

There are three basic types of switching networks for transmitting
information and they are:
i. Circuit switching

3 - 42
CS260 CHAPTER 3: TRANSMISSION MEDIA AND SWITCHING TECHNIQUES

ii. Message switching and

iii. Packet switching

Figure below shows a generic switching network which consists of a set of
nodes connected by transmission media. The nodes can be computers,
telephones, terminals or other data communicating devices. Data entering
the network from a station are routed to the destination by being switched
from node to node. For example, data from station b, intended for station
d, can be sent by routes 2-3-5 or 2-4-3-5. The classification as to whether
the network is circuit, message or packet depends on the way nodes switch
data
from one link between the source and destination.

3.2.1 Circuit Switching

In this kind of switching, a single path between two nodes is
exclusively dedicated to customers. For example if customer with
access to station b wants to transmit / receive data to/from another
customer attached to station d, he has to do the following:
i. Call set up

Customer with access to station b dials up customer with access

to station d. If customer at station d is ready to communicate, a
dedicated path will be established between customer at b and
customer at d.

ii) Data transfer

Figure 3-7

3 - 43
CS260 CHAPTER 3: TRANSMISSION MEDIA AND SWITCHING TECHNIQUES

Once the dedicated path is established, data transfer between

customer at b and customer at d can continue.

iii) Call termination

Once the data transmission is over and both the parties agree to
terminate the session, the call will be terminated. Once the call is
terminated, the dedicated path will be released and can be used
by somebody else.

Telephone conversation can be one good example of circuit

switching. Circuit switching can be inefficient, since the path is
dedicated to the customers even when no information is being
transmitted. For a voice connection, utilisation is high although
not 100% (conversations, in general, do not have long periods of
silence). Also note that while sending data, error control can be
performed only on an end-to-end
basis, after the complete message has been sent. In this kind of
switching, after the connection is established, the network is
transparent to the customer. That is, the network becomes a
conduit for whatever kind of information the customer wants to
send. As a result of this transparent conduit, the network and
the customer do not have to agree on a common format when
data is being transmitted. For message and packet switching
however, customer and network must agree on formats.

3.2.2 Message Switching

Message switching is generally used for exchange of digital data such

as computer to computer communications or sending a telegram. This
technique however does not require a dedicated path between two
communicating nodes. The switching node is designed to have the
capability to store the message in a buffer until a link to the next
node between the two customers is available. Referring to figure 3-7,
assume that station b corresponds to a computer that is sending
information to another computer at station d. Node 1 is accessed and
information is stored in a buffer until a line from either 1 to 3 or 1 to
2 becomes available. Assuming that message switching node 2
becomes available first, the information is then transferred over the
1-2 link. Then it is stored in a buffer at node 2, where it will be
transferred to node 4 and then to the computer at station d. Thus,
the computer at d will receive the information from the computer at
station b. Any ensuing delay depends on the message length and the
time it takes for the intervening switches to become available.

In a message-switched network, the message in a given switch is

stored and subsequently forwarded to the next switch when a link
becomes available. Thus, the message-switched network is also
referred to as a store-and-forward network. Similarly, a packet-

3 - 44
CS260 CHAPTER 3: TRANSMISSION MEDIA AND SWITCHING TECHNIQUES

switched network is also a store-and-forward network. The message

storage capability at the switches makes the network appear to be
non-blocking, since if the path to the next node is busy, the message
switch may still accept the message, depending on the switch buffer
size. Potential blocking at the switch is translated into a potential
delay. However, being delayed is far less serious than being blocked.
Messages can be statistically multiplexed and transmitted when the
path to the next node becomes available. The tendency to fill in the
gaps between messages on the path leads to extremely high
efficiency of network resources as well as high utilisation of the
inter-switch connection channels.

Even when no message is blocked, the delay through the network can
be long with a high variance (depending on the message length). This
delay is a major disadvantage, and message switching is not well
suited for real-time traffic (e.g. digital voice) and specially
interactive communication (e.g. airline reservation). Another
disadvantage is that the message switch requires large buffers for
long messages. The hardware in these buffers is wasted when short
messages are received. Even if a path to the next node is available, a
switch node can not begin to relay the message until the entire
message has been received. Finally, because complete messages are
stored in full at the message switches, security can be a problem.
Personnel working at these switching centres would have access to
this information.

The network and the customer now need to agree on a common

message format. This brings up the need for agreements between
the network provider and the manufacturer of customer equipment.
Such agreements are called rules or protocols.

3 - 45
CS260 CHAPTER 3: TRANSMISSION MEDIA AND SWITCHING TECHNIQUES

3.2.3 Packet Switching

Data transmission, including digital voice, is bursty, since there are

periods of time when no transmission occurs. The packet switching
technique uses this burstiness to combine the advantages of message
and circuit switching while minimising the disadvantages associated
with those approaches. In this switching scheme network delay is
reduced and line efficiency is increased. Information is broken into
packets, and these small messages are transmitted packet by packet.
As in message switching, described above, each packet must be
stored in buffers at the switching nodes. Different packets of the
same message may follow different paths during the same time
period. Delay is reduced, since packets are short in length, have a
short transmission time, and require a small amount of buffer space.

Referring to figure 3-7, assume that station b corresponds to a

computer that is transmitting data to another computer at station d.
Assume, further, that the message consists of three packets, x, y,
and z. The basic transmission operation will be described for packet
switching using figure 3-7. An actual transmission involves packets,
composed of bits, inserted into frames that contain additional
control information bits. In this simple description, we do not
distinguish between bits, frames, and packets.

The flow of the message is initiated by the transmission of packet x

to switching node 1. After the packet is completely stored in the
buffer of node 1, the switch (following a set of routing rules) will
transmit packet x toward its destination by sending it, in this case,
to packet switch 3. (It is assumed for the moment that switch 2 is
not available). In the mean time, packet y is moving into the buffer in
switch 1. During this time, the conditions in the network change (for
instance, a large amount of traffic from end office c arrives at
switch 3), so the second packet, y, of the message is routed from b
to d via switch 2. The third packet of the message, z , arriving at
switch 1 soon after the second packet, y, is similarly routed via
switch 2. Any information that is properly received is acknowledged
by the receiving switch. Thus, in figure 3-7, switch 3 has sent a short
acknowledgement, designated as ACKx, back to switch 1, informing
the original sender that the packet was received without error by
the next switch. The acknowledgement is the key to the error
mechanism, which ensures the integrity and accuracy of the
transmitted data packet. If an acknowledgement is not received
within a certain predefined time period, the time out period, the
sending switch presumes that the packet was received erroneously
and re-transmits that packet.

Returning to our example, packet x is successfully transmitted to

switch 4. After being received correctly by switch 2, packet y is

3 - 46
CS260 CHAPTER 3: TRANSMISSION MEDIA AND SWITCHING TECHNIQUES

transmitted to the destination switch, switch 4. But during that

transmission, an error occurs. When switch 4 receives packet y, the
error mechanism determines the existence of the error. This error
can be either within the packet or within the control bits sent along
with the packet. Since, switch 4 does not know where the error is,
the packet is discarded. The error mechanism detects only the
presence of errors, not the location of errors within the packet or
control bits. Switch 4 can not request retransmission of the damaged
packet, y, since the control bits could be in error.

However, switch 4 knows the number called a sequence number, of

the next expected packet from switch 2. This sequence number is
contained within the control bits that are sent along with each
packet. The error mechanism requests a retransmission when the
next error-free frame from switch 2 arrives. This procedure is
illustrated in figure 3-7 where the packet z is shown arriving at
switch 4, behind packet y, with no errors. (If packet z contained
errors, it would appear that switch 4 has no way of communicating to
switch 2 that packet z contains errors. The time out mechanism
avoids any problem. Switch 2 would retransmit packet y after the
timeout period etc.). Once it is determined that packet z contains no
errors, switch 4 examines the sequence number within the control
bits sent along with packet z. Since this number does not equal the
expected sequence number stored within the memory of switch 4,
packet z is also discarded. However, a negative acknowledgement is
sent to switch 2 requesting a retransmission of all packets starting
with the sequence number of packet y. This procedure is illustrated
in figure 3-7 by the negative acknowledgement NACKy. Switches
keep copies of all transmitted packets until packets are successfully
acknowledged. The buffer within switch 4 is shown. These packets
are delivered to the computer at d in the correct sequence - x, y, z.

As noted earlier, packet switching is also store-and-forward system,

because it permits different packets from a single message to arrive
at the destination by different routes and with different delays. (Of
course, this method does not preclude packets arriving by the same
route). It also permits the received packets to be in different order
than the original transmission. The process of sequencing the packets
may be performed at the receiving switch (packet switch 4 in figure
3-7 ) using packet sequence information that accompanies the user
data through the network. Or it may be done by the customer at
station d, depending on the kind of service provided to the customer.

Each message in a circuit switching, message switching, or packet

switching must include not only the information bits, but additional
bits referred to as overhead information. Overhead information
contains both packet - specific information (a packet header) and link
- specific information (the frame bits which surround the packet

3 - 47
CS260 CHAPTER 3: TRANSMISSION MEDIA AND SWITCHING TECHNIQUES

header and packet data bits). Frame bits contain the control
information necessary for error-free transmission between the
communicating switches (refer to figure 3-7 ). Within the packet
header, the overhead bits can identify the destination of a message,
or packet, so that each switching centre will know how to further
route the information. Overhead bits also can specify the source of
the message or packet, so that acknowledgement is possible, and the
user identification, so that the user can be charged for services.
Synchronisation bits are part of the switch overhead and must be
included to identify the beginning and end of a frame containing the
message or packet. The packet header also can contain numbers to
allow reassembly in proper sequence.

In a message-switched system, the overhead information is appended

to each message, where as in packet switching, each packet is
accompanied by overhead bits. Thus, if there is more than one packet
/ message, the packetized message has more overhead.

Accordingly, with respect to message switching, packet switching has

two disadvantages:
18. To transmit a given amount of information per unit time, packet
switching requires that bits be transmitted at a more rapid rate
than is required in message switching.

19. The switching hardware needed to packetize, add overhead,

depacketize, and reassemble is more complicated and must
operate more rapidly than the corresponding hardware needed in
message switching.

Even though packet switching requires more overhead bits, a packet

switching system will transmit a give amount of information with less
delay than message switching. This feature results from the fact
that a switching computer can not begin to retransmit a message
(relatively long) or a packet (relatively short) until the entire
message or packet has been received. With packet switching, the
long message is broken into smaller packets so that the buffering
delay at the switches is reduced.

3 - 48
CS260 CHAPTER 3: TRANSMISSION MEDIA AND SWITCHING TECHNIQUES

Note again, in a manner similar to message switching, that the

packet-switching network and the customer must agree on a common
packet format. Agreements (rules or protocols) must be met between
the packet network provider and the manufacturer of customer
equipment. The X.25 protocol is an example of such an agreement.

[Link] Packet Switching Concepts

There are two kinds of rules, or protocols, for transmitting
packets. One is based on the processing rules within the
network and is referred to as internal processing; the other
is based on the processing rules between the customers and
the network and is called external processing. The
techniques for internal processing are called datagram and
virtual circuit.
♦ Datagram Technique
• In the datagram approach, the internal packet switching
network treats each packet independently, just as the post office
treats the mailing of letters. Occasional packets will be lost, and the
customer may or may not know this unless there is some
acknowledgement from the recipient. For example, if a bill is paid and
gets lost in the mail, you can be sure that the party owed the money
will be in touch. On the other hand, if you request information about
a product by filing out an information post card which is subsequently
mailed and lost, you probably will not be notified since the mailing was
not expected (there is also no economic impact from this loss).
Thus, in this approach, packets from a single message can arrive at
the customer by different routes, with different delays, and out of
order. The customer has the responsibility to determine that the
packets are out of sequence and to reorder them.
♦ Virtual-Circuit Technique

• In the virtual-circuit approach, a logical connection is

established before any packets are sent. It is similar to the call
setup, such as making a telephone call where a route must be
selected and equipment is dedicated to the call. The set-up
procedure for the virtual-circuit approach establishes the route
between the caller and the receiver. It uses software to make
entries into tables stored within each switching node along the route.
A table specifies the addresses of adjacent switches for the call.
Once the set-up is complete, information is directed over the same
route by the appropriate table within each packet-switching node.
The packet format after call set-up does not require complete
addressing information since the virtual circuit directs the packets
to their destination.
The virtual-circuit is "phone-like" because there is a specific route
between the caller and receiver through the networks for the

3 - 49
CS260 CHAPTER 3: TRANSMISSION MEDIA AND SWITCHING TECHNIQUES

duration of the call. With datagrams, routing may be different for

each piece of data. However, it does not mean that there is

3 - 50
CS260 CHAPTER 3: TRANSMISSION MEDIA AND SWITCHING TECHNIQUES

dedicated path, as in circuit switching. Data frames will be buffered if

the flow of traffic at a switch becomes too high and queued for
transmission over the established route. Transmission begins when lines
become available. When two stations exchange large files, which can take
a long time, there are certain advantages for employing the virtual-
circuit approach, all related to relieving the stations of unnecessary
communications-processing function. All packets for a specified virtual
circuit follow the same route and arrive in sequence.

3.3 Review Questions

20. What is a co-axial cable? What is the difference between Baseband

and Broadband co-axial cable?

21. "Optical Fibre is the medium of choice", Comment on this statement.

22. What is the advantage of using satellite microwave transmission?

23. Where would you use terrestrial microwave as a transmission media?

24. What is circuit switching? What are the disadvantages of circuit
switching?
25. Why packet switching is more efficient than message switching?
What are the typical applications of message switching?
26. Describe in detail, the difference between the datagram and virtual
circuit techniques
27.

3 - 51
CHAPTER 4 : NETWORK SECURITY

Chapter Objectives

Upon completion of this chapter, you should be able to :

⇒ describe what network security is;

⇒ describe some of the main reasons of network faults;

⇒ describe how errors are detected and corrected during data

transmission;

⇒ describe what is flow control and some of the protocols related to flow
control;

⇒ describe different measures that are adopted to prevent human

interference in data communication.

4.1 Introduction

For a computer network, security is of utmost importance. This is because

the network is used as a highway of information movement. The failure and
success of many companies depend on the integrity of the information
transmitted. Because of this security is one of the prime considerations of
network design.

Network security is concerned with protecting the network from faults or

from human interference with the network. Faults might occur within the
network itself or it may be caused by natural disaster or human error. On
the other hand, human can interfere with the network to corrupt the data or
retrieve valuable information. Network faults are minimized by designing the
network in such a way so that it can overcome the problem inherent into a
computer network. On the other hand, different protective measures are
taken to prevent human beings from interfering the network. The technique
and measures for network security is discussed here in detail.

4.2 Network Faults

Most of the network faults are caused by transmission impairment, which

means that the received signal differs significantly from the transmitted
signal because of some random variations and degradation of signal quality.
There are essentially three parameters that is attributed to this impairment
and they are:
♦ Attenuation

♦ Delay Distortion
♦ Noise

4.2.1 Attenuation
As the transmission distance increases, the signal strength decreases. There
are mainly three aspects of attenuation and they are:

4 - 52
CS260 CHAPTER 4 : NETWORK SECURITY

♦ Received signal should be strong enough so that the receiving end can
detect and interpret the signal.
♦ The signal level should be above the noise level so that the
transmitted signal can be received without error.
♦ Increase in frequency increases attenuation.

The first and second issue is addressed by using amplifier or regenerators.

Amplifiers or regenerators are used on the transmission links to boost up
the weak signal. The number of amplifiers or regenerators to be used for a
particular transmission link depends on the transmission media. The third
issue is caused because of imperfect amplitude frequency response. This can
be avoided if all frequencies within the pass band are subjected to exactly
same loss or gain.

4.2.2 Delay Distortion

Propagation velocity of a signal through a guided medium varies with
frequency. The velocity tends to be greater near the centre
frequency of a band limited signal and velocity decreases towards
the edges of the band. This causes different components of a signal
to arrive at the destination at different times causing delay
distortion. This phenomenon is very critical for digital transmission.
This type of distortion can be avoided by equalising the velocity of
signal across the band.

4.2.3 Noise

Noise consists of any undesired signal in a communication circuit.

Noise happens to be the major limiting factor in system performance.
There are four major categories of noise and they are:

[Link] Thermal noise

Thermal noise occurs in all transmissions media and all

communication equipment. It arises from random electron
motion, and is a function of temperature. This is also
referred to as white noise and is uniformly distributed across
the frequency band. Thermal noise places an upper bound on
any communication system performance and it cannot be
eliminated.

[Link] Inter-modulation Noise

If two signals with frequencies F1 and F2 as passed through a

non linear device or medium, the result will be inter-
modulation product that are spurious frequency components
(F1+F2). These components may be present either inside or
outside of the band of interest for the device. These
components interfere with a signal at the frequency F1+F2.

4 - 53
CS260 CHAPTER 4 : NETWORK SECURITY

[Link] Crosstalk

Cross talk occurs when there is electrical coupling between

nearby twisted pair or coaxial cable carrying multiple signals.
Another kind of cross talk happens when microwave antennas
pickup unwanted signal. While using telephone, one can
experience crosstalk when he can listen to another
conversation.

[Link] Impulse noise

Impulse noise is non-continuous, consisting of irregular pulses
or noise spikes of short and of relatively high amplitude.
These spikes are often called "hits". Impulse noise degrades
voice telephony only marginally. However it may seriously
degrade error rate on data or other digital circuits.

4.3 Error Detection and Correction Techniques

Synchronous and Asynchronous transmissions are susceptible to noise. The
speed of synchronous transmissions is often higher, and therefore the
consequence of noise is more severe. Asynchronous transmission contain a
great deal of idle time, where as synchronous data are contiguous, adding
further reason for concern.

Transmission line noise is by nature bursty. Relatively long period of time will
elapse without a serious noise and then a click will occur that will destroy a
tenth of a second worth of data. In a 9600 bit/second transmission, this
click will destroy a 60 bits or 120 consecutive characters. This leads to the
conclusion that error detection and correction techniques are of prime
importance.
♦ The basic principle of error control code is to add redundancy to
information bit stream. The error control codes can be classified as :
• error detection code

• error correction code

• error detection and correction code.

There are many error control codes but only few of them are discussed
below:

4.3.1 Even/ odd Parity

This is the simplest error control code where one extra bit is added
to the character being transmitted. For even parity this bit is set
such that the total
number of 1's in each character is always even. On the other hand,
for odd parity this bit is set so that the total number of 1's in each
character is always odd. Even /odd parity scheme can only detect
error but cannot correct them.

Example:

4 - 54
CS260 CHAPTER 4 : NETWORK SECURITY

Apply even parity to every two names: Dole, Baer, Fan, Bush, Chen,
Cole, Hart, king. Also discuss error detection/correction capability,
using the given received bit stream
0000011 0100110 1001011 1101111.

Information: 000001 010011 100101 110111

I I I I

Parity insertion: 1 1 1 1

Transmitted signal: 0000011 0100111 1001011 1101111

Error detection/correction capability:

Received signal: 000001 1 010011 0 100101 1 110111 1

Parity Checking:

First name: 1 + 1 = Even number of 1's

Second name: 1 + 1 + 1 + 0 = Odd number of 1's

Third name : 1 + 1 + 1 + 1 = Even number of 1's

Fourth name : 1 + 1 + 1 + 1 + 1 + 1 = Even number of 1's

So from parity checking, we can see that second name is in error.

4.3.2 Longitudinal Redundancy Check (LRC) and Vertical Redundancy

Check (VRC)
The simplest error detecting scheme is one dimensional parity
checking where an extra bit is added to each character to make the
total number of 1 bits in the character even (for even parity) or odd
(for odd parity). The one-dimensional procedure is known as Vertical
Redundancy Checking (VRC). This technique is effective if a
substantial number of the errors that occur in transmission are
single bit errors. VRC is clearly defeated by an error that affects
any even number of bits.

An additional parity check can be performed on the columns, called

longitudinal redundancy check (LRC). The terms vertical and
longitudinal refer to paper tape systems in which these techniques
where developed. The VRC/LRC combination is an effective error-
detecting scheme. But this scheme can be defeated by certain types
of multiple bit errors. An important thing to note is that VRC is also
applied to the parity check character.

4 - 55
CS260 CHAPTER 4 : NETWORK SECURITY

The following example illustrates the methods, and notes some

situation in which they fail to detect an error.

Example
Apply odd parity, LRC and VRC to the ASCII "It's a bird!". Discuss
error detection / correction capability if the received signal is given
as shown.
ASCII code (in octal)
I = 111; t = 164; ' = 47; s = 163; Space = 40; a =141; b = 142; I = 151; r
= 162d = 144; ! = 41

Received signal (each character plus its VRC is transmitted following

another character plus its VRC .)

11010010 11101001 01001111 11100110 01000000 11000010 01000000

1100010011010011 11100101 11001000 01000011 10010100

Solution:

Transmitted signal Received signal Parity verification

I 10010010 11010010 VRC NG
t: 11101001 11101001 VRC OK
': 01001111 01001111 VRC OK
s: 11100110 11100110 VRC OK
Space: 01000000 01000000 VRC OK
a: 11000010 11000010 VRC OK
Space: 01000000 01000000 VRC OK
b: 11000100 11000100 VRC OK
i: 11010011 11010011 VRC OK
r: 11100101 11100101 VRC OK
d: 11001000 11001000 VRC OK
!: 01000011 01000011 VRC OK

Check

char: 10010100 10010100 VRC OK

|||||||||
LRC verification onoooooo
kgkkkkkk ng = no good

4 - 56
CS260 CHAPTER 4 : NETWORK SECURITY

♦ VRC verification

• OK No error or even erroneous bits in one character

• NG Odd erroneous bits in one character

♦ LRC verification
• OK : No error or even erroneous bits among all the 1st, the
2nd, the 3rd, Á , bits of all characters (parity check character
included)

• NG : Odd erroneous bits, instead

If the channel has a good performance record, that is, one bit in
error in thousands of bits transmitted, we can assume that one error
in at most may occur in transmitting this sentence: "It's a bird!".
Then from the matrix, one can conclude that the first received
11010010 = i
will be detected as an erroneous character and corrected as
10010010 = I

4.4 Assuring Error-free Data Communication

Errors in a communication facility can occur because of many reasons. A
method of detecting and correcting errors in transmission must be
developed to assure error-free data communication. There are two
different approaches to the problem and they are:
♦ Transmit enough extra bits with the actual data to allow detection of
errors by the receiver. Once an error is detected, the receiver requests
a retransmission of the message.

Or
♦ Transmit sufficient extra bits with the message so that the receiver
can detect and correct most errors with high probability. This might
lower the efficiency of transmission because a relatively large number of
non-information bits are sent with every transmission.
The characteristic of the communication facility determines the
efficiency of the first approach. If the probability of error is low
enough, the efficiency of this method exceeds that of the "error
correcting" method because relatively few non-information bits are sent.
If the probability of error is high then efficiency will be low.

Error detection is the preferred choice since the communication

channels are reasonably error-free and the bandwidth costs for the
correction overhead are not always economical.

4.5 The Automatic Repeat/ Request (ARQ) System

For detect only approach, the receiver will typically signal the acceptance or
rejection of a block of user information. If a block is rejected, the
transmitter is to re- transmit the block in question. Thus an automatic
repeat request is generated by a reject signal. This type of error control
system is referred to as an automatic repeat request system.

4 - 57
CS260 CHAPTER 4 : NETWORK SECURITY

ARQ systems can be either stop-and-wait or continuous. In stop-and-wait

ARQ, the transmitter waits after each transmitted block for an
acknowledgement or reject signal (ACK and NAK). In continuous ARQ, the
transmitter continues to send blocks while acknowledgements arrive on a
return channel. Some means of identifying the block being acknowledged or
rejected must be provided.

4 - 58
CS260 CHAPTER 4 : NETWORK SECURITY

4.6 Flow Control

In a transmission scenario, the transmitting and receiving end may not

operate at the same speed and the transmitter can transmit data faster
than the receiver can receive. Typically the receiver has a buffer area where
data is stored and processed before the buffer area is cleared.
Consequently, if the receiver is not fast enough to clear the data buffer, the
buffer will be overwritten by the incoming data causing overflow problem.
Some sort of flow control mechanism must be in place between the
transmitter and receiver.

♦ Simplest kind of flow control mechanism is known as Stop-and-wait

flow control mechanism. This scheme works as follows:

• transmitter transmits one frame and waits for and

acknowledgement from the receiver;

• once the receiver receives the frame successfully assuming

error free transmission, it sends back an acknowledgement;

• upon receiving the acknowledgement, transmitter transmits

the next frame.

This scheme works fine when messages are transmitted in few large frames.
This is because there will be fewer re-transmissions required if necessary,
making an efficient
use of the transmission bandwidth. But in practice messages are not
transmitted in large frames, as they are more susceptible to error. Other
reasons include limited buffer size at the receiving end and large frames
monopolise the transmission links. So in practice, message is transmitted in
smaller frames. So if we use this scheme for flow control, it may not be
efficient because of the number of re-transmissions.
Another more efficient scheme is known as Sliding Window Protocol. In this
scheme the transmitter and the receiver is considered to have a window. The
maximum allowable size

of the transmitter window is the number of blocks the receiver is prepared

to accept at any time. Each time the receiver acknowledges a block, the
transmitter window rotates by one and the transmitter is permitted to send
one more block. (This rotating of the window results in the concept of the
"sliding window").

4.6.1 ARQ Sliding Window Protocol

Let us assume that every out bound block contains a sequence number
in the range (0, 2n -1). 2n is called the modulus of the protocol. The
sender is required to keep a list of all blocks that have been sent but
not yet acknowledged. This structure is called transmitter window.
Likewise the receiver keeps a list of blocks that it can accept. The
receiver's window, as it is called, is of constant size, while the
transmitters window may grow to a maximum size agreed to by both
parties. Note that the window size is a flow control parameter. If for
example, the transmitter window size is restricted to 1, we have a

4 - 59
CS260 CHAPTER 4 : NETWORK SECURITY

stop-and-wait protocol, while if the transmitter window size is large,

a substantial number of blocks can be sent in short order. Thus by
adjusting this window size, the receiver can protect itself from being
overrun by a zealous transmitter.

If the receiver gets a block that is not in its window, the block is
discarded without comment. If the block is the one at lower edge of
the window, it is acknowledged and the window is rotated. The
following diagram illustrates the idea with a transmitter window size
of 2 and a receiver window size of 1.

Figure 4-1
Normal sliding window interchange. Transmitter (top row) has window
size = 2, receiver window size=1

(1) Sender transmits block 0, receiver waits 0.

(2) Sender transmits block 1. It must now wait until an

acknowledgement is received, Receiver gets 0, sends ACK and
advances window.

(3) Sender gets ACK(0), sends block 2. Receiver gets block 1, sends
ACK (1) and advances window.

(4) (5) Blocks sent and received normally.

4.7 Human Interference

It is very important to protect the network from hackers who may copy or
corrupt the information. some of the techniques that can be used to prevent
humans from doing so are discussed below.

4.7.1 Authentication
Typically authentication is achieved by means of identifying the user,
user terminal and the level of authorisation. But these techniques
cannot guarantee that this authentication information will not be
tapped by a hacker. This problem can be avoided by introducing a
public key which will be known to the user and the host system only.
Whenever the user tries to establish a session with the host system,

4 - 60
CS260 CHAPTER 4 : NETWORK SECURITY

the host picks a random number, encrypts it with the public key and
sends it to the user. The user is suppose to decrypt the data using
the public key and send it back to the host. As the user is the only
person who knows the public key, the user can be quite safe. The
hacker may try to record the data traffic and retransmit during
another session. But he will be unsuccessful as the random number
will be different.

Besides the public key, the host might require the user to attach
additional authentication information with the message. This
authentication information could be a sequence number, secret
password or time information of transmission. All these additional
authentication information makes it difficult for the hacker to
record the data traffic and play it back at a later time during
another session.

4.7.2 Encryption
Encryption involves the use of a key to process the original character
string to produce a cryptogram which is transmitted over the
medium. The receiver, knowing the key, can decrypt the message to
get back original character string.

Figure 4-2

The message to be encrypted is known as "text". Encryption process

output is known as Cipher text or Cryptogram which is transmitted
over the medium. Key K is a parameter to the function which convert
the text to Cipher text. Once the message is encrypted, it becomes
difficult for the hacker to decrypt it as they don't have the key.

There are two standard categories of encryption techniques and they

are
♦ Substitution Cipher

In this kind of encryption, a letter or a group of letters is

substituted by another letter or group of letters. One example
of substitution cipher is shown below
Text : Informatics

Each character in the text is substituted by a character which is

shifted by n =5 letters then the cyphertext becomes
♦ Cyphertext : Nsktwrfynhx

4 - 61
CS260 CHAPTER 4 : NETWORK SECURITY

Generally it is easy to break substitution cipher.

♦ Transposition Cipher
In transposition cipher the order of letters is changed but the
letters remain unchanged. This is unlike substitution cipher where
the order of plain text symbols are preserved. For this kind of
cipher, there will a key which will typically be a word without
containing any repeated letters. All the letters in the key are
assigned numbers according to their alphabetical order. The letter
closer to the beginning of alphabet is assigned number 1. These
numbers are used to number the columns. The text to be encrypted
is written down having as many columns as there are numbers in the
key. Then all the columns are written down horizontally starting with
column 1.

Example:

Text:
thissentenceisgoingtobeencryptedusingsubstitutioncipher
Key: KEYWORD

K E Y W O R D
3 2 7 6 4 5 1
t h i s s e n
t e n c e i s
g o i n g t o
b e e n c r y
p t e d u s i
n g s u b s t
i t u t i o n
c i p h e r a

4 - 62
CS260 CHAPTER 4 : NETWORK SECURITY

♦ Ciphertext:

nsoyitnaheoetgtittgbpnicsegcubieeitrssorscnnduthinieesup
Typically the transposition cipher is more difficult to break than
substitution cipher.

4.7.3 Data Compression

In data communication, data compression technique is used to utilise

the communication channel efficiently. This compression technique
also acts as a form of encryption against hackers. There are
different ways of achieving data compression depending on the
application. For example think about a scenario where a departmental
store communicates to its head office using a communication
software. At the end of the day, the departmental store needs to
send a report to the head office listing all the items that are in
demand. Assuming there are 1000 items with each item name being
10 characters long, 1000*10=10,000 characters needs to be
transmitted for the item names only. The situation can be improved
if we represent the items by a sequence number. So we can achieve
data compression by transmitting sequence number instead of names.
Another way to achieve data compression is by encoding the symbols
in text depending on their frequency of occurrence. It is quite
apparent that some symbols occur more frequently than consonants.
Keeping this fact in mind, more frequently occurring symbols are
encoded using smaller number of bits compared to other symbols.
Number of bits that will be used to encode a symbol depends on the
probability of occurrence. Greater probability of occurrence will
mean lesser number of bits to encode the symbol. Encoding symbols
this way helps in achieving data compression compared to the scheme
where each symbol is represented using same number of bits.

4.8 Review Questions

28. What is noise? Describe different types of noise.

29. How error is detected during data transmission? Once error is
detected, what are the different ways in which the error can be
corrected?

30. Why is flow control is required in data communication? Describe

different techniques used to control the flow of data transmission.

31. Describe what is sliding window protocol.

32. What is encryption? Why is encryption required?

33. Describe what computer crime is.

34. Why do we need authentication? What are the different types of
authentication?

4 - 63
CHAPTER 5 : OSI MODEL AND TCP/IP

Chapter Objectives

After completion of this chapter, you will be able to:

⇒ learn about OSI model and the different layers involved;

⇒ learn and apply the following concepts;

• the function of Physical Layer and the services it provides to Data Link Layer

• the function of Data Link Layer and the services it provides to Network Layer

• the function of Network Layer and the services it provides to Transport Layer

• the function of Transport Layer and the services it provides to Session Layer

• the function of Session Layer and the services it provides to Presentation Layer

• the function of Presentation Layer and the services it provides to Application

Layer

• the function of Application Layer and the services it provides to user application
process

⇒ learn about TCP/IP;

⇒ learn about FTP and SMTP.

5.1 Open Systems Interconnection

Rapid proliferation of protocols, modems, signalling methods, and network
architectures created many problems. Users were obliged to support
expensive redundant equipment. The standards organisation decided to
define common structure of data processing machines to alleviate some of
the problem. In 1978 the International Standards Organisation (ISO),
proposed a layered network architecture, called the Open Systems
Interconnection Reference Model (OSI RM). This model makes it possible
for data processing machines from different vendors to freely communicate
and interchange data. In the OSI environment the complicated
communication task is divided into a set of manageable functions called
layers. Each layer performs a subset of task required to communicate with
other system.

Open System Interconnection (OSI) is concerned with the exchange of

information between systems, in fact, "open" systems. Within OSI, a
distinction is made between "real" systems and "open" systems. A real
system is a computer system together with the associated software,
peripherals, and terminals. Whereas, open system is only a representation of
real system that is known to comply with the architecture and protocols as
defined by OSI. In fact, an open system is that portion of a real system that
is visible to other open system in their attempts to transfer and process

5 - 64
CS260 CHAPTER 5: OSI MODEL AND TCP/IP

information jointly. It should be mentioned that OSI is not only concerned

with interconnection and exchange of information between open systems, but
also with interconnection aspects of co-operation between systems which
includes activities such as inter-process communication.

Figure 5-1

5.2 Layered Architecture Concept

Figure 5-2 shows a collection of real systems that are interconnected by
using physical media. The overall objective of all software or hardware that
forms the open system is to enable interconnection between application
processes. Application processes are an abstraction of user programs that
implement certain user applications. However only certain aspects of
application programs are of interest- those that concern communication
between users. To design and implement complex mechanisms for information
transfer and co-operation, the entire network must be viewed as a series of
layers. Each open system is logically divided into a collection of subsystems,
each implementing a certain function. A layer is a collection of all subsystems
from different open systems that perform similar function.

5 - 65
CS260 CHAPTER 5: OSI MODEL AND TCP/IP

Figure 5-2

5 - 66
CS260 CHAPTER 5: OSI MODEL AND TCP/IP

5.3 Layers, Services and Functions

A network of computers may be viewed as a series of layers. Each layer

provides a collection of services to the subsystems in the higher layer - the
ability to enable users to exchange data without concern for its
representation. As we go up the layers, the kind of services made available at
each layer changes. Each layer must implement certain functions over and
above the service made available to it by the lower layers, thereby enlarging
the service. The concept of layering is similar to solving a larger problem as a
series of smaller manageable problems. As an example, one has to provide a
service with which a user may transfer reliably across network without
concern for representation of information. This problem is solved by
implementing specific functions in different layers to solve each aspect of
the problem. The design of a layer may be changed to incorporate new
technology without having to change the other layers.

In some ways the layered architecture approach is similar to a top-down

design of a hardware/software system. One breaks down the problem of
interconnection between systems into a number of smaller problems, each
problem being solved within layer. This approach has a number of advantages
like modular construction, independence from implementation of the
underlying layer, maintenance, and standardisation. Once the service made
available by each layer is defined, then it is easy to see what the function of
each layer is. A layer bridge the gap between the services made available by
it and those available to it.

Figure 5-3

5 - 67
CS260 CHAPTER 5: OSI MODEL AND TCP/IP

5.4 Entities, Protocols and Interfaces

A subsystem consists of one or more entities. An entity is a hardware or a

software module. Only that aspect of a module, which is visible from the
viewpoint of communication between entities, is considered part of the
entity. For example, three entities are required to handle three transmission
lines in each system. The three entities together form the subsystem that
uses the transmission media.

A network may be viewed as a collection of entities. Two entities in two

different systems may interact or communicate with each other, provided
they belong to the same layer. Such entities are called peer entities, and the
interaction between them is governed by what is known as a protocol. A
protocol specifies the nature of information that may be exchanged by two
peer entities, both in respect of semantics and syntax. It also specifies
under what conditions, entity may send or receive information. Similarly, two
entities within a real system may interact with each other provided they
belong to two adjacent layers. This form of interaction across a layer is
consistent with the service made available by an entity to the entity in the
next higher layer. Its specification is in terms of the collection of primitive
operations called primitives that either entity is allowed or capable of
supporting. These are primitive in the sense that they may not be broken
down any further. Associated with each primitive is a collection of
parameters that may be exchanged.

In aspects pertaining to the implementation of the service, primitives are

not part of the architecture specification. Neither is the representation of
the parameters. An implementation of the service is referred to as
interface.

Figure 5-4
5 - 68
CS260 CHAPTER 5: OSI MODEL AND TCP/IP

5.5 OSI Reference Model

The diagram below shows the OSI reference model allowing application
program A in host X to communicate with application program B in host Y.
The connection between A and B is shown by the dotted line. The model
includes seven layers as shown. The lower layers, one through three exists in
both the nodes and the host computer and are responsible to transport data
packet across the subnet. Layers four through seven are called end-to-end
layers and exist in the hosts only. The upper layers, five through seven, deal
with the way the application program communicates. The transport layer is a
liaison between the lower layers and the upper layers and provides an end-to-
end reliable path between the applications.

Each layer provides services to the layer above it and receives some services
from the layer below. A layer requests the services of the layer below it
through an interface shown by vertical arrows. Each entity talks with its
peer entity, which resides in a network node or another data station. The
peer communication is governed by a protocol showed by dotted horizontal
arrows. The services and protocols at each layer are specified by the ISO.
The peer entities communicate with one another over a logical connection.
Physical connection exists at the physical layer which the actual physical
movement of information occurs.

Figure 5-5

5 - 69
CS260 CHAPTER 5: OSI MODEL AND TCP/IP

5.5.1 Physical Layer

Physical layer is responsible for the transparent transmission of

information across the physical medium. The data unit at the physical
layer is a bit. The standards of this layer include EIA-232-C, EIA-
232-D and X.21.

The attributes of the physical layer interfaces can be grouped as

follows:
♦ Electrical:
This is concerned with electrical characteristics. e.g. voltage
levels, impedances, and the timing of the electrical changes to
represent a binary 0 or 1.
♦ Functional:
This is concerned with the meaning of the voltage levels on
certain pins of the connector. Examples of this functions are
control, timing, data and ground, the interchange circuits
between two stations (functional characteristics of RS-232-C)
♦ Mechanical:

This deal with wires, types of connectors, physical dimensions,

allocation of pins and so on (physical characteristics of RS-232-
C).
♦ Procedural:

The procedural specification define the rules applying to various

function, the sequence in which certain events may occur i.e. they
describe what the pins on the connectors must do and the
sequence of events required to effect actual data transfer
across the interface. Example of procedural specification of RS-
232-C can be found in chapter one.

[Link] Services Provided to the Data Link Layer

Following are some of the services provided by the Physical

Layer to the Data link layer:

♦ Physical connection: A physical connection effectively sets up a "Data

Circuit between physical connections end points.

♦ Physical connection end points: The physical layers provide

identifiers of physical connection end-points which may be used by the
Data Link Layer.
♦ Sequencing: The physical layers delivers bits in the same order in
which they were submitted.
♦ Data circuit identification: The physical layer provides identifiers
which uniquely specify the data circuits between two systems.

5 - 70
CS260 CHAPTER 5: OSI MODEL AND TCP/IP

♦ Fault condition notification: The data link layer is informed of fault

conditions detected within the physical layer.

5 - 71
CS260 CHAPTER 5: OSI MODEL AND TCP/IP

[Link] Functions Performed by the Physical Layer

The following functions are performed by the Physical Layer

which are independent of the actual physical medium used.
♦ Establishment, maintenance and release of connections: Activates,
maintains and deactivates physical connections for bit transmission
between Data Link entities possibly through intermediate system, each
relaying bit transmission within the physical layer.

♦ Transmission: The transmission of bit streams, which can be in full

duplex or half-duplex, synchronous or asynchronous.

♦ Management: The Physical Layer protocol deals with some aspects of

the management activities in the layer.

5.5.2 Data Link Layer

The data link layer shields the higher layer from the characteristics
of the physical transmission medium and provides a reliable and error
free data link connection. Within the data link layer, the physical
layer data streams are arranged into blocks of data called "frames".
Synchronization of the bits within the frame, error detection, error
correction (by re-transmission of frame), and flow control are all
data link layer functions. Some of the existing standards of this
layer are:
♦ ISO HDLC, which is a superset of several classes of data link
protocols

♦ IEEE 802.2, for local area network

[Link] Services Provided to the Network Layer

The following are some of the services provided by the Data
Link Layer to the Network layer:
♦ Data link connection:

The provision of one or more data link connection

between two network entities

♦ Sequencing:
Maintaining the frames of data in the correct sequence
♦ Error notification:
If an unrecoverable error is detected, the Network layer
is notified
♦ Flow control:

The Network Layer can dynamically control the rate at

which it receives frames of data.

5 - 72
CS260 CHAPTER 5: OSI MODEL AND TCP/IP

[Link] Functions Performed by the Data Link Layer

The following are some of the functions performed by the

Data Link Layer:
♦ Establishment and release:
Set up and release of data link connections.
♦ Delimiting and synchronisation:
This is essentially the framing function which organises
bits into frames. Within the Data Link Layer, the Physical
Layer bit streams are arranged into blocks of data called
"frames". Synchronisation of the bits within the frame is
maintained.
♦ Sequence control:
Maintains the sequential order of frames transmitted
over a data link.
♦ Error detection:

This function detects transmission, format and

operational errors, usually by re-transmission of frames.
♦ Error recovery:
This function attempts to recover from errors, usually by
re-transmission of frames.
♦ Flow control:

The flow control function is done by turning frame

transmission on and off according to the state of the
receiving system.
♦ Control of data circuit interconnection:

This function conveys to the Network Layer the

information necessary to control the interconnection of
data circuits within the physical layer.
♦ Management:

The Data Link Layer protocols deal with some

management activities of the layer.

5.5.3 Network Layer

A practical requirement for a network is to allow multiple users to

have access to multiple applications. The essential functions of the
Network Layer is to provide the transparent transmission of data
from a Transport Layer in one system (e.g. in a user terminal) to a
transport layer in another system (e.g. an application host computer).

In complex network, the transport layer communicating entities need

not be adjacent to each other but are connected via one or more
intermediate systems which perform a relaying function. In some
cases the network layer provides

5 - 73
CS260 CHAPTER 5: OSI MODEL AND TCP/IP

routing functions. Each intermediate system may be a network, and a

set of one or more intermediate system used for a connection is
termed a "sub-network". A practical example might be the
interconnection of a public data network, a private data network and
a local area network. Network addresses are used to identify
Transport layer entities to the network layer. The data unit in the
Network layer is known as packet. One example of network layer
protocol is CCITT X.25, the packet layer protocol.

[Link] Services Provided to the Transport Layer

Following are some of the services provided by the Network

Layer:
♦ Network Addresses. Network addresses are provided by the network
layer and are used by transport entities to uniquely identify other
transport entities.
♦ Network connections: Network connections provide the means of
transferring data between transport entities.
♦ Network connection end-point identifiers: The network layer
provides to the transport entities an identifier which identifies the
network connection end-point uniquely with the associated network
address.
♦ Error notification: Unrecoverable errors detected by the network
layer are reported to the transport layer.
♦ Sequencing: The network layer may provide sequenced delivery of
packets over a given network connection.
♦ Flow control: A transport entity which is receiving at one end of a
network connection can cause the network service to stop sending data
units. This flow control condition may or may not be propagated to the
other end of the network connection.
♦ Release: A transport entity may request release of a network
connection.

[Link] Functions Performed by the Network Layer

Network layer functions provide a wide variety of

configuration from point-to-point connections to complex
combinations of sub-network with different characteristics.
Following are some of the functions performed:
♦ Routing and relaying: Network connections are provided by network
entities in end systems but may involve intermediate systems which
provide relaying. Routing functions determine an appropriate route
between network addresses.

♦ Congestion control: The objective of this function is to maintain the

number of packets with the network below the level at which
performance falls off dramatically.

5 - 74
CS260 CHAPTER 5: OSI MODEL AND TCP/IP

♦ Network connections: This function provides network connection

between transport entities, making use of data link connections provided
by the data link layer.

♦ Segmenting and blocking: The network layer may segment and/or

form into blocks the Network service data units for the purpose of
facilitating the transfer.
♦ Error detection: This function is used to check that the quality of
service provided over a network connection is maintained. Error
detection in the network layer uses error notification from the data link
layer.
♦ Error recovery: This function provides for recovery from detected
errors and may vary depending on the quality of the network service
provided.

♦ Sequence delivery: Sequencing provides for the sequenced delivery

of packets over a given network connection.

♦ Flow control: A transport entity which is receiving at end of a

network connection can cause the network service to stop sending data
units. This flow control condition may or may not be propagated to the
other end of the network connection.

5.5.4 Transport Layer

The transport layer provides the functions necessary to bridge the
gap between the services available from the network layer and those
required by the layers above. The four lower layers provide the user
with a transport service, the transport layer role is to complement
the underlying network so as to ensure that the required quality of
service is made available to the user.

Transport functions are concerned with cost optimisation, error

control, flow control, sequencing and multiplexing. The transport
layer can also check for lost or duplicated information. If the
network connection is temporarily broken, the transport connection
may be held until the network connection is restored.

Transport layer protocols are designed to cope with a variety of

different networks of varying quality of service. The data unit at the
transport layers and the layers above it are a message. One example
of Transport Layer protocol is United States Department of Defence
(DOD) Transmission control Protocol(TCP) used in AR PANET.

[Link] Services Provided to the Session Layer

The transport layer uniquely identifies each session entity by

its transport address. The transport service provides the
means to establish, maintain and release transport
connections between a pair of transport addresses. The
connections are full duplex and more that one connection can

5 - 75
CS260 CHAPTER 5: OSI MODEL AND TCP/IP

be established between the same pair of addresses. Following

services are provided by the transport layer:
♦ Transport connection establishment: Transport connections are
established between session entities identified by transport addresses.
The quality of service of the connection is negotiated between the
session entities and the Transport Service.
♦ Data transfer: Provides data transfer in accordance with the agreed
quality of service.
♦ Transport connection release: Provides the means by which either
session entity can release a transport connection.

[Link] Functions Performed by Transport Layer

♦ Transport Layer performs the following functions:

• Establishment phase: During this phase the following
functions are performed
• Obtains a network connection which matches the
requirements of the session entity taking into account
cost and quality of service

• Establish as optimum transport packet size

• Selects the function which will be operational during
data transfer
• Maps transport addresses onto network addresses

• Provides identification of transport addresses

• Transfers data

♦ Data transfer phase: This is achieved by the transmission of packets

and the following services :
• Sequencing
• Blocking

• Concatenation
• Segmenting

• Flow control
• Error detection

• Error recovery
• Expedited data transfer

• Transport connection identification

♦ Release phase: Release phase includes the following functions.

• Notification of reason for release

• Identification of transport connection released

• Transfer of data

5 - 76
CS260 CHAPTER 5: OSI MODEL AND TCP/IP

5.5.5 Session Layer

Lower four layers of the OSI model are concerned with providing
end-to-end connection. The upper three layers deal with user-
oriented services. More specifically, the session layer solves the
problem of dialogue discipline for application process-to-application
process communication and provide value. The session service is
provided by the session protocol making use of the services available
from the transport layer. The session service provides the means for
organised and synchronised exchange of data between operating
session services users.

[Link] Services Provided to the Presentation Layer

Following are some of the services provided by the Session

layer:

♦ Session connection establishment: Enables two presentation entities

to establish a session connection between themselves.

♦ Session connection release: Allows presentation entities to release a

session connection in an ordinary way without loss of data.

♦ Normal data exchange: Allows a sending presentation entity to

transfer a session service data unit to a receiving presentation entity.

♦ Interaction management: Allows the presentation entities to control

explicitly whose turn it is to exercise certain control functions. The
following types of interaction are defined: (i) two way simultaneous, (ii)
two way alternative, (iii) one way.

♦ Exception reporting: Permits the presentation entities to be notified

of exceptional situations not covered by other services.

[Link] Functions Performed by Session Layer

Most of the functions required are implied by the services
provided.
♦ Mapping of sessions connections to transport connection: This is done
on a one- to -one basis.
♦ Session connection flow control: There is no peer flow control in the
Session Layer. To avoid overload, the Transport layer flow control is
used.

♦ Session connection recovery: In the event of a transport connection

failure, the Session layer may contain the functions necessary to
establish a new transport connection in order to continue the session.
♦ Session connection release: Enables the release of a session
connection without loss of data.

5.5.6 Presentation Layer

Computers may use their own way of representing data internally.

Examples of differences include:

5 - 77
CS260 CHAPTER 5: OSI MODEL AND TCP/IP

♦ Character sets: ASCII vs. EBCDIC

♦ Integers: binary vs. Binary coded decimal

♦ one's complement vs. Two's complement

♦ Word size: 16 bit vs. 32 bit

♦ Bit order: left to right vs. Right to left

♦ Byte order: left to right vs. Right to left

The question is, `ahow can the information be transferred and still
preserve the meaning?' Agreements and conversions are needed to
ensure that different computers can understand one another. So the
Presentation Layer will :
♦ Encode structured data from the internal format used in the
sending machine to a suitable pattern for transmission

♦ Decode the data to the required representation at the receiving

machine

♦ Two things are needed:

♦ A system independent presentation of data, i.e., abstract syntax

♦ A common syntax between the syntax for exchange, i.e., transfer

syntax

[Link] Purpose of Presentation Layer

This layer is responsible for ensuring that information is
presented to the eventual user in a meaningful way. This layer
is concerned with the way the data is represented and the
data structures employed in the representation. This layer is
concerned only with syntax and not with semantics. One
example of Presentation Layer is ISO 8822: Connection-
Oriented Presentation Service Definition.

5.5.7 Application Layer

The purpose of the application layer is to provide a window for
correspondent application processes to communicate via an OSI
environment. Application processes exchange information by means
of application entities, application protocols and presentation service

[Link] Services Provided to Application Process

In addition to information transfer, these may include:

♦ Identification of communication partners by name, address, and
description.
♦ Determination of the availability of the desired communication
partners.
♦ Establishment of the authority to communicate.

♦ Authentication of intended communication partners.

5 - 78
CS260 CHAPTER 5: OSI MODEL AND TCP/IP

♦ Determination of cost allocation methodology.

♦ Determination of adequacy of resources.

♦ Determination of acceptable quality of service.

♦ Synchronisation of applications.
♦ Selection of dialogue discipline.

♦ Agreement on the responsibility for error recovery.

♦ Agreement on procedures for control of data integrity.

♦ Identification of constraints on data syntax.

[Link] Functions Performed by Application Layer

The application layer contains all functions which are required

for communication between open system not provided by the
lower layers. Communication between application processes
takes place via application entities in the Application layer.
These entities are divided into user elements and application
service elements. The latter are categorised as:
♦ Common Application Service Elements (CASE)
♦ Specific Application Service Elements (SASE)

Example of SASE is, File Transfer Access and

Management (FTAM).

5.6 Introduction to TCP/IP

Transmission control protocol/internet protocol (TCP/IP) is best described
as a suite of protocols that offers host-host communication for users
connected to diverse networks. The networks could be Public Data Networks,
Department of Defence (DoD) ARPANET or high-speed local networks.

DoD uses the terminology "Internet" referring to their collection of

networks such as LANs, Long Haul and satellite nets. However, we will use
the term in the less restricted sense as applied to any linked set of
networks.

In 1969 the defence department (DARPA or Defence Advanced Research

Projects Agency) established a four-node store-and-forward packet
network. By 1972 the network included research and university sites whose
hosts had implemented several protocols for end-end communication. The
evolution of the protocol proceeded and by 1982 a suite of protocols had
been refined into what we call transmission control protocol (TCP) and the
internet protocol (IP).

5.6.1 Internetworking and TCP/IP

TCP/IP is called a suite of protocols because it allows hosts on
different networks to communicate with one another. The term
Internet is also used to describe any collection of linked networks.
TCP/IP can be viewed as the best current alternative to the ISO

5 - 79
CS260 CHAPTER 5: OSI MODEL AND TCP/IP

stack for interoperability between machines. Most manufacturers

have not implemented ISO protocols yet, but have access to TCP/IP.
Three major advantages of TCP/IP are:
♦ Interoperability between hosts
♦ Internetworking capability

♦ Simplicity
♦ Different networks are linked together by bridges and gateways.
Bridges are two-layer devices that interconnect local area
networks that differ at the medium access control (MAC) layer.
Gateways in the internet world are three-layer devices. More
correctly they should be called routers.

As an example, Host X wants to use the services of Host Y. Host X

opens TCP connection through the internet to the host. In addition,
Host Z needs a file stored on Host X. Another TCP connection is then
established to a file server on X.

Figure 5-6
5.6.2 TCP/IP vs. ISO Model
Although developed before the definition of the ISO model, TCP/IP
and its services fit closely into this architecture. This model allows
the communication software to be broken into modules as follows:
Layer 7: Application Layer

Layer 6: Presentation Layer

Layer 5: Session Layer

5 - 80
CS260 CHAPTER 5: OSI MODEL AND TCP/IP

Layer 4: Transport Layer

Layer 3: Network Layer

Layer 2: Data Link Layer

Layer 1: Physical Layer

Figure 5-7

All layers are integrated to accomplish information transfer. In the

same way, each layer of the TCP/IP model:
♦ Builds or adds functionality for its own use from the services of
the lower layers

♦ Provides a service to the overall network, which allows standard

interfaces and protocols to be placed into the appropriate layers

♦ Is independent and is defined by function

♦ Can be changed without affecting the other layers

5.6.3 File Transfer Protocol and Simple Mail Transfer Protocol

The internet protocol suite has a file transfer protocol (FTP). FTP
allows users to log onto a remote host, list remote directories, and
obtain help with remote machines file syntax. FTP also can convert
between file formats, e.g. EBCDIC to/from ASCII. FTP allows a user
to access multiple machines in a single session. It maintains separate

5 - 81
CS260 CHAPTER 5: OSI MODEL AND TCP/IP

TCP connections for control and data transfer. FTP is a sophisticated

protocol that does more than transfer files from one system to
another (from a server to client). It can for example, handle third
party transfers. It should also be mentioned that another file
transfer protocol exists that does not
provide all the expense and sophistication of FTP called trivial file
transfer protocol (TFTP). TFTP does not use a reliable service like
TCP; instead it uses another unreliable packet delivery system. It is
restricted to transferring small files.
The simple mail transfer protocol (SMTP) is a standard for sending
mail between hosts. The protocol does not specify how the mail
system accepts mail or how the user interface presents the user with
incoming mail, nor does it specify how mail is stored. Once
communication has been established between users, the sender can
transmit one or more mail messages or even make a request that the
roles of sender and receiver be reversed so messages can flow the
other
way. SMTP attempts reliable operation but does not guarantee
recover from hosts that drop files. While the main features are
concerned with message transfer there are some functions that deal
with destination verification and handling.

5.7 Review Questions

35. "OSI Reference model enables open systems to communicate"
explain.
36. Define the terms layer, interface, protocol, entity and primitives and
give examples of each.
37. Explain the difference in flow control performed at the data link
layer and flow control performed in the Network layer.
38. What are the functions performed by the presentation layer?

39. Why do we need a layered architecture in a networking environment?

40. Reliability in data transmission is of prime importance. What are the
layers that contribute to a reliable data transfer?
41. What is the commonality between OSI model and TCP/IP protocol
suite?
42. Describe how FTP makes use of TCP/IP.

5 - 82
CHAPTER 6 : INTRODUCTION TO LAN

Chapter Objectives

Upon completion of this chapter, you will be able to :

⇒ learn the evolution of LAN and different components involved in LAN;

⇒ learn different LAN topologies and apply the concept in work place;

⇒ learn about different channel allocation techniques and apply the concept
in work;

⇒ learn about the different medium access techniques and apply the
concepts.

6.1 LAN Evolution

In the 1960s, computing needs were handled by a central computer with

batch processing and time sharing capabilities. Dumb terminals were
connected via low-speed transmission facilities to a central computer. Time
sharing provided a dramatic improvement over the old batch systems, but
had its own set of inherent difficulties. Central processor systems had to
have the capability to perform large, compute-bound jobs and hence were far
from optimal for interactive tasks. By the 1970s, minicomputers were
commonly used as a means of off-loading the central facility. Such machines
were generally located close to the user groups utilising them. This allowed
simpler, less costly connection methods. In addition, the machines offered
better price/performance statistics for many types of application.

Of course, the multiple minicomputer approach to computing is not without

its own set of problems. Some jobs require more power than can be provided
on a single machine of this size. Locally generated data bases often contain
information of value to other departments that may use a different system.
Thus, the next evolutionary step involved the interconnection of multiple
minicomputers into a network. Generally speaking, these networks utilised
the same philosophy as the long haul networks, despite the fact that the
distances involved were usually relatively small.

The explosive growth of intelligent office equipment in the 1980s (e.g. word
processors, typesetters, and copiers) coupled with ever increasing number of
intelligent terminals, has created a situation where literally hundreds of
computers can be found in many business settings. This equipment give local
computing control to the user but does, however, lack some of the facilities
offered by a mainframe computer. Indeed, the requirement for inter-
machine communication is greater than ever before. Since it is most cost
effective to store files and programs on large rotating disk drives, the
personal computer terminal must be able to download needed information
and/or programs onto small floppy disk systems. Word processors must be
able to communicate directly with typesetting machines for maximum
efficiency. Larger line printers provide better price/performance ratio in
outputting voluminous material than the dot matrix devices found on the
typical personal computer terminal. Extensive communication capability must

6 - 83
CS260 CHAPTER 6 : INTRODUCTION TO LAN

be provided in order to make the most efficient use of the personal

computer terminal.

As we have observed, wide are network (WAN) technology was commonly

used in the early stages of local area networking. This was a consequence of
the availability of the software and hardware products necessary to
structure such networks, rather than because such a approach is optimal in
the local environment. However, the long haul approach is completely
inappropriate, when hundreds or even thousands of microprocessor-based
systems require communication services. A new communication system was
needed to deal with the personal computer environment and the LAN evolved
to satisfy the need.

6.2 Characteristics of a LAN

The Institute of Electrical and Electronics Engineering (IEEE) defines a LAN

as follows:

"A data communication system allowing a number of independent devices to

communicate directly with each other, within a moderately sized geographic
area over a physical communications channel of moderate data rates."
♦ The following characteristics can be seen in a typical LAN:
• Limited geographical communication area

- In a LAN, the communication takes place within a moderately

sized geographic area, between 0.5 km to 10 km in diameter. This
is one of the things that distinguish it from wide area networks
(WANs). LANs are typically confined to a single building or group
of building that are close together.

6 - 84
CS260 CHAPTER 6 : INTRODUCTION TO LAN

• Data rate on the LAN

- The physical communication channel supports moderate data rates,

ranging from 1 to 10 Mbps. These rate are between the very high-
speed rates of links between computers and peripheral devices (> 20
Mbps) and the rates supported by the wide area network (9.6 Kbps
to 1.5 Mbps).
• Dedicated physical communication channel

- A LAN communicates over a physical communications channel, such as

a dedicated cable or other communications medium that is used to
hook all the devices together. This physical channel can be baseband
transmission on a coaxial cable, unshielded twisted pair, shielded
twisted pair or optical fibre, or broadband transmission on optical
fibre or coaxial cable.

• Peer-to-peer communication
- A LAN allows a number of independent devices, such as any
combination of terminals, PCs, workstations, printers, storage
devices, and computers, to communicate directly with each other.
Therefore in a LAN, all communicating devices have the same status
and communicate as peers. In other words, no one device controls the
network. Users select only the data addressed to them over the
network and each user has equal opportunity to gain access to the
network.
• Low error rate

- Typically, the error rate is less than 1 bit error in 108 bits since the
LAN uses a common physical channel and signals are usually in digital
form.
• LAN ownership

- LANs are generally owned and operated by a single entity, i.e. the
user organisation, individual, or company. Public telecommunication
companies are not allowed.
6.3 Major Elements of a LANs

♦ The major elements of a LAN are:

• Cabling Systems

- There are three major types of cable used to connect PC stations

in a LAN environment; wire pair, coax, and fibre. Wire pair is the
least expensive with the lowest transmission speed capability and
fibre is the most expensive with the highest potential
transmission speed. Many LAN networks will use more than one
type of cable in their cabling plan such as fibre for the LAN
backbone and wirepair for connection to individual stations.
• Protocols

- Protocols, are the rules which govern communication on networks.

The three most popular protocols used in the LANs today are

6 - 85
CS260 CHAPTER 6 : INTRODUCTION TO LAN

CSMA/CD, Token-Ring, and Token Bus. CSMA/CD is the Ethernet

standard, Token-Ring is the IBM standard, and Token Bus is the
MAP standard used in the manufacturing networks originating
with General Motors.
• Topologies

- Topologies describe the structure of LANs. There are physical

topologies which deal with the physical layout and the logical
topology which governs the way data is transmitted on the LAN.
The three major topologies include, bus; ring; and star. In some
cases the physical and logical topologies on a particular LAN
differ. For example, the Token-Ring operates logically as a ring,
but is physical laid out in a star structure.

6 - 86
CS260 CHAPTER 6 : INTRODUCTION TO LAN

• Gateways/Bridges

- Gateways connect LANs to other networks with different

protocols acting as a form of protocol converter. This would
include connections to WANs or other LANs. Bridges are used to
connect two or more LANs with common protocol.

Protocols and topology will be covered in this chapter. Cabling has been
covered in chapter 3 and Gateways and Bridges have been covered in chapter
8.

6.4 LAN Topology

LAN topology represents the physical or logical arrangement of network

stations in relation to each other. For example, in the token bus LAN, the
physical topology is a bus but it behaves logically like a ring. In close
association with the topology is the concept of transmission control for the
interconnections of these stations, primarily whether it is distributed
control or centralised control. There are three basic network topologies:
star; ring; and bus/tree. In implementation, the topologies each have their
own advantages and disadvantages. Sometimes a hybrid of these topologies
may be used (e.g., ring-star).

6.4.1 Ring

The ring topology consists of a series of nodes connected by

unidirectional transmission link to form a closed path. It can be
viewed as a series of point-to-point links with each node as an active
tap listening for bits transmitted to it from a preceding node and
regenerating (or repeating) them bit by bit to the next node. It can
only do this in one direction on the ring so that the nodes, or
repeaters, have only the simple task of sequentially relaying the
information through the links.

A node identifies a message on the ring as belonging to it by

recognising its own identifier or address in the message. Usually the
source that sends the message on the ring will remove the repeated
message circulated back to it from the preceding section. However,
there is the possibility of unattended data or data fragments
indefinitely circulating round the network due to the closed loop
nature of the ring. This can be controlled by the procedure in the
access control mechanism.

The rings transmission control can be either distributed or

centralised. In the distributes control, the sending node can
determine on its own when it may begin transmitting based on the
status of the ring at that time. In this way, every node participates
in the procedure to control access to the shared channel provided by
the ring (e.g. token access control scheme). In the centralised
control, one master station is responsible for initiating all data
transfers (e.g. IBM 8100 communication loop). Here, the master node
polls the nodes around the loop, allowing each of them in turn to send

6 - 87
CS260 CHAPTER 6 : INTRODUCTION TO LAN

data if they have been waiting when they are polled. In this case the
loop acts similar to a star in logical operation.

The advantages of ring topology are:

♦ all wiring is point-to-point. This permits mixed media type of
implementations, especially optical fibres connections;
♦ simplicity in its connectivity and maintenance;

♦ messages are regenerated at each node, giving a longer distance of

transmission as well as minimising transmission errors.

The disadvantages are:

♦ it is susceptible to many single points of failure as any break in the
ring will bring down the system;

♦ more complex access mechanism is needed.

The physical media used with ring topology are twisted pair and
baseband coaxial cable. It is also well suited for optical fibre
implementation.

Active Taps

Figure 6-1

6.4.2 Star
The star topology uses a central controller node as the hub of the
network; all other nodes are wired directly to the hub in a radial or
star-like manner. Point-to-point communication schemes enabling each
node to exchange data with the hub are used. All communication
between the nodes passes through the hub. The star topology easily
lends itself to centralised control of the communication whereby the
hub will be responsible for managing and controlling all communication
that exists between all nodes. It can also have distributed control in
the sense that end-stations can make a request connection to the
central hub or another station. In this case, the hub behaves like a
switch providing the physical connections between two stations upon
request from one of the stations (e.g. PBX-LANs).

6 - 88
CS260 CHAPTER 6 : INTRODUCTION TO LAN

In the star topology, the central hub can be complex and the end-
stations need only possess simple communication requirements. It is
well suited for communication systems with terminal intensive
requirements. The reliance on the central hub by the whole network
may pose the danger of single point of failure. In addition, it may be
costly to install due to the fact that many connections are needed
from each end-station. The physical medium used for the link is
usually shielded or unshielded twisted pair.

Figure 6-2

6.4.3 Bus/Tree
The communication network in the bus/tree topology is simply the
transmission media i.e., no switches as in the star topology or
repeaters as in the ring topology. All stations are passively attached
through appropriate hardware interfacing, directly to a linear
transmission medium, or bus, in a multipoint or broadcast manner.
Information signals propagate away from the transmitting station in
both directions on the bus to the terminated ends of the bus. Each
node is tapped into the bus and copies the message as it passes that
point in the transmission medium. The stations will identify the
messages meant for them by recognising the addresses on each
message.

The bus is a special case of the tree topology with only one trunk and
no branches. Bus/tree generally uses distributed control e.g. via a
contention mechanism such as CSMA/CD or by token control scheme.
Centralised control is also possible by having a polling mechanism on a
multidrop-bus configuration where a master controller polls each
node to initiate data transfers.

The Bus/tree topology is well suited for broadband networking

techniques as well as those of the baseband techniques. In this way
it allows maximum channel utilisation. It is both simple and flexible to
insert or remove a station into or out of the network without service

6 - 89
CS260 CHAPTER 6 : INTRODUCTION TO LAN

interruption as passive interfacing components are used between the

stations and the bus.

However, the bus/tree does face the following problems:

♦ it has a more difficult access control problem because of the single
media used in a multi-access control problem;
♦ balancing electrical signals on the bus as the bus length varies with
the addition or removal of nodes.

The media used in baseband bus networks are shielded or unshielded

twisted pair and baseband coaxial cable. For a broadband bus/tree
network, broadband coaxial cable is used. Research is currently in
progress to make the optical fibre economically feasible as the
broadband medium for the bus/tree network.

6 - 90
CS260 CHAPTER 6 : INTRODUCTION TO LAN

♦ Bus

♦ Tree

Figure 6-3

6.5 Channel Allocation

Channel allocation is another area of interest to LAN . It is the procedure
for dividing the communication resources, the channel capacity, with the
maximum efficiency. The channel refers to the physical medium used for
transmission. The allocation procedure is independent of the access control
mechanism used for the channel. The access control mechanism is the way by
which access to the channel created by the allocation procedure is
controlled. For example, we may allocate a channel by dividing it into
frequency bands, but how these bands are used will be determined by the
access control mechanism. We may allow a master node to determine which
node pairs will own a given band or, alternatively, we may allow nodes to
contend for a given band. The concept of the access control mechanism will
be discussed later in this chapter.
♦ We have two ways of looking at channel allocation:

• by the allocation techniques used, i.e. space division multiplex

(SDM), time division multiplex (TDM), frequency division multiplex
(FDM), or a hybrid of these; and
• by signalling methods for transmission used in the network,
especially those with bus/tree topology. The signalling methods can
be either baseband or broadband signalling, which gives rise to the
names of networks such as broadband and baseband networks. In
these methods, the digital information to be transmitted over the

6 - 91
CS260 CHAPTER 6 : INTRODUCTION TO LAN

medium must first be encoded such that the bits are distinguishable
at the receiving node(s). The rate at which the encoded bit
information is applied to the medium by a sending node is referred to
as the transmission speed expressed in bits per second.

6.5.1 Channel Allocation Techniques

[Link] Space Division Multiplexing

This is merely a complex way of saying that there are
multiple physical channels, which, when considered together,
make up the LAN communication facility. Such an allocation
procedure, is of course, common in star topologies. But is
never found in bus or ring networks.

[Link] Frequency Division Multiplexing

This provides an effective way of sharing a channel capacity
among multiple users. FDM allocation procedures are
principally found in bus topologies, although they could be
used either in conjunction with space division multiplexing or
as a stand-alone basis to provide circuits between outlying
nodes and the central node in a star network. FDM is
impractical in the ring networks because of the point-to-point
connections of the channels.

[Link] Time Division Multiplexing (TDM)

This is the allocation mechanism of choice in the ring

topology, with nodes being granted (by the access control
mechanism) exclusive right to the channel for a time interval.
Fixed TDM could also be used to provide multiple independent
channels for outlying nodes in a star topology. TDM is an
applicable allocation procedure in a bus network where the
access control mechanism grants exclusive rights to the bus
to a single node for a time interval long enough to transmit its
data packet.

[Link] Combination of FDM and TDM

A network can use FDM to divide the channel bandwidth and

then use TDM to enable sharing of each frequency band. This
gives an extra mileage on the usage of channel. Of course,
the cost and complexity of such system would be more than
that which implements only TDM or FDM.

[Link] Code Division Multiplexing (CDM)

♦ CDM has only a recent history in the commercial usage of the

technique. It is digital multiplexing technique based upon spread-
spectrum theory of communication. It has two distinct categories:
• direct-sequence (DS) spreading, based on amplitude
modulation

6 - 92
CS260 CHAPTER 6 : INTRODUCTION TO LAN

• frequency hopping (FH), based on frequency

modulation
The signal meant for a given user is assigned with an
identification code which can be a distinct sequence (pattern)
of +1 or -1s for DS modulation or a distinct sequence or
frequencies for FH modulation that only user's receiver
recognises. The receiver knows in advance how the
transmitter will spread the frequency spectrum and acquires
the signal and continues to track the transmitted pattern.
For example user1 gets sequence S1, user2 gets sequence S2,
and so forth.

6 - 93
CS260 CHAPTER 6 : INTRODUCTION TO LAN

When a receiver wants to listen to user 1, it follows S1 i.e., it

sees all user 1's energy but only a small fraction of the
others' energy.

CDM offers advantages in better privacy in transmission, less

fading problems in built up areas, lesser vulnerability to
jamming of the signal, flexibility in that it does not need
precise time co-ordination between transmitters as in the
TDM, and potentially it can handle higher capacity than FDM
and TDM. It however has increasing error rates as the
number of data users increases. One of the area in which
CDM is used in is the wireless LANs implementations.

6.5.2 Baseband Network

The baseband network uses basement signalling techniques for
transmission, i.e., transmitting and receiving unmodulated signals. The
encoded signal is directly applied to the medium either as a continuos
stream of voltage transitions on a copper medium or as a stream of
light pulses on an optical fibre medium. Unlike the signals in
broadband network, baseband signals travel in both directions on the
medium, implying that only a single piece of the medium is needed .
Since baseband signals from each station are usually digital in nature,
it uses the entire physical channel bandwidth. This requires the
network to use some form of TDM methods to allocate the resource
among the stations.

Time on the channel may be allocated via one of several procedures.

Devices may be polled by some central "authority" or, alternatively,
may own specific time slots. Yet another interesting alternative is
having devices contend for the use of the channel.

An important variable to be considered when discussing baseband

networks is the allocated time interval. Time intervals can be of
fixed length, in which case the amount of data which a device can
transmit during its ownership of the channel is directly related to
the speed of transmission. A second possibility is to allow the device
to transmit a specified amount of data during its time slot. If the
data unit is of fixed length, there is obviously no distinction between
this approach and the fixed-time method.

Baseband data rates exceeding 100 Mbps are possible (e.g. FDDI
networks). However, practical limitations in some transmission media
(e.g. twisted pair) result in typical baseband network rates of up to
16 Mbps. Baseband signals must be periodically repeated over a long
distance to avoid data loss or interference due to signal degradation.

6 - 94
CS260 CHAPTER 6 : INTRODUCTION TO LAN

6.5.3 Broadband Networks

Networks that transmit and receive modulated signals are known as

broadband networks. The physical channel bandwidth can be divided
into frequency-band channels. Different devices can then employ
different carrier frequencies and multiple signals can simultaneously
reside on the transmission medium. Of course, the different
frequency channels can themselves be time division multiplexed, thus
providing an additional degree of sharing. LANs using modulation and
FDM are called broadband LANs. These are invariably bus/tree
topology.

Transmission on the medium is only in one direction. Any broadband

system requires two channels on which data must move. An outbound
channel is used for traffic leaving the headend. The system's
headend is simply a signal processor that takes an in-bound signal and
upconverts it to an outbound

6 - 95
CS260 CHAPTER 6 : INTRODUCTION TO LAN

channel. When a device transmits, it always does so on the in-bound

channel and listens and receives on the outbound channel. Both
channels can reside on a signal cable or can, as with Wang's
broadband system, use two cables - one for inbound traffic and, the
other for outbound traffic. Now the full range of frequencies
offered by the medium is available, but the cost of the system is
increased by the two-cable requirement and the necessity of having
two taps at every node. The transmission medium is typically cable
television (CATV) coaxial cable.

♦ With the availability of multiple sub-channels in a broadcast network,

different channels can be used to satisfy different requirements. There
are three kinds of sub-channel allocations for these requirements :
• Dedicated service type - a portion of cable bandwidth is
reserved for exclusive use by two devices.
• Switched service type - this requires the use of a number of
frequency bands. Devices are attached to a "frequency agile"
modems which can be accessed in a way similar to a dial-up
line.
• Multi-access service type - this is the most common and
allows a number of attached devices to be supported at the
same frequency. As with baseband, it needs some form of
access control mechanism to enable distributed, peer
communications among the devices.

6.5.4 Baseband vs. Broadband

The advantage of a broadband network is that it can carry analog as
well as digital information (e.g., full band-width video can be
distributed along with data). Another advantage of the broadband
network is its ability to accommodate multiple, special-purpose
applications on a non-interfering basis, even to the extent of
containing subnetworks operating on TDM principles. Broadband
networks also provide longer distance of transmission and potentially
more bandwidth than the baseband networks A disadvantage is the
need for radio-frequency (RF) modems which are often more
expensive than the relatively simple transceivers used in baseband
schemes. Another disadvantage is that the propagation delay through
a broadband network is likely to be much greater than that in a
baseband network. It is also more expensive than the baseband
network.

It should be noted that the term "broadband" was co-opted into the
local network vocabulary from the telecommunications world, but
with a change in meaning. In the local network context, the term
refers to the transmission techniques using analog signalling (such as
FDM, FDM/TDM) on coaxial cable. Broadband, as in BISDN, holds a
very different meaning in that it has nothing to do with transmission

6 - 96
CS260 CHAPTER 6 : INTRODUCTION TO LAN

techniques. In BISDN it means that the signals occupies a very large

frequency band spectrum that can be varied on a demand basis.

6.6 Access Control

♦ Access control deals with how to regulate the use of a shared

medium and is a key technical issue for LAN. Access control is exercised
at two main areas:
• Centralised Control - whereby a designated controller has the
authority to grant access to the network. A station wishing to
transmit must wait until it receives permission from the controller.

6 - 97
CS260 CHAPTER 6 : INTRODUCTION TO LAN

• Distribute control - whereby the stations collectively

perform a medium access control function to dynamically determine
the order in which stations transmit.

♦ Advantages of centralised control are:

• fewer problems in co-ordinating the activities of multiple
devices;
• greater access control in terms of priorities, overrides, and
guaranteed bandwidth to each station; and
• less intelligence in individual devices, thus simplifying their
network interfaces.
♦ Disadvantages of centralised control are:

• higher risk of single point of failure affecting the entire

network;

• reduced efficiency if the control point becomes a bottleneck;

and

• overhead may be unacceptable if propagation delay is high.

♦ The pros and cons for distributed control are mirror images of the
previously discussed points. The access control can be exercised by the
following access control techniques:

• Synchronous techniques - a specific capacity is dedicated to a

connection (e.g. in digital PBX). Such techniques are not optimal in a
broadcast network because the needs of the stations are generally
unpredictable.

• Asynchronous techniques - where the capacity of the channel

is allocated more or less in response to immediate needs. This is the
more preferred technique. Contention and round robin schemes fall
under this category.

6.6.1 Round Robin

♦ Token-passing Ring
A small frame or free token is circulated round the ring when all stations
are idle. A station wishing to transmit must wait until it detects a passing

6 - 98
CS260 CHAPTER 6 : INTRODUCTION TO LAN

token. When a station receives a token, it seizes it by changing one bit in

the token, and transforming it into a busy token, which is actually a
start-of-frame sequence for a frame. The station then appends and
transmits the remainder of the fields needed to construct a frame. The
transmitting station inserts a new token on the ring when both of the
following conditions are met:

6 - 99
CS260 CHAPTER 6 : INTRODUCTION TO LAN

• the station has completed transmission of its frame; and

• the leading edge of its transmitted frame has returned

(after making a complete circle of the ring) to the station.

When the transmitting station releases a new free token, the next
downstream station with data to send will be able to seize the token
and transmit. The use of the token guarantees that only one station
at a time may transmit. The token passing ring scheme may also be
used in buses with ordering of nodes.
♦ Collision avoidance

The collision avoidance scheme uses the distributed round robin

technique. A specific collision avoidance scheme is the carrier
sense multiple access with collision avoidance (CSMA/CA)
method. Each station listens to the carrier while a transmission is
in progress. After the transmission ends, each station waits for a
specific period of time, based on its position in a logical list of
stations. If no other station has started transmitting by the time
a particular station's time has elapsed, it may begin
[Link] example of the collision avoidance scheme is the
ISDN's Basic Rate Interface (BRI), whereby a passive bus is
used for connecting multiple ISDN terminals to the Network
Termination (NT) device.

6.6.2 Contention
In this scheme, no control is exercised to determine whose turn it is.
The applied traffic is non-deterministic and many stations may
simultaneously seek to access the medium. All stations randomly
contend for time on the medium, transmitting whenever the
transmission medium is available. Their principal advantage is that
they are simple to implement and efficient, under light to moderate
load. Contention techniques are usually appropriate for bursty
traffic. CSMA/CD is one example of contention technique.

Carrier sense multiple access with collision detection (CSMA/CD)

CSMA/CD with Collision Detection (CSMA/CD) is the most commonly
used contention scheme. CSMA/CD consists of two parts: the carrier
sense multiple access part in which the station uses the listen-
before-talk scheme; and the collision detect (CD) part which uses the
listen-while-talking scheme. In this scheme, a station wishing to
transmit listens to the medium to determine whether another
transmission is in progress. If the medium is idle, the station may
transmit. Otherwise, the stations back off for some period of time
and try again. A station continues to listen to the medium while it is
transmitting. If a collision is detected during transmission, the
station immediately ceases transmitting the message and transmits a
brief jamming signal to ensure all stations know that there has been
a collision. After transmitting the signal, the station waits for a
random amount of time and then attempts to transmit again.

6 - 100
CS260 CHAPTER 6 : INTRODUCTION TO LAN

6.7 Review Questions

43. What are the typical characteristics of a LAN?

44. What are the typical components of LAN?

45. What are the advantages of the ring topology compared to the star
topology?

46. Describe different channel allocation techniques.

47. What is the difference between channel allocation and access
control?
48. Why do we need a access control mechanism in LAN? Describe the
access control mechanism in token ring network.
49. What is the difference between CSMA/CA and CSMA/CD?

50.

6 - 101
CHAPTER 7

Chapter Objectives

After completing this chapter, you will be able to :

⇒ plan for a Local Area Network implementation;

⇒ make the right decisions about hardware and software;

⇒ learn the installation and configuration of network software and

hardware components;

⇒ administer a Local Area Network.

7.1 Planning for a LAN Implementation

When you plan for the implementation of a Local Area Network (LAN), the
three factors: people, process and organisations become more clearly
defined with addition of detailed requirements for the user, the job, and the
workgroup.

Your hardware and software choices must be considered in light of user

requirements, the need of the job, and the size of the organisational budget.
The best solutions will meet the need for adaptability, and scalability and will
provide workgroups with a platform for completing tasks without placing
unrealistic burdens on the organisation.
In the planning process, you should consider several things:
♦ The first consideration is making up a user group.
♦ Selecting workstation hardware

♦ Selecting a network hardware platform

♦ Selecting a network topology

♦ Designing the physical location of the equipment

♦ Selecting the Network Operating System (NOS)

7.1.1 Making Up User Groups

The initial consideration for making up a user group is to determine
what the makeup of the workgroup will be. Who will be working
together? You need to know this to determine who will need remote
access, or perhaps an independent server. The workflow among users
will determine how you set up workgroups and assign workgroup
managers. The workgroup requirements will determine what kind of
application software or Network Operating System you need, and
also gives you preliminary idea of printer set-up, menu design, and
user communication.

The number of users who will have access to the network and how
they will use it will be a determining factor in your decision about
which network software and hardware to purchase. This information

7 - 102
CS260 CHAPTER 7

is essential to avoid having to immediately upgrade the network or

justify having 100 or 50 extra user licenses that go unused.

7.1.2 Selecting Workstation Hardware

You should select hardware for the network based on the needs of
the application and the scenario for growth of the organisation. The
older and less powerful workstations can be connected with a
network to allow sharing of file and application. With the abilities
provided by the network operating system, their work life can be
extended.
♦ In the selection process for network components, you must answer
many questions concerning how will various machine, applications meet
the needs of the network. Often the best start is beginning with the
most basic considerations, such as :

• What are the goals for the organisation that the network
installation is supposed to support?

• How many people will need network access today; how many
next year; how many in the next five years? What are your
goals in expanding your business and your employee base?
• Will users need remote access?

• How important is security to your organisation? Who needs

access to critical information and who does not?

• How many of the existing computer hardware can be

incorporated into the new system.

Question such as these will lead to more specific ones concerning the
actual use of the individual workstations in the network.

7.1.3 Selecting Network Hardware

Local Area Networks run on a system of cables, hubs, routers,
servers and other peripherals. Different topologies and different
methodologies may require different network hardware and cabling.
♦ Local Area Networks normally are connected by one of the four
types of cable :

• Coaxial Cable
- It is relatively immune to noise (Electromagnetic
Interface) and is common choice in areas where the
environment is noisy. It is most often used in Ethernet
and ARCnet topologies.
• Unshielded Twisted-Pair (UTP)

- It can be used in nearly every situation, except those

requiring fibre-optic cable installation. UTP comes in
several grades and depending on the type of network,
some attention must be paid to grade.

• Shielded Twisted-Pair (STP)

7 - 103
CS260 CHAPTER 7

- It is noise resistant and capable of carrying signals over a

longer distance than UTP.
• Optical Fibre

- Optical fibre is the most efficient option and the most

expensive. If you need very high-speed transmission,
practical immunity to electromagnetic interference and
environmental hazards, and send signals over long
distance (100 KM), then this is the best choice. Because
of its large base of installed fibre, the telephone
company might be the best choice for information and
possible installation.

- Your selection of a network topology will be a guiding

factor in choosing which media will best suit the needs of
the network. The choice may also be influenced by what
type of cabling is already installed. Network design can be
an important factor in eliminating the need for hubs and
concentrators, which can be a real saving in effort and
money.

7.1.4 Selection of Network Topology

The topology you select for the network will have an impact on how
the workstations, servers, and other network components are
arrayed. You should consider the distance limitations for routing
different type of media. There will be a increased costs for the
addition of concentrators and hubs and additional signal-management
equipment to serve workstations over longer distances. The concept
of the right tool for the right job is important to consider.
♦ The selection of network topology also depends on the same three
factors : people, process and organisation. Below are the some factors to
consider in defining network topology selection :

• Physical location of the user

• Type and amount of network usage

• Presently available facilities to the organisation

• Level of reliability required for network operation

• Special requirement such as large data storage capacity,

security and need for remote access, or inter networking

• Future requirements such as to adopt the network to faster

backbone networks with fibre optics or other high bandwidth
transmission media
• Maintenance support required to ensure that the network is
managed and maintained

7.1.5 Physical Location of Network

Planning the physical location of workstations, servers, and the

peripherals is an important part of the design of the network. The

7 - 104
CS260 CHAPTER 7

physical location of components must allow for maintenance, security,

and access and also provide a location with adequate ventilation to
keep the machines from overheating.

Workstations
The physical location and configuration of the workstation and its
attached components; the monitor, keyboard, mouse, digitiser and so
on - are determined by considerations of user needs, comfort and
available workplace. The decisions you make about cabling determine
the distance the workstation; may be from server or hub and what
options are available for routing the cable. The selection of a
workspace for the user will be more of a factor then the length of
the cable attaching the workstation to the network. The workstation
should be ergonomically situated to facilitate a good physical working
environment for the user and to support workgroup logistics.

Server

The server is the heart of a Local Area Network. The Network

Operating System (NOS) and network application programs reside in
it. If the server is down, the network is down. Server performance
will be better and it will have a longer life if you take care in planning
its location. You must provide a clean environment free of dust with
good ventilation for cooling. You also need enough space for
maintenance to be performed easily, a clean stable power supply, and
a lockable door to protect the server from any tampering.

A major consideration for the location of the server is security. In

practice, the server operation is so mission critical to organisations it
should be kept in a "restricted room". This stops unauthorised
personnel from turning it off or tampering with it in other ways,
costing the organisation time and money.

Network Printers and Other Peripherals

Network printer and other peripherals such as plotters, should be
located in area accessible to users. If they are attached to the
server, they should be closed to it. Parallel cabling should not exceed
25 feet. If they cannot be located that close to
the server, then other method of attachment will be necessary such
as connecting directly to the LAN or to a dedicated workstation and
let the workstation serve as print server. The characteristics of the
printer selected, the type and size of the network, the capabilities
of the server, the needs of users, the job to be done, and the budget
constraints of the organisation should all be considered.

7.1.6 Selecting Network Operating System

The Network Operating System (NOS) resides in the server and

provides the connectivity that completes the networking system and
creates the environment in which the network operates.

7 - 105
CS260 CHAPTER 7

The selection of Network Operating System does depend on many

factors. Some of the important factors and guidelines are :

Platform Support and Broad Functionality

A facility for the implementation of technologies across multivendor
computing platform is one of the important advantages of Local Area
Networking. The Network Operating System must support a wide
variety of hardware and software to effectively provide a seamless
connection, one that makes the user interface consistent across
multivendor hardware, diverse Operating Systems, and even
different networks.
The Network Operating System must be able to exploit the existing
hardware and software to provide needed capacity and functionality
for a distributed processing environment.

A Consistent and Intuitive User Interface

An interface that looks the same across the network will help keep
user productive and reduce confusion. There are some instances
where a migration to Local Area Networking is prompted mainly by a
desire for a more intuitive, user-friendly interface.

Ease of Installation and Configuration

Most business cannot afford to disrupt operations for very long. The
easier the software is to install, configure, and learn to use the less
time your operation will lose.

Competitive and Flexible Pricing

There are simple, inexpensive implementations of Network Operating
System and there are also complex, expensive ones. The solution for
your needs will have to be something that fits within your budget
constraints. Be prepared to negotiate for price versus performance.

Application Scalability

It is the key to keeping your network at it's highest level of

performance. The application that you choose needs to be adaptable
to future needs. They need to be portable to different operating
systems and platforms. You must be able to expand them and provide
for adding of more users, dealing with changes in available computing
platforms and upgrading network components for increased
performance.

Modular Component Options

The NOS should be able to provide a scaleable solution for networks.

If you can buy what you need for now and still plug in what you buy
later to do more of the
job, then you will be able to pick and choose just those functions that
you need for your situation.

7 - 106
CS260 CHAPTER 7

Single-point Monitoring

A Single-point monitoring makes it possible for whole network to be

watched for problems from one console. It facilitates security and it
also provides data to be analysed by a protocol analyser or network
monitor.

Conflict-free Systems Coexistence

Conflict-free systems coexistence for the system already installed
and operating on your network is important. Backward compatibility is
another aspect of network that you can exploit to enhance the usage
and lifetime of components.

Security Management

A single interface security management should be defined, audited,

and maintained so that complete control of security can be managed
by policy administration.

Event Management

♦ Event management is a facility that responds to system events on a

network and is administered from a single point. Depending on a
predefined event policy, the Event Manager can take a number of
actions:

• Respond to messages
• Suppress messages

• Launch programs or scripts

• Forward messages to other platform

• Initiate execution of other platforms

• Establish database alerting

Problem Management
This facility allow you to maintain asset information, such as
warranty information, maintenance details concerning hardware and
software, and a log of trouble tickets that can be used to monitor
fault patterns.

Storage Management

Mainframe standard storage management should be maintained

across the network. You should have automated archive facility that
can define retention policies and monitor the media and the statistics
on tapes and drive errors.

Resource Accounting
Enables you to monitor resource usage on the network for budgeting
and charge back. The CPU usage can be monitored along with disk
spaces and processes.

7 - 107
CS260 CHAPTER 7

Performance Monitoring

Monitoring CPU usage and I/O rates will provide you with valuable
information on system performance.

7 - 108
CS260 CHAPTER 7

Standard

♦ Standard must be a part of the design to allow operation in diverse

environments on different computing platforms. You will need to consider
implementing the following standards in your network :
• SNMP ( Simple Network Management Protocol) for network
interoperability
• SQL (Standard Query Language) for database access

• OLE (Object Linking and Embedding), an RPC for application

interoperability

• CORBA (Common Object Request Broken Architecture), an

RPC locator for application interoperability

• TCP/IP, IPX/SPX (Transmission Control Protocol / Internet

Protocol), (Internet Packet Exchange / Sequenced Packed
Exchange) for heterogeneous communications.

7.2 Installation and Configuration of a LAN

This section describes the installation and configuration of a Local Area

Network (LAN). The goal of a network is to connect computers and devices
such as printers, scanners, communication devices, storage devices, and
other peripheral components in such a way as to maximise the utility and
availability of resources to the user. A network can range from small
workgroups consisting of few personal computers (PCs) attached to a single
file server to large enterprise-wide networks encompassing hundreds or even
thousands of PCs sharing information with various file servers, as well as mini
and mainframe computers. Such large networks can link systems in different
cities or even different continents.

PC users today can choose from several widely supported operating system.
The most popular of these "Desktop Environments" include Microsoft DOS,
MS Windows, Apple Macintosh, UNIX and OS/2. A good Network OS should
support all of these operating systems. This means that the network can
allow users of dissimilar applications, computers and operating systems to
work together. While this section focuses mainly on the DOS environment, it
can be beneficial to all network users.

There are numerous hardware options for use with network. Literally
hundreds of makes and models of PCs can be used as network workstations
and many of them are suitable as file servers. There are a variety of popular
methods for interconnecting these computers and number of manufacturers
of products that support each method.

7.2.1 Hardware Installation

♦ Installing the LAN hardware is one of the most difficult tasks of
installing a LAN environment. It often involves hiring outside contractors
for cable installation and requires some knowledge about how the PC
works internally. The types of tasks involved are :

7 - 109
CS260 CHAPTER 7

• Configuring and installing LAN adapter cards for each PC and

server that will be part of the LAN.
• Installing cabling between PCs, servers, and other hardware
such as hubs and adapters.
• Installing one or more servers with hard disks and attached
shared printers.
• Installing proper power supply and power conditions.

Hardware Requirements for File Server

CPU - A PC (or PC compatible) Pentium
processor running more than 166 MHz.

RAM - Servers are always hungry for

memory. The more you have, the
faster the operation and better the
performance of the network.
Recommendation is 128 MB of RAM.

HARD DISK - A hard disk with sufficient storage

for your network.

NIC - One Network Interface Card.

CABLING - Network cabling (Ethernet, ARCnet,

10 Base T, Token Ring etc.)

MODEM - Install a Modem to allow server

management software with remote paging
to page you if a problem arises.

CD-ROM - A CD-ROM drive will provide facility installing software and if required to
access -ROM stored data.

BACKUP UNIT - Install a Backup Unit to safeguard against data damage.

Configuration and Installation of a Network Interface Card (NIC)

Read carefully the installation instructions for the NICs before

attempting to install them. Stick to the exact cable specified by the
manufacturer of the NICs. Using the wrong cable can prevent the
network from operating at all or, cause intermittent problems that
may be very difficult to trace back to the cable.

Before you can begin the installation process, you will have to take
off the covers of all your machines. You will have to find out what
kind of adapter cards are already installed and if these cards use
IRQs or DMA. Then you have to configure each LAN adapter cards
for the machine it will go into. You should record this information on
a worksheet. Although NICs are preconfigured at the factory, in

7 - 110
CS260 CHAPTER 7

many cases you cannot use the defaults settings. First you have to
change the DIP switches by moving shorting blocks around different
sets of pins on the LAN adapter boards. Then you have to change the
following settings if necessary.
♦ Interrupt Request Signal (IRQ) line

♦ Direct Memory Access (DMA) channel

♦ Base I/O Address

♦ Connector / Cable type

♦ Remote Reset

♦ Station Number or Node Address

♦ Base Memory Address

♦ Cable Mediums and Cable Topology

Three types of cable are used in LANs today: Twisted-Pair, Coaxial
Cable and Fibre Optics. The cable used is determined by the
manufacturer of the LAN adapter card hardware. In most LANs,
these mediums are not interchangeable. Each type of cable has a
different cost and different transmission characteristics that
affect LAN performance and the size of the LAN.

Topology is the scheme used to interconnect the PCs on the LAN.

The topology used is determined by the LAN adapter manufacturer.
The major topologies are bus, star, ring and tree. The topology can
influence the ease of installation of the LAN. In general, bus and
tree topology are the simplest in both cases; and star and ring are
the more difficult. Difficulties in installation and expansion are also
affected by the size of the physical facilities in which the LAN is
being installed.

Hard Disk
The file server must contain at least one hard disk drive. This
contains the Network Operating System's files and utilities as well
as the application software and data. The hard disk is prepared with
Network OS own format routine as part of the installation.

To protect data from the hard disk controller failure if you want to
duplicate data over one or more hard disk, you need an extra hard
disk.

Ensure Proper Power Supply and Power Conditions

1. Check the operating environment and power requirement for

network equipment in the following areas:

• Temperature / Humidity
• Power Consumption

• Power Frequency

7 - 111
CS260 CHAPTER 7

• Heat Dissipation

• Maximum Altitude
• Power Source

• Power Requirements

2. Use dedicated power lines and grounded outlets to connect

network components only.

3. Use power-conditioning equipment:

♦ Use Uninterruptable Power Supply (UPS) to protect server
hardware from power fluctuations with a regulating UPS. In
addition to protecting server hardware from damage caused
by power surges and voltage spikes a UPS protects data held
in RAM during a power failure.
♦ If using a UPS is not feasible, try to equip network hardware
with power conditioning devices such as Line-surge
suppressers or Ferro-resonant isolation transformers.
Connect network hardware through at least one of these
devices to protect them from minor power surge.

♦ Protect network equipment from static electricity by taking

following protective measures :

• Treat carpets with anti-static chemicals.

• Use protective covers for carpets such as anti-static
type.
• Ground equipment through a one meg. Ohm resister to
bleed off the static slowly.
• Make sure that the personnel working on open equipment
chassis take precautionary measures such as wearing
grounded wrist straps.

Documentation
♦ After installation and configuration of hardware, record the
following hardware information for future reference :
• File server : name, make and model

• RAM : Memory size

• Non-network boards : type and setting

• Network boards : associated LAN drivers, network number,

I/O address, memory address, interrupt, and station address.

• Floppy disk drives : diskette size and storage size

• Internal disks : make, model and storage size

• Disk coprocessor boards : DCB drivers and I/O address

7 - 112
CS260 CHAPTER 7

• Disk subsystems : number of drives, drive type, storage size,

number of heads and cylinders
• Mirrored disks

7.2.2 Loading and Configuration of Network Operating System

The network operating system provides a platform for network
services that combines network access with operating system
software. This is usually not application software but rather an
integrated operating system. The Network Operating System (NOS)
runs in the file server and controls system resources and information
processing for the network. The most popular Network Operating
System is Novell's NetWare (more than 70 percent) but there are
other manufacturers of NOSs. The major NOS are :
♦ Novell's NetWare

♦ Banyan VINES
♦ IBM LAN Server

♦ Microsoft Windows NT Advanced Server

♦ The following is the procedure for Installing of Network Operating
System (NOS) for Novell's NetWare which is most popular.
• Fulfil the hardware requirement

• Ensure proper power supply and power condition

• Set up hardware

• Make working copies if NOS comes in diskettes

• Decide file server booting method

• Run installation program for NOS

• Create and format DOS partition
• Name the file server
• Assign an IPX internal network number

• Load appropriate disk driver

• Create network disk partition table

• Mirror or duplex the disk (optional)

• Create and mount volumes

• Copy SYSTEM and PUBLIC files to the hard disk

• Load LAN driver with appropriate frame type

• Create file server boot files

7.2.3 Installation and Configuration of Workstations

♦ Prepare Workstation Hardware :

7 - 113
CS260 CHAPTER 7

• A IBM PC or compatible (XT, AT, 8088, 286, 386, 486 or

Pentium etc.)
• Ensure the workstation has 1.2 MB of free disk space

• Install the network board and keep record of the following

information:

- Base I/O address

- Interrupt (IRQ)

- Frame Type
- Node Address

• Install workstation client software

• Load appropriate LAN driver

• Install necessary files needed to connect to the network

• Customise network connection by editing configuration files

• Login to the network

7.3 Network Administration

The job of network administration - maintaining and operating the network,

falls to the Network Administrator. This is the person tasked with
overseeing the present needs of the organisation, predicting the future,
enhancing the network and administering day-to-day network operations. The
network administrator's job begins with the initial set-up of the network and
configuration of the hardware and software components to operate as a
network. The configuration of the network operating system (NOS) is one of
the first important steps in creating user environment.

7.3.1 Creating Directory Structure

The network administrator should establish and maintain a workable
directory structure. A server's disk is divided into hierarchy of
directories and files. The directory has a system of access and
function to promote the best utilization of
the resources by the users on the network. This organisation of the
directory system is important. It needs to not only to be logical for
the user but it must facilitate best use of the hardware and data
resources as well.

7 - 114
CS260 CHAPTER 7

During the installation process, NOS creates four system directories

: LOGIN, SYSTEM, PUBLIC and MAIL.

7.3.2 Setting Up User Accounts

Its a good idea to set up accounts with corresponding access rights.

The system default, created in the initial phase of setting up of the
network, that will apply to the majority of users on the network.
Default setting will apply to new users added to the network later
and facilitate the granting of rights and access to directories and
application and data files. You must also configure for access to
network peripherals.

7.3.3 Setting Up File Management Procedures

♦ Network administrator of the network have access and rights that

none of the users may have. These rights are required to perform the
following file management tasks :
- Adding, deleting or making other changes to the directory

- Creating sub-directories
- Defining directory and file ownership

- Naming trustee
- Modifying files

7.3.4 Allocating Network Resources

♦ You also need to allocate disk space for applications and users. Here
are some suggestion to consider :
• Group your network application together.

• Group your utility program together.

7 - 115
CS260 CHAPTER 7

• Be stingy with allocating user access and rights.

• Create public directories for network data files. Then

restrict access to those files so that only those users who
have access to the application files can use those data files.
• Make a place in the structure for user to keep private files.

• Allocate disk space to users to promote better storage

housekeeping.

7.3.5 Making Backups

Making backups is the process of creating duplicate copies of data to
protect against data loss. A network contains many system and data
files that need to be backed up. As a network administrator, you are
responsible for developing a backup strategy that is appropriate for
your particular situation.

7.3.6 Establish Routines

♦ As the network administrator, you must establish daily and weekly
maintenance routines. Tasks that should be done :
• Adding new users

• Cleaning up hard drives

• Installing upgrades

• Doing backup and archiving

• Restoring damaged or lost files and data

• Monitoring the network's traffic flow

• Troubleshooting

• Collecting data for accounting and optimisation

• Generating reports for management

7.3.7 Maintaining the Network

♦ A big part of administrator job is to maintain the network, these
includes :
• Establishing network documentation

• Providing support for the network user

• Developing training

• Managing the configuration of the network

• Experimenting with new technologies

7.3.8 Establishing Security Procedure

♦ Passwords and file/directory rights and access controls will provide a
level of protection from intruders into the network, but attention must
be given to physical security of the network components as well. The

7 - 116
CS260 CHAPTER 7

network file server and other network components such as bridges,

routers, terminal connectors and UPS system should be locked in a room
and access given only to the network administrator and maintenance
personnel. The best way to ensure server security is to practice good
configuration management by observing the following :

• Load the file server only from authorised distributed

software

• Make only authorised changes to configuration

• Maintain a copy of the current LAN configuration worksheet

• Periodically review device configurations

• Maintain a backup of the server software

• Test software before introducing it into network

7.3.9 Planning for the Future

♦ In addition to the preceding jobs, the person who is responsible for

the management of the LAN must keep abreast of new technology and be
able to respond to an ever-increasing user demand for the network. His
attention must be focused toward:

• Increasing network performance

• Assessing new hardware and software products

• Expanding network services

• Connecting with other networks

7.4 Review Questions

51. Why is planning important for the implementation of a Local Area
Network? Describe the steps in the planning process for the
implementation of a LAN.

52. What is the selection criteria for a network operating system?

53. When you design the physical layout of a LAN, what are the factors
that you must consider?
54. Describe the detailed procedures for the installation and
configuration of a LAN.
55. Describe the role of a network administrator during the installation
and configuration of a LAN.
56. Describe the responsibilities of a network administrator in the post-
implementation of a LAN.
57. Is network topology important in the implementation plan? If yes,
then explain the reasons.
58. How do you safeguard your network equipment against damages from
power disturbances or from intruders?

7 - 117
CS260 CHAPTER 7

59. What are the security measures you should apply in a LAN
environment?
60. Describe the hardware and software requirement for the file server
and workstation.

7 - 118
CHAPTER 8 : TELECOMMUNICATION SERVICES AND
NETWORK PRODUCTS
Chapter Objectives

Upon completion of this chapter, you will be able to:

⇒ learn about different types of Telecommunication networks;

⇒ learn the typical telecommunication services provided by a telecommunication service provider and apply
the concepts in real life;

⇒ learn and apply the concepts of the Router, Bridge and Gateway;

⇒ learn and apply the concepts of StarLAN 10 and IBM Token Ring network.

8.1 Types of Telecommunication Networks

There are many different ways to define various types of telecommunication networks. This matrix may be
used for a simple classification. In one dimension, we may define the networks by application such as voice or
data services. The other dimension defines the domain by identifying whether the network is public or
private.

The public switched telephone network (PSTN) is the most familiar one. It provides the voice services as well
as voice grade data services. Data services in the public domain include the circuit-switched public data
network (CSPDN) and the packet-switched public data network (PSPDN).

In the private domain, the private line networks (PLN) may provide either voice or data services. They are
generally used by big corporations. These networks may be configured either as switched services or non-
switched services. Non-switched services are also referred to as dedicated services, nailed-up services,
channelised services, or simply special services.

Local area networks (LAN) evolved from computer networking. They are generally limited in a small
geographical area and transmit high-speed data. There are special considerations with respect to the
topology, access mechanism and protocol. At the centre of the diagram is the Integrated Services Digital

Figure 8-1

8 - 119
CS260 CHAPTER 8 : TELECOMMUNICATION SERVICES AND
NETWORK PRODUCTS

Network (ISDN), where voice and data service (either private or public) may be integrated by means of
standardized interfaces.
8.2 Services Provided by Telecommunication Service Providers

Typically telecommunication service providers offer a wide range of services which include telephone as well
as data communication related services. Some of the typical services provided are discussed below:
♦ Domestic and International Telephone Services
The user can use a modem for data communication using this service.
♦ Digital Leased Line Service
This digital leased line service is specially engineered for private data communication networks. Its
applications are wide ranging- voice cum data communication, disaster recovery networks, high-speed
remote printing and digital PBX network. Digital leased line can support different speed for the users
which can range from 2,400 bps to 140 Mbps or more. When the subscriber subscribes to this service, he
has to install a digital interface unit which is the interface between the subscribers computer and the
digital leased line. Once a subscriber subscribes to this service, the line becomes a dedicated link and no
modem is required for data communication. Typically these line will be subscribed by companies who want
to have their own network of computers. For this type of service, service providers charge the following:
• Installation charge of digital interface unit;

• Monthly rental;
• Monthly access charge.

♦ Analog Leased Line Services

Analog leased line services are similar to digital leased line services. The only difference is that the
subscriber needs to use modem for data communication. Subscribers will be charged based on monthly
rental and monthly access. Typically these sorts of services will be used by companies where transmission
speed required is minimal.
♦ Packet Switch Service

Typically the service provider will have a packet switched network. Access to this packet switched
network is via leased circuit (up to 64 Kbps) or dial-up(2.4 Kbps). A subscriber to this type of packet
switched network can have the following types of connection:
• Public Dial-up Access:

The subscriber is connected to the packet switched network exchange via the public switched
telephone network (PSTN). For connection, the subscriber needs to simply dial the public port number
of the packet switched network. Each subscriber is given a confidential user name known as Network
User Identity. It serves as a password for access.

• Private Dial-up Access:

Similar to the public dial-up access except that the subscribers are given a dedicated port number to
dial in. The number is exclusive for their use.
• Dedicated Access:

The subscriber is connected to the packet switched network exchange by a leased line. The
subscriber occupies a dedicated port of the packet switch network which is known as Network User
Address. This is to facilitate receiving calls from others.
• Telex Access:

8 - 120
CS260 CHAPTER 8 : TELECOMMUNICATION SERVICES AND
NETWORK PRODUCTS

The subscriber establishes a connection to the packet switched network by calling the telex access
port using a telex machine. Similar to the public dial-up access, the subscriber requires a Network
User Identity.
♦ The subscriber to the packet switched network typically needs to pay for the following:
• Monthly subscription charge (depends on the type of connection)

• Usage charge (includes volume and duration)

• Line rental charge (includes line installation and monthly rental)

8.3 Devices for LAN Interconnection

When two systems are not directly connected to each other and connectivity is through an intermediary, the
intermediary is called a relay in the ISO terminology. There can be many relays on the path between two
systems. OSI reference model uses a seven-layer protocol suite for the network architecture. If a relay
shares a common layer k protocol with other systems, but does not participate in a layer k+1 protocol in the
process of relaying information, the relay is known as layer k relay.

One of the most basic interconnection devices is a repeater. This is layer one relay device i.e. it operates at
the physical layer. It can not control or route information. A bridge operates at the data link layer. Typically,
it looks at the packet address to see if it is destined for another subnet. A router on the other hand
operates at the network layer, has more intelligence capabilities because it can handle several levels of
addressing. A gateway operates at the top four layers of the OSI reference model. It can interconnect
networks or media with different architectures and protocols. We will discuss each of these basic types in
the next few pages.
Not all vendor products fit neatly into these basic categories. Many vendors have developed hybrid products
that perform functions traditionally associated with two or more of these categories. For instance, a router
performs the functions of a bridge and also has some router capabilities.

To add to the confusion, the word gateways, bridges, and routers are often used in generic sense. What
Novell's Netware refers to as a bridge is a router in OSI terminology. A gateway in TCP/IP terminology is a
router in OSI terminology. Thus one has to be careful in reading trade magazines and vendor information.

8.3.1 Bridges

A bridge is a device that operates at the data link layer and relays frames (packets) between LANs. A
bridge has a number of physical ports and, at most, one sub network of an internet is connected to
each such port of the bridge. Although many bridges operate on networks of similar architecture,
such as Ethernet to Ethernet, or token ring to token ring, bridges can also be between LANs with
dissimilar MAC sub layer protocols as long as they follow same LLC protocol. For instance, a bridge
can connect an Ethernet LAN with a Token Ring LAN since they share a common 802.2 LLC format.
However, such interconnection between dissimilar networks often require a higher level protocol
conversion than provided by bridges. Therefore, bridges connecting dissimilar LANs could present
reliability problems.

A bridge does not provide flow control and there can be congestion within a bridge. To deal with
congestion, bridge must have the ability to temporarily buffer frames. Protocols at the higher layers
must then deal with the issue of recovering frames that are either lost or discarded by a bridge when
its buffer gets full.

8 - 121
CS260 CHAPTER 8 : TELECOMMUNICATION SERVICES AND
NETWORK PRODUCTS

Figure 8-2

♦ Bridges are commonly used to interconnect LANs of same type (e.g. all coax Ethernets) or to break up
a large single LAN into a number of smaller interconnected LANs that are all of the same type. The break
up of a single LAN into interconnected LANs is often desirable for the following reasons:
• The performance of a single LAN deteriorates rapidly when the total network traffic
increases beyond a certain level. By breaking up the LAN into smaller LANs in which the major traffic
is within each smaller LAN with some traffic between LANs, as a result delay and throughput of each
individual LAN and the performance of the network for the user can be significantly improved.
Bridges are the main vehicles used for interconnecting the individual smaller LANs (all of the same
type) in such cases.
• A group of interconnected smaller LANs is less vulnerable to failures than a single LAN. If a
component LAN fails, the rest of the internet can continue to function, perhaps at a decreased level
of effectiveness whereas a failure in the case of a single larger LAN spanning the whole organisation
can bring the operation of the whole organisation to a halt.
• Many organisations are wary of sensitive information such as personnel or financial
information, being placed on the LAN. Isolating such information on smaller LANs and providing
interconnection to other LANs through bridges makes the security issue more manageable. The
bridges can ensure that only authorised users have access to sensitive information on the LANs to
which they are connected.

8 - 122
CS260 CHAPTER 8 : TELECOMMUNICATION SERVICES AND
NETWORK PRODUCTS

Figure 8-3

Consider a bridge connecting LAN X and LAN Y. When the bridge receives a frame from LAN X, it either
forwards the frame to LAN Y (i.e. allows to pass through) or filters it (i.e. does not allow it to pass through).
A bridge must therefore make a decision on whether to forward/filter a frame. When a bridge is connected
to more than two LANs, it must also decide to which LAN or LANs the frame will be forwarded if the frame
is not filtered.
♦ There are two methods commonly used for making such decisions. These are:
• Transparent bridging method

In this approach the bridges of the internet carry the required "intelligence" to make the decision
about forwarding a frame. The forwarding mechanism is thus transparent to the communicating
stations. The bridge has a greater processing responsibility in this method and the stations have an
easier task since they need not concern themselves with the bridges' decisions.

• Source routing method

In this approach the source station makes the decision as on how to route the frames through the
bridges. The bridges merely carry out the routing determined by the source. Thus in this method the
processing responsibility on the source station is increased while the task of the bridges is simpler.

8.3.2 ROUTERS
In many cases, an organization may need to access devices on various networks. For instance, LANs may have
a hierarchical architecture. There may also be a need to access devices on a WAN over a public or private
switched network, for example, to access public databases. For such purposes, a suitable internet working
device is a router.

8 - 123
CS260 CHAPTER 8 : TELECOMMUNICATION SERVICES AND
NETWORK PRODUCTS

System X Router
System Y

Figure 8-4
♦ A router is a device operating at the OSI network layer (i.e. layer three) and can be used to provide
internetworking among dissimilar networks. The routers can accommodate several differences among the
networks that they can interconnect. Some of those are listed below:
• Different addressing schemes: Since bridges generally use a flat addressing scheme, they
need a global addressing scheme for the entire internet. In contrast, typically routers use a
hierarchical addressing scheme. In the simplest scheme, the address of a station of the internet with
routers is a pair of the form (subnetwork, address within subnetwork). Thus the subnetworks can use
different addressing schemes as long as there is a global way of identifying the subnetworks.

• Different maximum packet sizes: Different subnetworks have different restrictions on the
maximum packet sizes they can handle. Thus, the router must break up packets from one subnetwork
into smaller packets for another subnetwork. This process is referred to as segmentation. At the
destination, packets that have been segmented will have to be reassembled by the network layer
before it can be passed to the higher layer.
A router must share a common network protocol with the stations and routers that are directly
attached to it. In addition, for successful communication between end stations, the stations must
share the same protocols above the network layer.

8.3.3 Gateways
Gateways are devices that may operate at all seven layers of OSI. They may be responsible for
connecting incompatible proprietary networks, electronic mail systems, converting and transferring
files from one system to another, or enabling interoperability between dissimilar operating systems
or database management systems.
♦ Some sample gateway products are:

• Gateways between networks such as SNA and DECnet, or between SNA and TCP/IP; and

8 - 124
CS260 CHAPTER 8 : TELECOMMUNICATION SERVICES AND
NETWORK PRODUCTS

• LANs in different organisation or even within the same organisation may be using different e-
mail systems such as cc:Mail, uucp and EasyLink. These systems have differences in the formats,
addressing schemes, routing, etc. In order to provide transparent connectivity between such
dissimilar e-mail systems, e-mail gateway products such as SoftSwitch Central have been developed.
Protocol standards such as X.400 help in the conversion process.

Figure 8-5

8.4 LAN Products

♦ StarLAN
10StarLAN 10 is a local area network with a star wiring topology in which each computer is connected to a
central hub. The advantage of this topology is that it is simple to install. Also, if one network device fails,
the rest of the network is unaffected. The network hardware is supported with software. Thus meeting
the requirements for quick and easy access to shared information and resource. The platform is built on
standard that combine the strengths of UNIX and MS-DOS to support integrated, portable applications
on an open multivendor architecture.
AT&T offers two versions of StarLAN Network hardware: a 1Mbps version for smaller configuration and
a 10Mbps version for larger networks and higher traffic volume. The 10Mbps version also operates over
optical fibre and is compatible with Ethernet systems. The discussion will be based on StarLAN 10.
♦ Features and Application
The StarLAN network offers a complete solution with the hardware, networking, and application
software. The StarLAN 10 network can connect up to 12 devices. Fibre optics is used for distant
connections and to address security concerns. Network users may access resources such as applications
software and files up to 15.9 Gbytes through UNIX operating system services. Asynchronous
communication facilities and terminal emulation software allow users to access both on-net and off-net
hosts. StarLAN 10 network gives a wide range of connectivity options for implementing networking
strategies. Users may access the network from a remote PC. Networks may be bridged to one another,
linked to computer networks, or able to access remote networks via X.25 facilities. The networks are

8 - 125
CS260 CHAPTER 8 : TELECOMMUNICATION SERVICES AND
NETWORK PRODUCTS

compatible with Ethernet. Applications and Communication software and Network Management package
are other features offered with the StarLAN 10 network.
• IBM Token-Ring Network
IBM Token-Ring network is IBM's major entry into the LAN market and a critical element in their overall
data communications strategy. It will serve as one of the major building blocks of IBM Networks in the
next couple of decades.
• The token-Ring is a star-wired ring topology.

• The protocol it uses is the token passing protocol of the IEEE 802.5 standard.
• It utilises shielded or unshielded twisted pair, or optical fibre.

• It uses baseband signalling and can operate at either 4MB or 16 MB depending on the type of
adapter card used.

• It will support 72 stations on unshielded wire pair or 260 stations on shielded pair on a single
LAN.

• PC LAN program in the DOS environment and the OS/2 Server/Requester software on OS/2
stations are the main software packages used with the Token-Ring to handle the client/server
functions.
• The LAN Manager program is used for the network management functions.

• A network station can serve as a bridge to another token ring using IBM Token-Ring network
bridge program. Several LANs can be joined with a backbone LAN using bridges.

8.5 Review Questions

61. Describe the different types of telecommunication networks.

62. What are the typical services provided by telecommunication service providers?
63. Why is a bridge used in a network?

64. What is the difference between transparent bridging and source routing method?
65. What are the functions of a router?
66. What is the difference between a bridge and a router?
67. Why do we need a Gateway?

68. Write down some of the features of the IBM Token Ring network.
69. What are some of the considerations for choosing a particular LAN?

8 - 126