COMMUNICATION NETWORK

A communication network, consists of a set of equipment and facilities that provide a

service, the transfer of information between users located at various geographical points.
For example telephone networks provides telephone service, computer networks,
television broadcast networks, cellular networks and Internet.

Communication network and transportation networks have become essential

infrastructure in every society.

The capacity to transfer high volume of data for a long distance almost immediately are
the features of network-based services. For example Internet provides e-mail,
information search and retrieval, various forms of electronic service.

EVOLUTION OF NETWORK ARCHITECTURE AND SERVICES

A communication service which is used of transfer of information. Different services
differ in the details of how and in what form information is transferred.

Let us consider three example networks

i) Telegraph networks
ii) Telephone networks
iii) Computer networks

TELEGRAPH NETWORKS AND MESSAGE SWITCHING

In the year 1987 Samuel B. Morse demonstrated a practical telegraph that provided the
basis for telegram service, the transmission of text message over long distance. Here the
text was encoded into sequence of dots and dashes. Each dot and dash which is used to be
represented by short and long pulses of electrical current over a copper wire. By relying
on two signals, telegraphy made use of digital transmission system. In Morse code the
pause between letter was 3 dots and for word is 5 dots.

In store-and-forward process, the message is completely received and stored at an

intermediate point and then forward to the next node depending on the availability of
link. The process of selecting an appropriate link is known as routing. If the information
is message, then it is known as message switching. In simpler terms, message switching
is a switching method that sends data from point-to-point with each intermediate node
storing the data , waiting for a free transmission channel and forwarding the data to the
next point until the destination is reached.

In the beginning communication channel used to carry information sent by one person.
Naturally transmission rate (in bits/second) was less. To increase the transmission rate
multiplexing was developed. Multiplexing is a technique used to place multiple signals
on a single communication channel. Multiplexing partitions a channel into many separate
channels, each capable of transmitting its own independent signal, thereby enabling many
different transmission over a single medium.

One approach of multiplexing involves modulation. Binary symbols can be transmitted

by sending a sinusoidal signal of a given frequency for a given period of time. Multiple
sequence of binary symbols could be transmitted simultaneously by using multiple pairs
of frequencies for the different telegraphy signals. This concept is used in Modems.

We know characters can be represented by ASCII ( American Standard Code for

Information Interchange) , is a 7 bit code. The equipments can communicate with ASCII
code.

The telegraph service involves the transmission of text messages between geographically
far places. To provide the service , the architecture of the telegraph network contains the
following key elements or functions.

i) Digital transmission takes in the network that is information is represented

either 0 or 1. It can make use of ASCII code also. The transmission medium
can be copper wire and radio.
ii) To identify the messages exactly, messages are covered by beginning and
ending by sequence of characters. This is known as framing.
iii) There should be destination address that is to whom it want to deliver.
Routing procedure determines the path that a message follows across a
network of telegraph stations interconnected by digital lines

TELEPHONE NETWORKS AND CIRCUIT SWITCHING

In 1876 Alexander Graham Bell developed a device that can transmit voice signals. The
device that is known as Telephone. The telephone network provides a two-way, real-time
transmission of voice signals across a network.

Telephone service became popular due to its voice service and expert operator with
knowledge of Morse code is not required like in the case of telegraph. Here voice signals
are converted into equivalent electrical signal and passed through cable.
 COMMUNICATION FUNDAMENTALS

Communication concepts

Communication is the process of transmitting information. It’s a big, vague, fuzzy word
and broadly applied. The “Communications Department “ at a university may include
speech teachers, public relations specialists, writers and broadcasters, who have nothing
to do with hardware. By that definition , a writer is in the communication business.

Evolution of communications

We the human beings are communicating from early days that is by drawing the pictures
on cave walls. It can be how to attack the animal or how to the save the life from the
threats or it may be a route to go to another cave.

The earlier long-distance communications was by signal fires that relayed simple
information. The signal used to represent “yes” or “no” – the question being asked had to
be known beforehand. It’s only a single bit information, not a detailed message.

A written letter or a human messenger could carry more information. The first systems
we might call telecommunications were series of hilltop towers, built by French engineer
Claude Chappe in the 1790s. The towers had to be in sight of each other, with an
operator in each one. The operator relayed a message by moving arms on top of the
tower. The operator of the next tower looked through a telescope to watch the arms
move, reading the message, recording it, then moving the arms of his tower to relay it to
the next tower. Chappe invented a code that used arm position to indicate letters or
certain common words. His system was called an optical telegraph.

Samuel Morse’s electrical telegraph eventually replaced Chappe’s optical one. It also
required operators, but they used keys that completed an electrical circuit, sending bursts
of currents through a wire. The Morse code was a series of dots an dashes ( short and
long pulses) that represented letters.

The electrical telegraph spread across the continents and in 1866 across the Atlantic. Its
wires formed a network running between major cities. Telegraphers received signals and
either sent them for local delivery or relayed them to more distant stations. We can think
of each dot and dash of the telegraph as a bit of information.

Later Alexander Graham Bell developed telephone. He borrowed some principles from
the telegraph , but instead of transmitting dots and dashes, it sent a continuously changing
electrical current that represented a speaker’s voice. The telephone network was bigger
and more complex than the telegraph network.
After that radio waves are used to carry signals through air. Radio first carried telegraph
signals. Wires worked fine on land, but only radio could send telegrams to ships at sea
and relay urgent messages such as request for help. It was radio rescue calls that sent
ships to help the survivors of the sinking of the Titanic. Radio communications started
for ships and airplanes in the world war II.

Radio communication started at low frequencies, but gradually moved to higher ad higher
frequencies as electronics improved. The higher the frequency the more information the
signal can carry. Pictures need more transmission capacity than sound alone, so television
are broadcast at higher frequencies than audio radio.

By the 1970s , satellites were beaming radio signals around the globe. The telephone
system had become global and known popularly “telecommunication network” . We
could make phone calls to anywhere in the world.

After that fiber optics arrived. Fiber provides high bandwidth. It carries high capacity
information at faster rate. Now it is a integral part of telecommunication network.

Signals and systems

Information is the one transmitted via communication. It may be very simple message
like “yes” or “no”. It can be huge and complex message, such as he files containing the
book or a television broadcast of a movie. It doesn’t matter if the “information” doesn’t
contain anything a critic might agree was “information”.

A signal transmits that information. Signals may take many forms, such as optical,
acoustic, electronic or radio-frequency. Signals may be converted from one form into
another and still contain the same information. When we make a long-distance telephone
call, the sound waves from our mouth are converted to an electrical signal at the
telephone, that electrical signal is converted to optical form at the telephone switching
office, back to electrical form at the local switching office on the other end and back to
sound waves at the other person’s telephone.

A system is the equipment that performs a task, like transmitting signals. We often speak
of telephone system as if it is the whole network of telephone equipment from our phone
to the high-capacity fiber-optics cables that carry long-distance calls. However, parts of
the telephone system also may be called systems, such as a switching systems that directs
phone calls. We can assume a system which is a collection of stuffs designed to work
together.

Basic components of data communication system

A data communication system is made up of five components.

 i) Message
ii) Sender
iii) Receiver
iv) Medium
v) Protocol

Message - the message is the information to be communicated

Sender – The sender is the device that sends the data message. It can be computer,
workstation, telephone etc.,
Receiver – The receiver is the device that receives the message. It can be a computer ,
workstations, telephone, fax machine
Medium – The transmitting medium is the physical path by which a message travels from
sender to receiver . It can consists of twisted pair wire, coaxial cable, fiber optic cable or
radio waves.
Protocol – A protocol is a set of rules that govern data communication. It represents an
agreement between the communicating devices. Without a protocol two devices may be
connected but not communicating , just as a person speaking Malayalam cannot be
understood by a person who speaks only Kannada.

Protocols and standards

The Communication occurs between entities in different systems. An entity is anything

capable of sending or receiving information. But two entities cannot just send bit streams
to each other and expect to be understood. For communication to occur the entities must
agree on a protocol. The key elements of a protocol are syntax, semantics and timing.

Syntax : Syntax refers to the structure or format of the data, meaning the order in which
they are presented. For example , a simple protocol might expect the first eight bits of
data to be the address of the sender , the second eight bits to be the address of receiver
and rest of the stream to be the message itself.

Semantics : Semantics refers to the meaning of each section of bits. How is a particular
pattern to be interpreted and what action is to be taken based on that information.

Timing : Timing refers to two characteristics. When data should be sent and how fast it
can be sent. For example , if a sender produces data at 100 Mbps but the receiver can
process data at only 1 Mbps, the transmission will overload and receiver and data will be
largely lost.

Standards

Standards are essential in creating and maintaining an open and competitive market fro
equipment manufacturers and in guaranteeing national and international operational
ability of data and telecommunication technology and processes. They provide
guidelines to manufacturers, vendors, government agencies and other service providers to
ensure the king of interconnectivity necessary in today’s marketplace and in international
communications.

Data communications standards fall into two categories

1. de facto ( by fact)
2. de jure ( by law)

De jure standards are those that have been legislated by an officially recognized body.

De facto standards are often established originally by manufacturers seeking the

functionality of a new product or technology.

Standards organization

Standards are developed by cooperation among standard creation committees forums and
government regulatory agencies.

i) The International Standards Organization (ISO)

ii) The International Telecommunication Union (ITU-T)
iii) ANSI (American National Standards Institute)
iv) The Institute of Electrical and Electronics Engineer (IEEE)
v) The Electronic Industries Association (EIA)

The ISO is an organization dedicated to worldwide agreement on international

standards in variety of fields. ITU-T is an international standards organization related
to the United Nations that develops standards for telecommunications. Two popular
standards developed by ITU-T are the V series and X series. ANSI , a nonprofit
organization is the United States, voting representative to both the ISO and ITU-T.
The IEEE is the largest national professional group involved in developing standards
for computing , communication, electrical engineering and electronics. It sponsored
an important standard for Local Area Network (LAN) called project 802. EIA is an
association of electronics manufacturers in the United States. It is responsible for
developing the EIA-232-D and EIA-530 standards.
System Functions

Communications systems serve many functions. It is possible to divide into four

categories on functions.
a) Point-to-point
b) Broadcasting
c) Switched
d) Networked

Point-to-point

A point-to-point configuration provides a dedicated link between two devices. The entire
capacity of the channel is reserved for transmission between those two devices. Most
point-to-point line configuration use an actual length of wire or cable to connect the two
ends.

Broadcasting

A broadcast system send the same signal to everyone who receives it. In its usual form ,
transmission is one way, from the signal source to the individual. Local radio and
television transmission are good examples of pure one-way broadcast systems.

Switched systems

A switched system makes temporary connections between terminals so they can

exchange information. The telephone is a good example.

Networking

A network connects many terminals that can communicate with each other. For example
Local Area Network is belongs to this category.

Services

We know the operation of modern communication networks is a very complex process

that involves the interaction of many systems. A communication network , is a set of
equipment and facilities that provides a service, the transfer of information between users
located at various geographical points. The ability of communication networks to transfer
communication at extremely high speeds, allows users to gather information in large
volumes almost instantly with the help of computers for long distances.

Radio and television broadcasting are probably the most common communication
services. Various stations transmit an ensemble of signals simultaneously over radio or
cable distribution networks.
Telephone service is the most common real-time service provided by a network. Two
people are able to communicate by transmitting their voices across the network. The
service is considered as connection-oriented in the sense that users must first interact with
the network to set up a connection

Telephone networks provide a broad class of call management services that use the
originating number or the destination number to determine the handling of a call. For
example , in call return the last originating number is retained to allow it to be
automatically called by the destination user at a later point in time. Caller ID allows the
originating number and sometimes name, of the originating call to be displayed to the
destination user when the receiving device is display capable. Voice mail allows a
destination user to have calls forwarded to a message receiving device when the
destination user is not available.

Cellular telephone service extends the normal telephone service to mobile users who are
free to move within a regional area covered by an interconnected array of smaller
geographical areas called cells. Each cell has a radio transmission system that allows it to
communicate with users in its area. Cellular system should handle “handoff” when user
moves from one area to another area.

Electronic mail (e-mail) is another popular network service. The user provides a text
message and a name and/or address to a mail application. The application interacts with a
local mail server, which in turn transmits the message to a destination server across a
computer network. The destination user retrieves the message by using a mail
application. E-mail is not a real-time service , where large delays can be tolerated. Here
security and privacy is matter of concern.

Many applications that involve an interaction between processes running in two

computers may be characterized by client/server interaction. Internet is an example of
distributed system. The World Wide Web (WWW) is accessed through a browser
program that displays the document and allows the user to access other documents by
clicking the links. Each link provides the browser with a uniform resource locator (URL)
that specifies the name of the machine where the document is located as well as the name
of the file that contains the requested document.

Video on demand characterization another type of interactive service. The objective of

the service is to provide access to a video library. The user initiates the service by
accessing a menu from which a selection is made. Video which requires lot of bandwidth
to pass from source computer to the destination.

Streamed audiovisual services over the internet provide an example of a service with
some of the feature of video on demand. An application like “real player “ can be used to
access a channel that provides an audio visual stream to the client machine.
Another class of service is audio-visual conferencing. Here both audio and video are
transmitted over a distance. This service requires real-time response requirement. This
service is being used by many companies for training, client interaction from a site, group
discussion etc.

Broadcast television

Satellites TV Disk

Mobile Phone Standard Telephone

LAN TV Cable Television

Broadcast
Radio
P.C.

FAX Fax Phone Line

PROTOCOL

Step 1
Step 2

CHANNEL

SOURCE COMPUTER DESTINATION COMPUTER

 Chapter 5

Computing environments
5.0 Introduction

We know, in information technology era, sharing of resources and easy communication

are acting as a backbone of any network. Popular example is our Internet. With the help
of internet we are able to exchange information and share resources. Hence in this
chapter we study different advantages of the network and different forms of the network
and their features.

The learning objectives of this chapter are to know

What is a computer network ?
How network is advantageous ?
What are different forms of networks ?
What is a e-mail ?

5.1 Networking of computers and its advantages.

Computer network is defined as an interconnection of autonomous

computer. Here autonomous means, there is no master and slave relationship. All
computers are equal. Computer network enables to share the resources. Computer
networking also refers to connecting computers to share data, application software and
hardware devices. Networks allow sharing of information among various computers and
permit users to share files. For example a students accesses compilers sitting at one place,
where compiler may be stored on the other machine. The students takes printout with the
help of one printer connected to the network. The printer can be shared among many
students.

Network offers the opportunity to communicate more efficiently with

others through electronic mail. Networks allow companies to share software and
peripherals such as printers, plotters, scanners and so on. With networking all the
computers in an office can be connected to a single printer and scanner. It also helps in
using storage devices efficiently.

Computer network acts as a very powerful communication medium. It

means people exchange their information. When compared to mainframe computers,
network of computers saves money.

5.2 Types of networks

Depending the nature of the distances, protocols ( the set of rules used for
communication) the network can be classified into LAN (Local Area Network ), MAN
(Metropolitan Area Network ) and WAN ( Wide Area Network).
Local Area Network (LAN)
A LAN (local area network) is a group of computers and network devices connected
together, usually within the same building. By definition, the connections must be high-
speed and relatively inexpensive (e.g., token ring or Ethernet). They function to link
computers together and provide shared access to printers, file servers, and other services.

Any individual computer connected to a network is called workstation. A workstation

may not need a floppy disk or hard disk. A LAN or local area network connects
computers and peripherals in a limited area. LAN requires cables to connect
workstations. For example LAN is used in a hall or within a building. Figure 5.1 shows
Local Area Network , where various departments are connected.

Fig 5.1 Local Area Network

Metropolitan Area Network (MAN)

A MAN ( Metropolitan Area Network) is used to connect computers to cover the city or
town. The range may be approximately 50 Kilometers. Normally cables and fiber optic
cables are used to connect computers. The routing of the messages are fast. Normally
central library in a city may be connected by a MAN, so users can access the information.
Figure 5.2 shows a typical view of Metropolitan Area Network.
Figure 5.2 Metropolitan Area Network

Wide Area Network (WAN)

A WAN (Wide Area Network) covers large distance like state, country or
continents. The WAN uses the fiber optics, cables and even satellites also. Here
communication circuits are connected with the help of hardware device called routers.
Routers forward small pieces of information called packets from one to another. Internet
is the popular one comes under WAN. Some of the examples makes use of internet are
reservation of airplane tickets, railway tickets and even cinema tickets. Another facility
called e-commerce, where business is carried out through internet. Here people can buy
books, articles and so on through registering their wants through the internet. Figure 5.3
shows a typical view of Wide Area Network.

Fig 5.3 Wide Area Network

5.3 Internet
The internet is a massive wide area network ,connecting thousands of computer
networks around the world. The internet is a world wide “network of networks”. It is a
collection of thousands of smaller networks in different countries around the world. It
links thousands of academic government, military and public computers, enabling
millions of people to share information and other resources.
 Internet pathways are used to exchange digitized computer data. The basic
services that form the foundation of the internet are e-mail , telnet and FTP. With internet
we can easily exchange electronic mail with friends and family anywhere in the world.

Telnet allows you to connect to a remote computer. We can access any of the
public services or tools and library databases at the remote site. FTP(File Transfer
Protocol) provides for transferring files from one computer to another across the internet.

Internet has many uses. For individuals, the most important uses of internet are e-
mail and surfing the Web. One can read the topics of interest like sports, a hobby, a
country or any place of interest.

E-mail:
The e-mail stands for electronic mail. One of the major features of computer
networking is that messages can be sent electronically to various terminals on the
network. The messages are sent very quickly and accurately. E-mail uses the concept of
Storing and forwarding messages. It saves a lot of money for the users. Here user
registers his/her account with one of the providers. The e-mail account normally contains
username and the provider name. For example abc @ yahoo.com , represents abc is the
name of the user , who is registered in yahoo provider.

5.4 Summary:
• Computer network is defined as a interconnection of autonomous computers.
• Computer networking enables us to connect computers to share data, application
software and hardware devices.
• Networks allow sharing of information among various computers and permit
users to share files.
• A LAN or local area network connects computers and peripherals in a limited
area.
• A MAN covers the city or town.
• A Wide Area Network covers a large area.
• The internet is a massive wide area network, connecting thousands of computer
networks around the world.
• E-mail is a cost effective communication

5.5 Quiz questions

1. Computer network defined as a interconnection of ---------- computers

2. Computer network allows user to ----- the resources
3. -------- network confined to hall or building.
4. ------- is one example of WAN.
5. ------ business is carried our through internet.
6. Small piece of information called ---- are forwarded by router in Internet.

Answers
1. Autonomous
2. share
3. Local area
4. Internet
5. E-commerce
6. packets

5.6 Exercise

1. What is a computer network ? What are the advantages of computer network ?

2. Briefly explain the Local area network, Metropolitan area network and wide area
network.
3. What e-mail ? How e-mail works ?
4. What are the uses of Internet ?
 Applications and Layered architecture
Communication network must support wide range of services. Normally people use
networks to communicate, send e-mails, transfer of files and so on. Industry people
use communication network for transfer of funds, update information about the
product and so on. Hence, to provide support for current service and future services, a
complete plan is required. This necessitates developing a complete flexibility in
network architecture.

communications functions are grouped into the following tasks

• The transport across a network of data from a process in one machine to the
process at another machine.
• The routing and forwarding of packets across multiple hops in a network
• The transfer of a frame of data from one physical interface to another.

To reduce their design complexity, most networks are organized as a series of layers
or levels, each one built upon its predecessor. The number of layers, the name of the
each layer, the contents of each layer and the function of each layer differ from
network to network.

Interaction between the layers must be defined precisely. Interaction is done with
definition of the service provided by each layer and to the layer above. Interface
between layers through which a service is requested and through which results are
conveyed. New services that build on existing services can be introduced even at the
later stage. The layered approach accommodates incremental changes readily.

We know , in all networks, the purpose of each layer is to offer certain services to the
higher layers. The entities comprising the corresponding layers on different machines
are called peer processes. Between each pair of adjacent layers there is an interface.
The interface defines which primitive operations and services the lower layer offers to
the upper one. The set of layers and protocols is called the network architecture.

A protocol is a set of rules that governs how two or more communicating devices are
to interact. HTTP protocol enables retrieval of web pages and TCP protocol enables
the reliable transfer of streams of information between computers.

HTTP

Let us consider a client/server architecture, a server process in a computer waits for

incoming requests by listening to a port. Port is an address that identifies which
process is to receive a message that is delivered to a given machine. The server
provide response to the requests. The server process always runs a process in the
background called daemon. httpd refers to server daemon for HTTP. The documents
are prepared using Hyper Text Markup Language (HTML) which consists of text,
 graphics and other media are interconnected by links that appear within the
documents. The www is accessed through a browser program that displays the
documents and allows the user to access other documents by clicking one of these
links. Each link provides the browser with a uniform resource Locator (URL) that
specifies the name of the machine where the document is located and the name of the
file that contains the requested document. The HTTP ( Hyper Text Transfer Protocol )
specifies rules by which the client and server interacts so as to retrieve a document.

In HTTP, we use two-way connection that transfer a stream of bytes in correct

sequential order and without errors. The TCP protocol provides this type of
communication service between two processes in two machines connected to a
network. Each HTTP inserts its messages into a buffer and TCP transmits the
contents of the buffer to the other TCP in blocks of information called segments.
Each segment contains port number information in addition to the HTTP message
information. The following figure shows how communication is carried between
HTTP client and HTTP server.

Courtesy : Data communication networks by Alberto Leon - Garcia

DNS query

First of all, client needs to perform a DNS query to obtain the IP address
corresponding to the domain name. This step is done by sending a message to a DNS
server. The Domain Name System (DNS) is a distributed database that resides in
multiple machines on the Internet and is used to convert between names and
addresses and to provide e-mail routing information. Each DNS machine maintains its
own database and acts as a DNS server that other query machine can query. Initially
requesting machine accesses a local name server. In case if it not possible to resolve
names then it will be sent to the root server. When the root server is unable to resolve
, it will be sent to authoritative name server. Every machine on the Internet is required
to register with at least two authoritative name servers.

Steps of DNS query

1. Application requests name to address translation.

2. Resolver composes query message.
3. Resolver sends UDP datagram encapsulating the query message
4. DNS server looks up address and prepares response
5. DNS sends UDP datagram encapsulating the response message.

DNS query and response messages are transmitted by using the communication
service provided by the User Datagram Protocol (UDP). Port 53 is used for
communication. The UDP service is connectionless, no connection setup is required
and the datagram is sent immediately.

SMTP

Simple Mail Transfer Protocol is used for sending mails. The user prepares an e-mail
message that includes the recipient’s e-mail address, a subject line and a body. User
clicks send button to send the information. Let us consider the steps are used in
sending the e-mail.

• The mail application establishes a TCP connection (port 25) to its local SMTP
server.
• SMTP daemon issues the message to the client, indicating that it is ready to
receive mail
• Client send s a HELO message and identifies itself.
• SMTP daemon issues a 250 message, indicating the client may proceed.
• Client sends sender’s address
• If successful, SMTP daemon replies with a 250 message
• Client sends recipient’s address.
• A 250 message is returned
• Client sends a DATA message requesting permission to send the mail message
• The daemon sends a message giving the client permission to send
• Client sends the actual text
• Daemon indicates that the message is accepted for delivery. A message ID is
returned.
• Client indicates that mail session is over.
• Daemon confirms the end of the session.

SMTP works best when the destination machine is always available, for this reason,
users in a PC environment usually retrieve their e-mail from a a mail server using the
Post Office Protocol version 3 (POP 3).

TCP and UDP transport Layer services

UDP provides connectionless transfer of datagrams between processes in hosts

attached to the Internet. UDP provides port numbering to identify the source and
destination processes in each host. UDP is simple and fast, but provides no guarantees
in terms of delivery or sequence addressing.

TCP provides for reliable transfer of a byte stream between processes in hosts
attached to the Internet. TCP involves the establishment of a connection between the
two processes. To provide their service, the TCP entities implements error detection
and retransmission as well as flow control algorithms. TCP implements congestion
control, which regulates the flow of segments into the network.

Peer-to-peer file sharing

File sharing applications such as Napster and Gnutella are popular for sharing the
MP3 audio and other files. Here clients can also acts as transient server while the
application is activated. When a peer is interested in finding a certain file, it sends a
query. The response provides a list of peer s that have the file and additional
information such as the speed of each peer’s connection to the Internet. The
requesting peer can then set up a TCP connection to one of the peers in the list and
proceeded to retrieve the file.

The OSI reference model

The model proposed by ISO OSI ( Open System Interconnection) reference model
because it deals with connecting open systems that is systems that are open for
communication with other systems. The OSI model has seven layers.
The concepts related to layers are
i) A layer should be created where a different level of abstraction is needed.
ii) Each layer should perform a well defined function
Iii) The function of each layer should be chosen with an eye toward defining
internationally standardized protocols.
 Iv) The layer boundaries should be chosen to minimise the information flow across
the interface.
v) The number of layers should be larger enough that distinct functions need not be
thrown together in the same layers out of necessity and small enough that the
architecture does not become unwieldy.

The following figure shows OSI model.

Courtesy : Data communication networks by Alberto Leon - Garcia

To remember the layers in sequence, we can use the sentence like

Please Do Not Touch Sita’s Pet Animal , Where the beginning letter of each word
represent the layer name. Like P in Please word stands for Physical layer, D letter in
Do word stands for Data link layer and so on. Let us see the functions of the layers.

Physical layer
The physical layer deals with the transfer of bits over a communication channel.
Physical layer concerned with the representation of signals. How to represents 1 and
how represent 0. Which type of encoding mechanism should be used for representing
the bits.This layer is concerned with the procedures to set up and release the physical
connections.

Data link layer

The Data link layer provides the frames (blocks of information) across a transmission
link that directly connects two nodes. Data link layer inserts a framing information in
the sequence of transmitted bits to handle boundaries. Data link layer inserts control
and address information in the header and check bits to enable recovery from
transmission errors and flow control.

High level data link control (HDLC) and Point-to-point protocol (PPP) are the two
standard data link protocol that are wide in use. A flat addressing space is used to
enable machines to listen and recognize frames that are destined to them.

Network layer
The network layer provides for the transfer of data in the form of packets across a
communication network. It uses the hierarchical way of addressing scheme. Routing
protocol is the procedure that is used to select paths across a network. The nodes in
the network must work together to perform the routing effectively. Network layer is
responsible for dealing with congestion that occurs from time to time in the network.
When the two machines are connected to the same packet-switching network, single
address space and routing procedure are used.

If two machines are connected to the different networks, the transfer of data must
traverse two or more networks and hence internetworking protocols are necessary to
route the data between gateways/routers that connects the intermediate networks.

The internetworking protocols must understand the differences in addressing and

differences in the size of the packets that are handled within each network.

Transport layer
The transport layer is responsible for the end-to-end transfer of messages from a
process in the source machine to a process in the destination machine. The transport
layer protocol accepts messages from its higher layers and prepares blocks of
information called segments or datagram's for transfer between end machines.

The transport layer provides number of services

• Transport layer may provide a connection-oriented service that involves the error-
free transfer of a sequence of bytes or messages.
• The associated protocol carries out error detection and recovery and sequence and
flow control.
• The transport layer is responsible for setting up and releasing connections across a
network.
Session layer
The session layer can be used to control the manner in which data are exchanged. It is
mainly concerns with establishing the sessions and termination the session. Certain
applications require half-duplex dialog . Certain applications requires the introduction
of synchronization points that can be used to mark the progress of an interaction and
can serve as points from which error recovery can be initiated.

Presentation layer
The presentation layer provide the application layer with independence from
difference in the representation of data. Presentation layer should convert the
machine-dependent information provided by application A into a machine-dependent
form suitable for application B, because different computers use different codes for
representing characters and integers.

Application layer
The application layer is to provide services that are frequently required by
applications that involve communications. For example browser application use
HTTP application layer protocol to access a WWW document. Application layers
protocols have been developed for file transfer, virtual terminal, electronic mail, name
service, network management and other applications.

In general each layer adds a header and possibly a trailer to the block of information it
accepts from the layer above.

Summary of the layers

The following table shows the different functions carried out by different layers OSI
model.
 Layer Description Examples
FT
Application Protocols that are designed to meet the communication requirements of HTTP, , SMTP,
specific applications, often defining the interface to a service. P
CORBA IIOP
Presentation Protocols at this level transmit data in a network representation that is Secure Sockets
independent of the representations used in individual computers, which may (SSL),CORBA Data
differ. Encryption is also performed in this layer, if required. Rep.
Session At this level reliability and adaptation are performed, such as detection of
failures and automatic recovery.
Transport This is the lowest level at which messages (rather than packets) are handled. TCP, UDP
Messages are addressed to communication ports attached to processes,
Protocols in this layer may be connection-oriented or connectionless.
Network Transfers data packets between computers in a specific network. In a WAN IP, ATM virtual
or an internetwork this involves the generation of a route passing through circuits
routers. In a single LAN no routing is required.
Data link Responsible for transmission of packets between nodes that are directly Ethernet MAC,
connected by a physical link. In a WAN transmission is between pairs of ATM cell transfer,
routers or between routers and hosts. In a LAN it is between any pair of hosts. PPP
Physical The circuits and hardware that drive the network. It transmits sequences of Ethernet base- band
binary data by analogue signalling, using amplitude or frequency modulation signalling, ISDN
of electrical signals (on cable circuits), light signals (on fibre optic circuits)
or other electromagnetic signals (on radio and microwave circuits).

We want OSI reference model because due to the following reasons.

• Summary ( why layered network)

• Reduces complexity
• Standardizes interfaces
• Facilitates modular engineering
• Ensures interoperable technology
• Accelerates evolution

Unified view of layers, protocols and services

The development of OSI reference model leads the world to the development of unified
view of layers, protocols and services. In each layer a process on one machine carries out
a conversation with a peer process on the other machine across a peer interface. This is
shown below.
Courtesy : Data communication networks by Alberto Leon - Garcia

The processes at layer n are referred to as layer n entities. Layer n entities

communicate by exchanging protocol data units (PDUs). Each PDU contains header,
which contains protocol control information and usually user information. The
communication between peer processes is usually virtual in the sense that no direct
communication link exists between them.

For communication to take place, the layer n+1 entities make use of the services
provided by layer n. The transmission of the layer n+1 PDU is done by passing a
block of information from layer n+1 to layer n through a software port called the
layer n service access point (SAP) across a service interface. SAP is identified by a
unique identifier. The block of information passed between layer n and layer n+1
entities consists of control information and a layer n service data unit (SDU), which
is the layer n+1 PDU itself. The layer n SDU, which is the layer n+1 PDU, is
encapsulated in the layer n PDU. The service provided by layer n typically accepting
a block of information from layer n+1, transferring the information to its peer
process.
 Courtesy : Data communication networks by Alberto Leon - Garcia

The service provided by the layers can be connection oriented or connectionless.

A connection-oriented service has three phases.
Step 1: Establishing a connection between two layer n SAPs. The setup involves
negotiating parameters like sequence numbers, flow control.
Step 2: Transferring n-SDUs using the layer n protocol.
Step 3: Tearing down the connection and releasing the various resources allocated to
the connection.

In connectionless service which does not require a connection setup and each SDU is
transmitted directly through the SAP. Suppose a layer n SDU is too large to be
handled by the layer n-1 and so segmentation and reassembly are applied. The layer n
SDU is segmented into multiple layer n PDUs that are then transmitted using the
services of layer n-1. The layer n entity at the other side must reassemble the original
layer n SDU from the sequence of layer n PDUs it receives. It is also possible that the
layer n SDUs are so small as to result in inefficient use of the layer n-1 services and
so blocking and unblocking applied. The layer n entity may block several layer n
SDUs into a single layer n PDU . The layer n entity on the other side must then
unblock the received PDU into the individual SDUs.

Multiplexing involves the sharing of a layer n services by multiple layer n+1 users.
The N+1 layer users passes its SDUs for transfer using the services of a single layer
n entity. Demultiplexing is carried out by the layer n entity at the other end.

Courtesy : Data communication networks by Alberto Leon - Garcia

Overview of TCP/IP architecture

The tcp/ip network architecture is a set of protocols that allows communication across
multiple diverse networks. Tcp/ip provides robustness with regard to failures in the
network and on flexibility in operating over heterogeneous networks. We know
Internet consists of different types computer systems and networks. Tcp/ip which
consists of 4 layers and OSI model consists of 7 layers. The following figure shows
the TCP/Ip architecture.

Courtesy : Data communication networks by Alberto Leon - Garcia

Application layer
The application layer provides services that are used by the applications. Examples
are remote login, e-mail, ftp, network management.The TCP/IP application layer
incorporates the functions of the top three OSI layers. HTTP is a TCP/IP application
layer protocol.
 Transport Layer

The TCP/IP application layer programs are directly run through transport layer.
Two basic types of services are provided by transport layer.
The first one is connection-oriented , which is provided by Transmission Control
Protocol (TCP).
The second service is best-effort connectionless transfer of individual messages,
which is provided by the user datagram protocols (UDP). Here in connection-less
service no mechanism for error recovery or flow control.
UDP is used for applications that require quick but not necessarily reliable delivery.
TCP/IP model does not require strict layering. It means the application layer has the
option of bypassing intermediate layers.

Internet layer

The Internet layer handles the transfer of information across multiple networks
through the use of gateways/routers . The internet layer corresponds to the network
layer of OSI , it mainly concerns with the transfer of packets between machines that
are connected to different networks. Internet layer deals with deal with the routing of
packets from router to router across the networks. Internet layer uses a concept of
globally unique addresses for machines that are attached to the Internet. The internet
layer provides a single service that is best-effort connectionless packet transfer. IP
packets are exchanged between routers without a connection setup, the packets are
routed independently. So packets may travel in different paths. Hence IP packets are
also called datagrams. The connectionless approach makes the system robust, if
failure occurs in the network, the packets are routed around the points of failure, there
is no need to set up the connections again .

Network interface layer

The network interface layer is concerned with the network-specific aspects of the transfer
of packets. It is similar to part of OSI network and data link layer services. The network
interface layer is particularly concerned with the protocols that access the intermediate
networks. At each gateway the network access protocol encapsulates the IP packet into a
packet or frame of the underlying network or link. The IP packet is recovered at the exit
gateway of the given network. This gateway must then encapsulates the IP packet into a
packet or frame of the type of the next network or link.

Hence, a clear separation of the internet layer from the technology –dependent network
interface layer. This allows the internet layer to provide a data transfer service that is
transparent in the sense of not depending on the details of the underlying the network.
The following figure shows protocols of the TCP/IP protocol suite.
Courtesy : Data communication networks by Alberto Leon - Garcia

Comparison between OSI and tcp/ip

• OSI
Three concepts are central to OSI model services, interfaces, protocols.
Protocols in the OSI model are better hidden than in the TCP/IP model and can be
replaced relatively easily as the technology changes.
Model first and then next protocol.
OSI has seven layers.
OSI supports both connectionless and connection oriented communication in the
network layer.

• TCP/IP
Services, interface and protocols are not distinguished properly.
Protocols comes first and model next.
TCP/IP has four layers.
TCP/IP model has connectionless in the Internet layer and both modes in the transport
layer.
tcp/ip - How layers work

Let us consider a network , which consists of server, workstation, router and personal
computer. The router which is used to connect two networks. The ordered pair (1,1)
which represent network 1 and the host id 1. Router which contain two interfaces one
is connected to network 1 and other is connected to pc via ppp link. It constitutes the
other network.

Courtesy : Data communication networks by Alberto Leon - Garcia

From the point of view of IP , the Ethernet LAN and the point-to-point link constitute
two different networks, shown in the following figure.
Courtesy : Data communication networks by Alberto Leon - Garcia

Each host in the Internet is identified by a globally unique addresses. IP address

identifies the host’s network interface rather than the host itself. A node that is
attached to two or more physical networks is called the router. Router attaches to two
networks with each network interface assigned to a unique IP address.

An IP address is divided into two parts

• A network id
• Host id

The network id must be obtained from an organization authorized to issue IP address.

On a LAN, the attached device to the network is often identified by a physical
address. The format of the Physical address depends on the particular type of the
network. Ethernet MAC address is 48 bits.Each Ethernet network interface card
(NIC) is issued a globally unique medium access control (MAC) or physical address.

Sending and receiving IP datagrams

Let us consider a case, in which the workstation wants to send an IP datagram to the
server. The IP datagram has the workstation’s IP address and the server’s IP
addresses in the IP packet header. Let us assume, IP address of the server is known.
The IP entity in the workstation looks at its routing table to see whether it has an
entry for the complete IP address. Workstation finds the server is directly connected
to the same network and the server has physical address s. The IP datagram is passed
to the Ethernet device driver, which prepares an Ethernet frame as shown in the next
figure.
Courtesy : Data communication net
The header in the frame contains the source physical address, w, and the destination
physical address s. The Ethernet frame is then broadcast over the LAN. The server’s
NIC recognizes that the frame is intended for its host, so the card captures the frame
and examines it. The NIC finds that the protocol type field is set to IP and therefore
passes the IP datagram up to the IP entity. Now let us see how the server sends IP
datagram to the personal computer. Suppose the server knows the IP address of the
PC and that the IP addresses on either side of the link were negotiated when the link
was set up. Otherwise, IP entity then checks to see whether it has a routing table entry
that matches the network id portion of the IP portion of the IP address of the PC.
Suppose that the IP entity does not find such an entry. The IP entity then the checks to
see whether it has an entry that specifies a default router that is to be used when no
other entries are found. Suppose that entry exists and that it specifies the router with
address (1,3). The IP datagram is passed to the Ethernet device driver, which prepares
an Ethernet frame. The header in the frame contains the source physical address, s,
and the destination physical address, r. IP datagram in the frame contains the
destination IP address of the PC (2,2), not the destination IP address of the router. The
Ethernet frame is then broadcast over the LAN. The router’s NIC captures the frame
and examines it. The routing tables at the router show that the machine with address
(2,2) is connected directly on the other side of the point-to-point link. The router
encapsulates the IP datagram in a PPP frame that is similar to the Ethernet frame. The
PPP receiver at the PC receives the frame, checks the protocol type filed and passes
the IP datagram to its IP entity.
Application layer protocols and TCP/IP utilities

Application layer protocols are high level that provide services to user applications.
Application protocols are user written. The popular protocols are Telnet, File transfer
protocol (FTP), HTTP and SMTP.

Telnet

Telnet is a TCP/IP protocol that provides a means of accessing resources on a remote

machine where the initiating machine is treated as local to the remote host. Telnet
protocol is based on the concept of a network virtual terminal (NVT), which is an
imaginary device that represents a lowest common denominator terminal. Each
machine initially maps its characteristics to that of an NVT and negotiates options for
changes to the NVT or other enhancements, such as changing the character set. Telnet
uses one TCP connection. A TCP connection is identified by a pair of port numbers, a
server is capable of supporting more than one Telnet connections at a time.

FTP

File transfer Protocol (FTP) is another commonly used application protocol. FTP
provides for the transfer of a file from one machine to another. FTP requires two TCP
connections to transfer a file. One is the control connection that is established on port
21 at the server. The second TCP connection is a data connection used to perform a
file transfer. A data connection must be established for each file transferred. Data
connections are used for transferring a file in either direction or for obtaining lists of
files in either direction or for obtaining lists of files or directories from the server to
the client. FTP works across different systems because it can accommodate several
different file types and structures. FTP commands are used to specify information
about the file and how it will be transmitted. Three types of information must be
specified. File type – FTP supports ASCII, EBCDIC, image (binary) or local.

HTTP and the WWW

The world wide web provides a framework for accessing documents and resources
that are located in computer connected to the Internet. The HTML is used to prepare
the documents. Each link provides the browser with a uniform resource locator
(URL) that specifies the name of the machine where the document is located as well
as the name of the file that contains the requested document HTTP is a stateless
protocol in that it does not maintain any information ( state) about the clients. The
HTTP server handles each request independently of all other requests.
The following are the few commands are used
GET
Retrieve information (object) identified by the URL
POST – Send information to a URL and retrieve result, used when a user fills out a
from in a browser.
PUT – Store information in location named by URL
TRACE – Trace HTTP forwarding through proxies, tunnels

Cookies and web sessions

It was indicated that the HTTP protocol is stateless and does not maintain information
about the prior requests from a given client. The use of cookies makes it possible to
have web sessions where a user interacts with a web site in a manner that takes into
account the user’s preferences. Cookies are data that are exchanged and stored by
clients and servers and transferred as headers lines in HTTP messages. The header
lines provide context for each HTTP interaction. When a client first accesses a web
server that uses cookies, the server replies with response message that includes a Set-
cookie header line. This header line includes a unique ID number for the given client.
If the client software accepts cookies, the cookies is added to the browser’s cookie
file. Each time the client software accepts cookies, the cookie is added to the
browser’s cookie file. Each time the client makes a request to the given site, it
includes a Cookie header line with the unique ID number in its requests. The server
maintains a separate cookie database where it can store which pages were accessed at
what date and time by each client.

IP utilities

A number of utilities are available to help in finding out about IP hosts and domains
and to measure Internet performance. We discuss utilities like ping, traceroute,
ipconfig, netstat, tcpdump.

Ping

Ping is a utility which is used to determine whether a host is reachable or not. Ping
makes use of Internet control message Protocol (ICMP) messages. The purpose of
ICMP is to inform sending hosts about errors encountered in Ip datagram processing
or other control information by destination hosts or by routers. Ping sends one or
more ICMP echo messages to a specified host requesting a reply. Ping is often to
measure the round-trip delay between the hosts.
The syntax is ping <hostname>

The round-trip delay is indicated and time-to-live value. The TTL is the maximum
number of hops an IP packet is allowed to remain in the network. Each time an IP
packet passes through a router, the TTL is decreased by 1. When the TTL reaches 0,
the packet is discarded.

Traceroute

Traceroute utility allows users to determine the route that a packet takes from the
local host to a remote host, as well as latency and reachability from the source to each
hop. Traceroute is used as a debugging tool by network administrator. Traceroute
makes use of both ICMP and UDP.
Ipconfig
The ipconfig utility available on windows operating system. This utility displays the
TCP/IP information about a host. It displays host IP address, subnet mask, default
gateway for the host. The utility can also be used to obtain information for each IP
network interface fro the host, DNS hostname, Ip address of DNS servers, physical
address of the network card, IP address for the network interface and whether DHCP
is enabled for automatic configuration of card’s Ip address.

netstat

The netstat queries a host about its TCP/IP network status. netstat can be used to find
the status of the network drivers and their interface cards, such as the number of
packets, out packets, error packets and so on. It is also be used to find out the state of
the routing table in a host, which TCP/IP server processes are active in the host as
well as the which TCP connection are active.

Tcpdump

The tcpdump program can capture and observe IP packet exchanges on a network
interface. The program usually involves setting an Ethernet network interface card
into a active mode so that the card listens and captures every frame that traverses the
Ethernet broadcast network. A packet filter is used to select the IP packets that are of
interest in a given situation. These IP packets and their higher-layer contents can then
be observed and analyzed. Tcpdump utility can be used as a protocol analyzer.
Protocol analyzer are extremely useful in teaching the operation of protocols by
providing a means of examining traffic from a live network. Network protocol
analyzer give the ability to capture all packets in a LAN and in doing so provide an
opportunity to gain unauthorized access to network information. These tools should
be used in a responsible manner.
 DATA COMMUNICATIONS
COMMUNICATION FUNDAMENTALS

Communication concepts

Communication is the process of transmitting information. It’s a big, vague, fuzzy word
and broadly applied. The “Communications Department “ at a university may include
speech teachers, public relations specialists, writers and broadcasters, who have nothing
to do with hardware. By that definition , a writer is in the communication business.

Evolution of communications

We the human beings are communicating from early days that is by drawing the pictures
on cave walls. It can be how to attack the animal or how to the save the life from the
threats or it may be a route to go to another cave.

The earlier long-distance communications was by signal fires that relayed simple
information. The signal used to represent “yes” or “no” – the question being asked had to
be known beforehand. It’s only a single bit information, not a detailed message.

A written letter or a human messenger could carry more information. The first systems
we might call telecommunications were series of hilltop towers, built by French engineer
Claude Chappe in the 1790s. The towers had to be in sight of each other, with an
operator in each one. The operator relayed a message by moving arms on top of the
tower. The operator of the next tower looked through a telescope to watch the arms
move, reading the message, recording it, then moving the arms of his tower to relay it to
the next tower. Chappe invented a code that used arm position to indicate letters or
certain common words. His system was called an optical telegraph.

Samuel Morse’s electrical telegraph eventually replaced Chappe’s optical one. It also
required operators, but they used keys that completed an electrical circuit, sending bursts
of currents through a wire. The Morse code was a series of dots an dashes ( short and
long pulses) that represented letters.

The electrical telegraph spread across the continents and in 1866 across the Atlantic. Its
wires formed a network running between major cities. Telegraphers received signals and
either sent them for local delivery or relayed them to more distant stations. We can think
of each dot and dash of the telegraph as a bit of information.

Later Alexander Graham Bell developed telephone. He borrowed some principles from
the telegraph , but instead of transmitting dots and dashes, it sent a continuously changing
electrical current that represented a speaker’s voice. The telephone network was bigger
and more complex than the telegraph network.
After that radio waves are used to carry signals through air. Radio first carried telegraph
signals. Wires worked fine on land, but only radio could send telegrams to ships at sea
and relay urgent messages such as request for help. It was radio rescue calls that sent
ships to help the survivors of the sinking of the Titanic. Radio communications started
for ships and airplanes in the world war II.
Radio communication started at low frequencies, but gradually moved to higher ad higher
frequencies as electronics improved. The higher the frequency the more information the
signal can carry. Pictures need more transmission capacity than sound alone, so television
are broadcast at higher frequencies than audio radio.
By the 1970s , satellites were beaming radio signals around the globe. The telephone
system had become global and known popularly “telecommunication network” . We
could make phone calls to anywhere in the world.
After that fiber optics arrived. Fiber provides high bandwidth. It carries high capacity
information at faster rate. Now it is a integral part of telecommunication network.

Broadcast television

Satellites TV Disk

Mobile Phone Standard Telephone

LAN TV Cable Television

Broadcast
Radio
P.C.

FAX Fax Phone Line

Signals and systems

Information is the one transmitted via communication. It may be very simple message
like “yes” or “no”. It can be huge and complex message, such as he files containing the
book or a television broadcast of a movie. It doesn’t matter if the “information” doesn’t
contain anything a critic might agree was “information”.

A signal transmits that information. Signals may take many forms, such as optical,
acoustic, electronic or radio-frequency. Signals may be converted from one form into
another and still contain the same information. When we make a long-distance telephone
call, the sound waves from our mouth are converted to an electrical signal at the
telephone, that electrical signal is converted to optical form at the telephone switching
office, back to electrical form at the local switching office on the other end and back to
sound waves at the other person’s telephone.

A system is the equipment that performs a task, like transmitting signals. We often speak
of telephone system as if it is the whole network of telephone equipment from our phone
to the high-capacity fiber-optics cables that carry long-distance calls. However, parts of
the telephone system also may be called systems, such as a switching systems that directs
phone calls. We can assume a system which is a collection of stuffs designed to work
together.

Basic components of data communication system

PROTOCOL

Step 1
Step 2

CHANNEL

SOURCE COMPUTER DESTINATION COMPUTER

A data communication system is made up of five components.

vi) Message
vii) Sender
viii) Receiver
ix) Medium
x) Protocol

Message - the message is the information to be communicated

Sender – The sender is the device that sends the data message. It can be computer,
workstation, telephone etc.,
Receiver – The receiver is the device that receives the message. It can be a computer ,
workstations, telephone, fax machine
Medium – The transmitting medium is the physical path by which a message travels from
sender to receiver . It can consists of twisted pair wire, coaxial cable, fiber optic cable or
radio waves.
Protocol – A protocol is a set of rules that govern data communication. It represents an
agreement between the communicating devices. Without a protocol two devices may be
connected but not communicating , just as a person speaking Malayalam cannot be
understood by a person who speaks only Kannada.

Protocols and standards

The Communication occurs between entities in different systems. An entity is anything

capable of sending or receiving information. But two entities cannot just send bit streams
to each other and expect to be understood. For communication to occur the entities must
agree on a protocol. The key elements of a protocol are syntax, semantics and timing.

Syntax : Syntax refers to the structure or format of the data, meaning the order in which
they are presented. For example , a simple protocol might expect the first eight bits of
data to be the address of the sender , the second eight bits to be the address of receiver
and rest of the stream to be the message itself.

Semantics : Semantics refers to the meaning of each section of bits. How is a particular
pattern to be interpreted and what action is to be taken based on that information.

Timing : Timing refers to two characteristics. When data should be sent and how fast it
can be sent. For example , if a sender produces data at 100 Mbps but the receiver can
process data at only 1 Mbps, the transmission will overload and receiver and data will be
largely lost.

Standards
Standards are essential in creating and maintaining an open and competitive market fro
equipment manufacturers and in guaranteeing national and international operational
ability of data and telecommunication technology and processes. They provide
guidelines to manufacturers, vendors, government agencies and other service providers to
ensure the king of interconnectivity necessary in today’s marketplace and in international
communications.

Data communications standards fall into two categories

3. de facto ( by fact)
4. de jure ( by law)

De jure standards are those that have been legislated by an officially recognized body.

De facto standards are often established originally by manufacturers seeking the

functionality of a new product or technology.

Standards organization

Standards are developed by cooperation among standard creation committees forums and
government regulatory agencies.

vi) The International Standards Organization (ISO)

vii) The International Telecommunication Union (ITU-T)
viii) ANSI (American National Standards Institute)
ix) The Institute of Electrical and Electronics Engineer (IEEE)
x) The Electronic Industries Association (EIA)

The ISO is an organization dedicated to worldwide agreement on international

standards in variety of fields. ITU-T is an international standards organization related
to the United Nations that develops standards for telecommunications. Two popular
standards developed by ITU-T are the V series and X series. ANSI , a nonprofit
organization is the United States, voting representative to both the ISO and ITU-T.
The IEEE is the largest national professional group involved in developing standards
for computing , communication, electrical engineering and electronics. It sponsored
an important standard for Local Area Network (LAN) called project 802. EIA is an
association of electronics manufacturers in the United States. It is responsible for
developing the EIA-232-D and EIA-530 standards.
System Functions

Communications systems serve many functions. It is possible to divide into four

categories on functions.
e) Point-to-point
f) Broadcasting
g) Switched
h) Networked

Point-to-point

A point-to-point configuration provides a dedicated link between two devices. The entire
capacity of the channel is reserved for transmission between those two devices. Most
point-to-point line configuration use an actual length of wire or cable to connect the two
ends.

Broadcasting

A broadcast system send the same signal to everyone who receives it. In its usual form ,
transmission is one way, from the signal source to the individual. Local radio and
television transmission are good examples of pure one-way broadcast systems.

Switched systems

A switched system makes temporary connections between terminals so they can

exchange information. The telephone is a good example.

Networking

A network connects many terminals that can communicate with each other. For example
Local Area Network is belongs to this category.

Services

We know the operation of modern communication networks is a very complex process

that involves the interaction of many systems. A communication network , is a set of
equipment and facilities that provides a service, the transfer of information between users
located at various geographical points. The ability of communication networks to transfer
communication at extremely high speeds, allows users to gather information in large
volumes almost instantly with the help of computers for long distances.
Radio and television broadcasting are probably the most common communication
services. Various stations transmit an ensemble of signals simultaneously over radio or
cable distribution networks.

Telephone service is the most common real-time service provided by a network. Two
people are able to communicate by transmitting their voices across the network. The
service is considered as connection-oriented in the sense that users must first interact with
the network to set up a connection

Telephone networks provide a broad class of call management services that use the
originating number or the destination number to determine the handling of a call. For
example , in call return the last originating number is retained to allow it to be
automatically called by the destination user at a later point in time. Caller ID allows the
originating number and sometimes name, of the originating call to be displayed to the
destination user when the receiving device is display capable. Voice mail allows a
destination user to have calls forwarded to a message receiving device when the
destination user is not available.

Cellular telephone service extends the normal telephone service to mobile users who are
free to move within a regional area covered by an interconnected array of smaller
geographical areas called cells. Each cell has a radio transmission system that allows it to
communicate with users in its area. Cellular system should handle “handoff” when user
moves from one area to another area.

Electronic mail (e-mail) is another popular network service. The user provides a text
message and a name and/or address to a mail application. The application interacts with a
local mail server, which in turn transmits the message to a destination server across a
computer network. The destination user retrieves the message by using a mail
application. E-mail is not a real-time service , where large delays can be tolerated. Here
security and privacy is matter of concern.

Many applications that involve an interaction between processes running in two

computers may be characterized by client/server interaction. Interne is an example of
distributed system. The World Wide Web (WWW) is accessed through a browser
program that displays the document and allows the user to access other documents by
clicking the links. Each link provides the browser with a uniform resource locator (URL)
that specifies the name of the machine where the document is located as well as the name
of the file that contains the requested document.

Video on demand characterization another type of interactive service. The objective of

the service is to provide access to a video library. The user initiates the service by
accessing a menu from which a selection is made. Video which requires lot of bandwidth
to pass from source computer to the destination.
Streamed audiovisual services over the internet provide an example of a service with
some of the feature of video on demand. An application like “real player “ can be used to
access a channel that provides an audio visual stream to the client machine.

Another class of service is audio-visual conferencing. Here both audio and video are
transmitted over a distance. This service requires real-time response requirement. This
service is being used by many companies for training, client interaction from a site, group
discussion etc.

Computer networks
Early computers were expensive, so techniques were developed to allow them to be
shared by many users. In the 1960s tree-topology terminal-oriented networks were
developed to allow user terminals to connect to a single central shared computer. As the
cost of computers were reduced , it became necessary to connect to different computers
for different applications. Here dumb terminals are replaced with intelligent terminals. So
techniques were developed so that networks are more flexible and provide
communication among many computers .
The ARPANET (Advanced Research projects Agency of U.S. Department of Defense)
was the first network to interconnect computers over a wide geographical area. Here, they
used intelligent computer to process and the resources were available to store
information. Later Internet protocol was developed , that enabled communications
between computers were attached to different networks. IP made it possible to transfer
information in the form of packets across many dissimilar networks.

Protocol

A protocol is a set of rules that governs how two or more communicating parties are to
interact. Let us consider an example an example of a protocol
Example

HyperText Transfer Protocol (HTTP) specifies how a web client and server are to
interact.
File Transfer Protocol for the transfer of files
Simple Mail Transfer Protocol (SMTP) for e-mail
Internet Protocol (IP) for the transfer of packets.
Domain Name System (DNS) for IP address lookup.

Terminal-oriented networks
In the figure above figure a, shows the arrangement that allows a number of
terminals to share a host computer. Terminal can be teletype printer or a video
terminal is connected by a set of wires. Later Modems allowed the devices for
transmitting digital information so that terminals could access the host computer via
the telephone network. This is shown in the figure b.
Medium access control methods were developed to allow a number of terminals at
different locations to communicate with a central computer using the shared
communication line. The access to the line, it has to be controlled so that different
devices do not interfere with each other by transmitting at the same time.
This arrangement is also called master/slave. Here central computer sends a poll
message to a specific terminal on the outgoing line. All terminals listen to the
outgoing line, but only the terminal that is polled replies by sending any information
that it has ready for transmission on the incoming line.
Statistical multiplexers/concentrators provided another method for sharing
communication line among terminals. Messages from a terminal are encapsulated
inside a frame that consists of a header in addition to the user message. The header
provides an address that identifies the terminal. The messages from the various
terminals are buffered by the multiplexer ordered into a queue and transmitted one at
a time over the communication line to the central computer. The central computer
sorts out the messages from each terminal, carries out the necessary processing and
returns the result inside a frame.

Early data transmission systems that made use of telephone lines had to deal with
errors in transmission systems that made use of telephone lines had to deal with
errors in transmission arising from a variety of sources. It may be thermal nose or
electromagnetic noise. So Error-control techniques were developed to detect the
errors and correct the errors. Hence some redundant bits called check bits are
appended to the frame. If the data is changed, due to check bit it was possible to
detect and receiver is used to request retransmission.
Finally elements of the terminal-oriented networks are
A digital transmission system to transfer binary information. The transmission system
can consists of a telephone system and specialized data transmission equipment.
A method for the transmission of frames of information between terminals and central
computers. The frames contain address information to identify terminals, check bits
to enable error control. To access the same communication line, the systems need a
medium access control to coordinate the transmissions of information from terminals
into the shared communication line.

Computer-to-computer networks

The basic service provided by computer networks is the transfer of messages from
any computer connected to the network to any other computer connected to the
network. A switching technique in which messages are partitioned into smaller units
called packets, which contain addressing information as well as sequence numbers.
Packets are then sent to the destination node one at a time, at any time and not
necessarily in a specific order. The channel used to transmit a packet is occupied only
for the duration of the packet’s transmission.
The ARPANET
The packet consisted of a header with a destination address attached to user
information. It is transmitted as a single unit across a network. ARPANET consisted
of packet switches interconnected by communication lines that provided multiple
paths for interconnecting host computers over wide geographical distances. The
packet switches were implemented by dedicated minicomputers and each packet
switch was connected to at least two other packet switches to provide alternative
paths in case of failure.

Packets were encapsulated in frames that used special characters to delineate the
frame, as well as check bits to enable error control. If errors are present, then packets
used to get discarded. The sender used to retransmit the packets once again.
Each packet switching node implemented a distributed routing algorithm to maintain
its routing tables. The algorithm involved the exchange of information between
neighbouring nodes and the calculation of a consistent set of routing tables that
together directed the flow of packets from source to destination.

This enables the update the routing tables, in case if there is a change in traffic or
topology of the network. Due to change in the routes, it was possible for packets to
arrive out of order at the destination.
Each packet switch in ARPANET contained a limited amount of buffering for
holding packets. To prevent packet switches from becoming congested, congestion
control was developed.
Flow-control methods used to regulate the rate at which a source computer sent
messages to a destination computer. This action prevented a sending computer from
overflowing the buffers of a destination computer.

APRANET provided the application like e-mail, remote login and file transfer.

In summary ARPANET have the following features

High speed 56Kbps digital transmission lines provide for the transfer of streams of
binary information. Packets between adjacent packet switches are transferred in terms
of frames. Destination are identified by unique addresses that are used by routing
tables to perform routing decisions to select the next hop of each packet as it traverse
the network.
Routing tables at the packet switches are calculated dynamically in response to
change in network traffic and topology. Messages are segmented into packets and
transmitted. End-to-end congestion control mechanism are used to prevent congestion
inside the network by limiting number of packets a host can have in transit. Flow
control methods are used. Applications that rely on the transfer of messages between
computers are developed.

The computers are connected for a small distances , low cost, very high speed and
relatively error-free communication are characteristics of LAN.
To identify computers, NETWORK INTERFACE CARD (NIC) which is used to
contain globally unique address. Broadcasting technique was used in LAN. A
medium access control protocol was used to coordinate access to the transmission
medium in order to prevent collisions between frames.

Local area networks

A variety of topologies can provide the broadcasting feature required by LANs,
including ring and tree networks. In LAN, which involves the transmission over a bus
topology coaxial cable. Stations with messages to transmit would first sense the cable
for the presence of ongoing transmissions. If no there is no carrier , the station
would proceed to transmit its message encapsulated inside a frame. The station would
continue to monitor the cable in an attempt to detect collisions.
In star topology, computers are connected by copper wires , the computers transmit
packets using the random access procedure. The collisions may occur at the hub,
where the wires converge. A LAN provides a message-transfer service between
computers and other devices that are attached to the LAN.

The elements of LAN architecture involve

A high-speed 10 megabits/second or above , digital transmission system that support
broadcast transmissions.
An addressing structure that provides station with a unique address and supports
broadcast transmissions.
A frame structure to delineate individual transmission and a medium access control
procedure to coordinate transmissions into the shared broadcast mediu.

The Internet

An internet is the interconnection of multiple networks into a single large network.

The Internet Protocol (IP) was developed to provide for the connectionless transfer
packets called datagram across an internetwork. In IP the component networks are
interconnected by special packet switches called gateways or routers. Routers adopt
to change of the network and directs the transfer of IP packets across the internet.

After a routing decision is made, the packets are placed in a buffer to await
transmission over the next network. Packets from different users are statistically
multiplexed in these buffers. The underlying networks are responsible for transferring
the packets between routers.. IP provides best-effort service. It means IP makes every
effort to deliver the packets but takes no additional actions when packets are lost,
corrupted, delivered out of order or even misdelivered. Hence IP sometimes called
unreliable.
IP uses a hierarchical address space that has “grouping “ information embedded in the
structure. IP addresses consists of 4 bytes usually expressed in dotted-decimal
notations. IP addresses consists of two parts
• A network ID
• Host ID
The internet also provides a name space to refer to machines connected to the Internet
for example nie.ac.in . Automatic translation of names to addresses is provided by the
Domain Name System (DNS)
The User Datagram Protocol (UDP) allows applications to transfer individual blocks
of user information using datagrams. UDP takes the user information, appends
appropriate information that identifies the application in the destination host and then
uses IP to transfer the datagram across an internet. UDP may not be suitable for some
application.
The transmission Control Protocol (TCP) was developed to provide reliable transfer
of stream information over the connectionless IP. TCP operates in a pair of end hosts
across an IP internet. TCP provides for error and flow control on an end-to-end basis
that can deal with the problems that can arise due to lost, delayed or misdelivered IP
packets.
Examples that work on TCP are FTP, HTTP and SMTP.
Finally, to provide communication modern network architecture consists of
Digital transmission lines of the transfer of streams of binary information.
Exchange of frames of information between adjacent equipment. Frames contain data
as well as check bits
Medium access control regulate the transmission of frames from multiple users to a
shared broadcast medium.
Addresses to identify points of attachment to a network or internetwork.
Exchange of packets of information between packet switches in network. Routing
tables in packet switches are used to select the path of each packet as it traverses the
network.
Dynamic calculation of routing tables at the packet switches in response to changes in
network traffic and topology.
Congestion control mechanism may be used to prevent congestion inside the network
Internetworking provides connectivity across multiple, possibly dissimilar networks
by using gateways or routers.
Reassembly of messages into packets are carried.
 Applications and Layered architecture
Communication network must support wide range of services. Normally people use
networks to communicate, send e-mails, transfer of files and so on. Industry people
use communication network for transfer of funds, update information about the
product and so on. Hence, to provide support for current service and future services, a
complete plan is required. This necessitates developing a complete flexibility in
network architecture.

communications functions are grouped into the following tasks

• The transport across a network of data from a process in one machine to the
process at another machine.
• The routing and forwarding of packets across multiple hops in a network
• The transfer of a frame of data from one physical interface to another.

To reduce their design complexity, most networks are organized as a series of layers
or levels, each one built upon its predecessor. The number of layers, the name of the
each layer, the contents of each layer and the function of each layer differ from
network to network.

Interaction between the layers must be defined precisely. Interaction is done with
definition of the service provided by each layer and to the layer above. Interface
between layers through which a service is requested and through which results are
conveyed. New services that build on existing services can be introduced even at the
later stage. The layered approach accommodates incremental changes readily.

We know , in all networks, the purpose of each layer is to offer certain services to the
higher layers. The entities comprising the corresponding layers on different machines
are called peer processes. Between each pair of adjacent layers there is an interface.
The interface defines which primitive operations and services the lower layer offers to
the upper one. The set of layers and protocols is called the network architecture.

A protocol is a set of rules that governs how two or more communicating devices are
to interact. HTTP protocol enables retrieval of web pages and TCP protocol enables
the reliable transfer of streams of information between computers.

HTTP

Let us consider a client/server architecture, a server process in a computer waits for

incoming requests by listening to a port. Port is an address that identifies which
process is to receive a message that is delivered to a given machine. The server
provide response to the requests. The server process always runs a process in the
background called daemon. httpd refers to server daemon for HTTP. The documents
are prepared using Hyper Text Markup Language (HTML) which consists of text,
 graphics and other media are interconnected by links that appear within the
documents. The www is accessed through a browser program that displays the
documents and allows the user to access other documents by clicking one of these
links. Each link provides the browser with a uniform resource Locator (URL) that
specifies the name of the machine where the document is located and the name of the
file that contains the requested document. The HTTP ( Hyper Text Transfer Protocol )
specifies rules by which the client and server interacts so as to retrieve a document.

In HTTP, we use two-way connection that transfer a stream of bytes in correct

sequential order and without errors. The TCP protocol provides this type of
communication service between two processes in two machines connected to a
network. Each HTTP inserts its messages into a buffer and TCP transmits the
contents of the buffer to the other TCP in blocks of information called segments.
Each segment contains port number information in addition to the HTTP message
information. The following figure shows how communication is carried between
HTTP client and HTTP server.

Courtesy : Data communication networks by Alberto Leon - Garcia

DNS query

First of all, client needs to perform a DNS query to obtain the IP address
corresponding to the domain name. This step is done by sending a message to a DNS
server. The Domain Name System (DNS) is a distributed database that resides in
multiple machines on the Internet and is used to convert between names and
addresses and to provide e-mail routing information. Each DNS machine maintains its
own database and acts as a DNS server that other query machine can query. Initially
requesting machine accesses a local name server. In case if it not possible to resolve
names then it will be sent to the root server. When the root server is unable to resolve
, it will be sent to authoritative name server. Every machine on the Internet is required
to register with at least two authoritative name servers.

Steps of DNS query

6. Application requests name to address translation.

7. Resolver composes query message.
8. Resolver sends UDP datagram encapsulating the query message
9. DNS server looks up address and prepares response
10. DNS sends UDP datagram encapsulating the response message.

DNS query and response messages are transmitted by using the communication
service provided by the User Datagram Protocol (UDP). Port 53 is used for
communication. The UDP service is connectionless, no connection setup is required
and the datagram is sent immediately.

SMTP

Simple Mail Transfer Protocol is used for sending mails. The user prepares an e-mail
message that includes the recipient’s e-mail address, a subject line and a body. User
clicks send button to send the information. Let us consider the steps are used in
sending the e-mail.

• The mail application establishes a TCP connection (port 25) to its local SMTP
server.
• SMTP daemon issues the message to the client, indicating that it is ready to
receive mail
• Client send s a HELO message and identifies itself.
• SMTP daemon issues a 250 message, indicating the client may proceed.
• Client sends sender’s address
• If successful, SMTP daemon replies with a 250 message
• Client sends recipient’s address.
• A 250 message is returned
• Client sends a DATA message requesting permission to send the mail message
• The daemon sends a message giving the client permission to send
• Client sends the actual text
• Daemon indicates that the message is accepted for delivery. A message ID is
returned.
• Client indicates that mail session is over.
• Daemon confirms the end of the session.

SMTP works best when the destination machine is always available, for this reason,
users in a PC environment usually retrieve their e-mail from a a mail server using the
Post Office Protocol version 3 (POP 3).

TCP and UDP transport Layer services

UDP provides connectionless transfer of datagrams between processes in hosts

attached to the Internet. UDP provides port numbering to identify the source and
destination processes in each host. UDP is simple and fast, but provides no guarantees
in terms of delivery or sequence addressing.

TCP provides for reliable transfer of a byte stream between processes in hosts
attached to the Internet. TCP involves the establishment of a connection between the
two processes. To provide their service, the TCP entities implements error detection
and retransmission as well as flow control algorithms. TCP implements congestion
control, which regulates the flow of segments into the network.

Peer-to-peer file sharing

File sharing applications such as Napster and Gnutella are popular for sharing the
MP3 audio and other files. Here clients can also acts as transient server while the
application is activated. When a peer is interested in finding a certain file, it sends a
query. The response provides a list of peer s that have the file and additional
information such as the speed of each peer’s connection to the Internet. The
requesting peer can then set up a TCP connection to one of the peers in the list and
proceeded to retrieve the file.

The OSI reference model

The model proposed by ISO OSI ( Open System Interconnection) reference model
because it deals with connecting open systems that is systems that are open for
communication with other systems. The OSI model has seven layers.
The concepts related to layers are
i) A layer should be created where a different level of abstraction is needed.
ii) Each layer should perform a well defined function
Iii) The function of each layer should be chosen with an eye toward defining
internationally standardized protocols.
 Iv) The layer boundaries should be chosen to minimise the information flow across
the interface.
v) The number of layers should be larger enough that distinct functions need not be
thrown together in the same layers out of necessity and small enough that the
architecture does not become unwieldy.

The following figure shows OSI model.

Courtesy : Data communication networks by Alberto Leon - Garcia

To remember the layers in sequence, we can use the sentence like

Please Do Not Touch Sita’s Pet Animal , Where the beginning letter of each word
represent the layer name. Like P in Please word stands for Physical layer, D letter in
Do word stands for Data link layer and so on. Let us see the functions of the layers.

Physical layer
The physical layer deals with the transfer of bits over a communication channel.
Physical layer concerned with the representation of signals. How to represents 1 and
how represent 0. Which type of encoding mechanism should be used for representing
the bits.This layer is concerned with the procedures to set up and release the physical
connections.

Data link layer

The Data link layer provides the frames (blocks of information) across a transmission
link that directly connects two nodes. Data link layer inserts a framing information in
the sequence of transmitted bits to handle boundaries. Data link layer inserts control
and address information in the header and check bits to enable recovery from
transmission errors and flow control.

High level data link control (HDLC) and Point-to-point protocol (PPP) are the two
standard data link protocol that are wide in use. A flat addressing space is used to
enable machines to listen and recognize frames that are destined to them.

Network layer
The network layer provides for the transfer of data in the form of packets across a
communication network. It uses the hierarchical way of addressing scheme. Routing
protocol is the procedure that is used to select paths across a network. The nodes in
the network must work together to perform the routing effectively. Network layer is
responsible for dealing with congestion that occurs from time to time in the network.
When the two machines are connected to the same packet-switching network, single
address space and routing procedure are used.

If two machines are connected to the different networks, the transfer of data must
traverse two or more networks and hence internetworking protocols are necessary to
route the data between gateways/routers that connects the intermediate networks.

The internetworking protocols must understand the differences in addressing and

differences in the size of the packets that are handled within each network.

Transport layer
The transport layer is responsible for the end-to-end transfer of messages from a
process in the source machine to a process in the destination machine. The transport
layer protocol accepts messages from its higher layers and prepares blocks of
information called segments or datagram's for transfer between end machines.

The transport layer provides number of services

• Transport layer may provide a connection-oriented service that involves the error-
free transfer of a sequence of bytes or messages.
• The associated protocol carries out error detection and recovery and sequence and
flow control.
• The transport layer is responsible for setting up and releasing connections across a
network.
Session layer
The session layer can be used to control the manner in which data are exchanged. It is
mainly concerns with establishing the sessions and termination the session. Certain
applications require half-duplex dialog . Certain applications requires the introduction
of synchronization points that can be used to mark the progress of an interaction and
can serve as points from which error recovery can be initiated.

Presentation layer
The presentation layer provide the application layer with independence from
difference in the representation of data. Presentation layer should convert the
machine-dependent information provided by application A into a machine-dependent
form suitable for application B, because different computers use different codes for
representing characters and integers.

Application layer
The application layer is to provide services that are frequently required by
applications that involve communications. For example browser application use
HTTP application layer protocol to access a WWW document. Application layers
protocols have been developed for file transfer, virtual terminal, electronic mail, name
service, network management and other applications.

In general each layer adds a header and possibly a trailer to the block of information it
accepts from the layer above.

Summary of the layers

The following table shows the different functions carried out by different layers OSI
model.
 Layer Description Examples
FT
Application Protocols that are designed to meet the communication requirements of HTTP, , SMTP,
specific applications, often defining the interface to a service. P
CORBA IIOP
Presentation Protocols at this level transmit data in a network representation that is Secure Sockets
independent of the representations used in individual computers, which may (SSL),CORBA Data
differ. Encryption is also performed in this layer, if required. Rep.
Session At this level reliability and adaptation are performed, such as detection of
failures and automatic recovery.
Transport This is the lowest level at which messages (rather than packets) are handled. TCP, UDP
Messages are addressed to communication ports attached to processes,
Protocols in this layer may be connection-oriented or connectionless.
Network Transfers data packets between computers in a specific network. In a WAN IP, ATM virtual
or an internetwork this involves the generation of a route passing through circuits
routers. In a single LAN no routing is required.
Data link Responsible for transmission of packets between nodes that are directly Ethernet MAC,
connected by a physical link. In a WAN transmission is between pairs of ATM cell transfer,
routers or between routers and hosts. In a LAN it is between any pair of hosts. PPP
Physical The circuits and hardware that drive the network. It transmits sequences of Ethernet base- band
binary data by analogue signalling, using amplitude or frequency modulation signalling, ISDN
of electrical signals (on cable circuits), light signals (on fibre optic circuits)
or other electromagnetic signals (on radio and microwave circuits).

We want OSI reference model because due to the following reasons.

• Summary ( why layered network)

• Reduces complexity
• Standardizes interfaces
• Facilitates modular engineering
• Ensures interoperable technology
• Accelerates evolution

Unified view of layers, protocols and services

The development of OSI reference model leads the world to the development of unified
view of layers, protocols and services. In each layer a process on one machine carries out
a conversation with a peer process on the other machine across a peer interface. This is
shown below.
Courtesy : Data communication networks by Alberto Leon - Garcia

The processes at layer n are referred to as layer n entities. Layer n entities

communicate by exchanging protocol data units (PDUs). Each PDU contains header,
which contains protocol control information and usually user information. The
communication between peer processes is usually virtual in the sense that no direct
communication link exists between them.

For communication to take place, the layer n+1 entities make use of the services
provided by layer n. The transmission of the layer n+1 PDU is done by passing a
block of information from layer n+1 to layer n through a software port called the
layer n service access point (SAP) across a service interface. SAP is identified by a
unique identifier. The block of information passed between layer n and layer n+1
entities consists of control information and a layer n service data unit (SDU), which
is the layer n+1 PDU itself. The layer n SDU, which is the layer n+1 PDU, is
encapsulated in the layer n PDU. The service provided by layer n typically accepting
a block of information from layer n+1, transferring the information to its peer
process.
 Courtesy : Data communication networks by Alberto Leon - Garcia

The service provided by the layers can be connection oriented or connectionless.

A connection-oriented service has three phases.
Step 1: Establishing a connection between two layer n SAPs. The setup involves
negotiating parameters like sequence numbers, flow control.
Step 2: Transferring n-SDUs using the layer n protocol.
Step 3: Tearing down the connection and releasing the various resources allocated to
the connection.

In connectionless service which does not require a connection setup and each SDU is
transmitted directly through the SAP. Suppose a layer n SDU is too large to be
handled by the layer n-1 and so segmentation and reassembly are applied. The layer n
SDU is segmented into multiple layer n PDUs that are then transmitted using the
services of layer n-1. The layer n entity at the other side must reassemble the original
layer n SDU from the sequence of layer n PDUs it receives. It is also possible that the
layer n SDUs are so small as to result in inefficient use of the layer n-1 services and
so blocking and unblocking applied. The layer n entity may block several layer n
SDUs into a single layer n PDU . The layer n entity on the other side must then
unblock the received PDU into the individual SDUs.

Multiplexing involves the sharing of a layer n services by multiple layer n+1 users.
The N+1 layer users passes its SDUs for transfer using the services of a single layer
n entity. Demultiplexing is carried out by the layer n entity at the other end.

Courtesy : Data communication networks by Alberto Leon - Garcia

Overview of TCP/IP architecture

The tcp/ip network architecture is a set of protocols that allows communication across
multiple diverse networks. Tcp/ip provides robustness with regard to failures in the
network and on flexibility in operating over heterogeneous networks. We know
Internet consists of different types computer systems and networks. Tcp/ip which
consists of 4 layers and OSI model consists of 7 layers. The following figure shows
the TCP/Ip architecture.

Courtesy : Data communication networks by Alberto Leon - Garcia

Application layer
The application layer provides services that are used by the applications. Examples
are remote login, e-mail, ftp, network management. The TCP/IP application layer
incorporates the functions of the top three OSI layers. HTTP is a TCP/IP application
layer protocol.
 Transport Layer

The TCP/IP application layer programs are directly run through transport layer.
Two basic types of services are provided by transport layer.
The first one is connection-oriented , which is provided by Transmission Control
Protocol (TCP).
The second service is best-effort connectionless transfer of individual messages,
which is provided by the user datagram protocols (UDP). Here in connection-less
service no mechanism for error recovery or flow control.
UDP is used for applications that require quick but not necessarily reliable delivery.
TCP/IP model does not require strict layering. It means the application layer has the
option of bypassing intermediate layers.

Internet layer

The Internet layer handles the transfer of information across multiple networks
through the use of gateways/routers . The internet layer corresponds to the network
layer of OSI , it mainly concerns with the transfer of packets between machines that
are connected to different networks. Internet layer deals with deal with the routing of
packets from router to router across the networks. Internet layer uses a concept of
globally unique addresses for machines that are attached to the Internet. The internet
layer provides a single service that is best-effort connectionless packet transfer. IP
packets are exchanged between routers without a connection setup, the packets are
routed independently. So packets may travel in different paths. Hence IP packets are
also called datagrams. The connectionless approach makes the system robust, if
failure occurs in the network, the packets are routed around the points of failure, there
is no need to set up the connections again .

Network interface layer

The network interface layer is concerned with the network-specific aspects of the transfer
of packets. It is similar to part of OSI network and data link layer services. The network
interface layer is particularly concerned with the protocols that access the intermediate
networks. At each gateway the network access protocol encapsulates the IP packet into a
packet or frame of the underlying network or link. The IP packet is recovered at the exit
gateway of the given network. This gateway must then encapsulates the IP packet into a
packet or frame of the type of the next network or link.

Hence, a clear separation of the internet layer from the technology –dependent network
interface layer. This allows the internet layer to provide a data transfer service that is
transparent in the sense of not depending on the details of the underlying the network.
The following figure shows protocols of the TCP/IP protocol suite.
Courtesy : Data communication networks by Alberto Leon - Garcia

Comparison between OSI and tcp/ip

• OSI
Three concepts are central to OSI model services, interfaces, protocols.
Protocols in the OSI model are better hidden than in the TCP/IP model and can be
replaced relatively easily as the technology changes.
Model first and then next protocol.
OSI has seven layers.
OSI supports both connectionless and connection oriented communication in the
network layer.

• TCP/IP
Services, interface and protocols are not distinguished properly.
Protocols comes first and model next.
TCP/IP has four layers.
TCP/IP model has connectionless in the Internet layer and both modes in the transport
layer.
tcp/ip - How layers work

Let us consider a network , which consists of server, workstation, router and personal
computer. The router which is used to connect two networks. The ordered pair (1,1)
which represent network 1 and the host id 1. Router which contain two interfaces one
is connected to network 1 and other is connected to pc via ppp link. It constitutes the
other network.

Courtesy : Data communication networks by Alberto Leon - Garcia

From the point of view of IP , the Ethernet LAN and the point-to-point link constitute
two different networks, shown in the following figure.
Courtesy : Data communication networks by Alberto Leon - Garcia

Each host in the Internet is identified by a globally unique addresses. IP address

identifies the host’s network interface rather than the host itself. A node that is
attached to two or more physical networks is called the router. Router attaches to two
networks with each network interface assigned to a unique IP address.

An IP address is divided into two parts

• A network id
• Host id

The network id must be obtained from an organization authorized to issue IP address.

On a LAN, the attached device to the network is often identified by a physical
address. The format of the Physical address depends on the particular type of the
network. Ethernet MAC address is 48 bits.Each Ethernet network interface card
(NIC) is issued a globally unique medium access control (MAC) or physical address.

Sending and receiving IP datagrams

Let us consider a case, in which the workstation wants to send an IP datagram to the
server. The IP datagram has the workstation’s IP address and the server’s IP
addresses in the IP packet header. Let us assume, IP address of the server is known.
The IP entity in the workstation looks at its routing table to see whether it has an
entry for the complete IP address. Workstation finds the server is directly connected
to the same network and the server has physical address s. The IP datagram is passed
to the Ethernet device driver, which prepares an Ethernet frame as shown in the next
figure.
Courtesy : Data communication net
The header in the frame contains the source physical address, w, and the destination
physical address s. The Ethernet frame is then broadcast over the LAN. The server’s
NIC recognizes that the frame is intended for its host, so the card captures the frame
and examines it. The NIC finds that the protocol type field is set to IP and therefore
passes the IP datagram up to the IP entity. Now let us see how the server sends IP
datagram to the personal computer. Suppose the server knows the IP address of the
PC and that the IP addresses on either side of the link were negotiated when the link
was set up. Otherwise, IP entity then checks to see whether it has a routing table entry
that matches the network id portion of the IP portion of the IP address of the PC.
Suppose that the IP entity does not find such an entry. The IP entity then the checks to
see whether it has an entry that specifies a default router that is to be used when no
other entries are found. Suppose that entry exists and that it specifies the router with
address (1,3). The IP datagram is passed to the Ethernet device driver, which prepares
an Ethernet frame. The header in the frame contains the source physical address, s,
and the destination physical address, r. IP datagram in the frame contains the
destination IP address of the PC (2,2), not the destination IP address of the router. The
Ethernet frame is then broadcast over the LAN. The router’s NIC captures the frame
and examines it. The routing tables at the router show that the machine with address
(2,2) is connected directly on the other side of the point-to-point link. The router
encapsulates the IP datagram in a PPP frame that is similar to the Ethernet frame. The
PPP receiver at the PC receives the frame, checks the protocol type filed and passes
the IP datagram to its IP entity.
Application layer protocols and TCP/IP utilities

Application layer protocols are high level that provide services to user applications.
Application protocols are user written. The popular protocols are Telnet, File transfer
protocol (FTP), HTTP and SMTP.

Telnet

Telnet is a TCP/IP protocol that provides a means of accessing resources on a remote

machine where the initiating machine is treated as local to the remote host. Telnet
protocol is based on the concept of a network virtual terminal (NVT), which is an
imaginary device that represents a lowest common denominator terminal. Each
machine initially maps its characteristics to that of an NVT and negotiates options for
changes to the NVT or other enhancements, such as changing the character set. Telnet
uses one TCP connection. A TCP connection is identified by a pair of port numbers, a
server is capable of supporting more than one Telnet connections at a time.

FTP

File transfer Protocol (FTP) is another commonly used application protocol. FTP
provides for the transfer of a file from one machine to another. FTP requires two TCP
connections to transfer a file. One is the control connection that is established on port
21 at the server. The second TCP connection is a data connection used to perform a
file transfer. A data connection must be established for each file transferred. Data
connections are used for transferring a file in either direction or for obtaining lists of
files in either direction or for obtaining lists of files or directories from the server to
the client. FTP works across different systems because it can accommodate several
different file types and structures. FTP commands are used to specify information
about the file and how it will be transmitted. Three types of information must be
specified. File type – FTP supports ASCII, EBCDIC, image (binary) or local.

HTTP and the WWW

The world wide web provides a framework for accessing documents and resources
that are located in computer connected to the Internet. The HTML is used to prepare
the documents. Each link provides the browser with a uniform resource locator
(URL) that specifies the name of the machine where the document is located as well
as the name of the file that contains the requested document HTTP is a stateless
protocol in that it does not maintain any information ( state) about the clients. The
HTTP server handles each request independently of all other requests.
The following are the few commands are used
GET
Retrieve information (object) identified by the URL
POST – Send information to a URL and retrieve result, used when a user fills out a
from in a browser.
PUT – Store information in location named by URL
TRACE – Trace HTTP forwarding through proxies, tunnels

Cookies and web sessions

It was indicated that the HTTP protocol is stateless and does not maintain information
about the prior requests from a given client. The use of cookies makes it possible to
have web sessions where a user interacts with a web site in a manner that takes into
account the user’s preferences. Cookies are data that are exchanged and stored by
clients and servers and transferred as headers lines in HTTP messages. The header
lines provide context for each HTTP interaction. When a client first accesses a web
server that uses cookies, the server replies with response message that includes a Set-
cookie header line. This header line includes a unique ID number for the given client.
If the client software accepts cookies, the cookies is added to the browser’s cookie
file. Each time the client software accepts cookies, the cookie is added to the
browser’s cookie file. Each time the client makes a request to the given site, it
includes a Cookie header line with the unique ID number in its requests. The server
maintains a separate cookie database where it can store which pages were accessed at
what date and time by each client.

IP utilities

A number of utilities are available to help in finding out about IP hosts and domains
and to measure Internet performance. We discuss utilities like ping, traceroute,
ipconfig, netstat, tcpdump.

Ping

Ping is a utility which is used to determine whether a host is reachable or not. Ping
makes use of Internet control message Protocol (ICMP) messages. The purpose of
ICMP is to inform sending hosts about errors encountered in Ip datagram processing
or other control information by destination hosts or by routers. Ping sends one or
more ICMP echo messages to a specified host requesting a reply. Ping is often to
measure the round-trip delay between the hosts.
The syntax is ping <hostname>

The round-trip delay is indicated and time-to-live value. The TTL is the maximum
number of hops an IP packet is allowed to remain in the network. Each time an IP
packet passes through a router, the TTL is decreased by 1. When the TTL reaches 0,
the packet is discarded.

Traceroute

Traceroute utility allows users to determine the route that a packet takes from the
local host to a remote host, as well as latency and reachability from the source to each
hop. Traceroute is used as a debugging tool by network administrator. Traceroute
makes use of both ICMP and UDP.
Ipconfig
The ipconfig utility available on windows operating system. This utility displays the
TCP/IP information about a host. It displays host IP address, subnet mask, default
gateway for the host. The utility can also be used to obtain information for each IP
network interface fro the host, DNS hostname, Ip address of DNS servers, physical
address of the network card, IP address for the network interface and whether DHCP
is enabled for automatic configuration of card’s Ip address.

netstat

The netstat queries a host about its TCP/IP network status. netstat can be used to find
the status of the network drivers and their interface cards, such as the number of
packets, out packets, error packets and so on. It is also be used to find out the state of
the routing table in a host, which TCP/IP server processes are active in the host as
well as the which TCP connection are active.

Tcpdump

The tcpdump program can capture and observe IP packet exchanges on a network
interface. The program usually involves setting an Ethernet network interface card
into a active mode so that the card listens and captures every frame that traverses the
Ethernet broadcast network. A packet filter is used to select the IP packets that are of
interest in a given situation. These IP packets and their higher-layer contents can then
be observed and analyzed. Tcpdump utility can be used as a protocol analyzer.
Protocol analyzer are extremely useful in teaching the operation of protocols by
providing a means of examining traffic from a live network. Network protocol
analyzer give the ability to capture all packets in a LAN and in doing so provide an
opportunity to gain unauthorized access to network information. These tools should
be used in a responsible manner.
Circuit switching networks

Circuit switching networks provide dedicated circuits that provide the flow of
information between users. Multiplexing in the physical layer involves the sharing of
transmission systems by several connections or information flows. Multiplexing is
required when the bandwidth of individual connections is smaller than the bandwidth
of the available transmission system. Multiplexing helps to use bandwidth in a
efficient way and also reduces cost incurred in using the separate connections.

Frequency-division multiplexing

In Frequency division multiplexing technique we partition the available transmission

frequency range into narrower bands ( subfrequencies), each of which is a separate
channel. FDM based transmissions are parallel in nature. Examples of FDM are
broadcast radio and broadcast and cable television, where each station has an
assigned frequency band.

Time division multiplexing

In Time division multiplexing technique that assigns to each node connected to a

channel an identification number and a small amount of time in which to transmit.
TDM based transmissions are serially sequenced. TDM was started in the telephone
network in the early 1960s. The T-1 carrier system that carries 24 digital telephone
connections. The T-1 system uses a transmission frame that consists of 24 slots of
eight bits each. Each slot carries one PCM sample for a single connection. The
beginning of each frame is indicated by a single framing bit.

The T-1 carrier system was used to carry the traffic between telephone central offices.
Later growth of telephone network traffic and the advances in digital transmission led
to the development of a standard digital multiplexing hierarchy.
In North America and Japan, the digital signal 1 (DS1) , which corresponds to the
output of T-1 multiplexers, became the basic building block. The DS2 signal is
obtained by combining 4 DS1 signals and then adding 136 kilobits of synchronization
information. The DS3 signals is obtained by combining 7 DS2 signals and adding 522
kilobits of synchronization.

Summarized table
DS1 = 1.544 Mbps
DS2 = 4 * DS1 + 136 kilobits = 6.312 Mbps
DS3 = 7 * DS2 + 552 kilobits = 44.736 Mbps

Wavelength division multiplexing

A multiplexing method used with fiber-optic cables. Involves the simultaneous
transmission of light sources over a single fiber-optic channel. Light sources of
different wavelengths are combined by a WDM multiplexer and transmitted over a
single line. When the signals arrive, a WDM demultiplexer separates them and
transmits them to their respective destination receivers. Normally speed of
information is in terms of tens of Gbps. Fiber provides a high bandwidth.

SONET

The advances in optical transmission technology led to the development of SONET

( Synchronous Optical Network). An ANSI physical layer standard that provides an
international specification for high-speed digital transmission via optical fiber.
Initially developed in North America. Later CCITT (Consultative Committee for
International Telegraphy and Telephony) developed a corresponding set of standards
called Synchronous Digital Hierarchy ( SDH). Current backbone networks in US are
SONET and in Europe are SDH.

The SONET simplifies the handling of lower-level digital signals and reduces the
overall cost of multiplexing in the network. In SONET , incorporation of overhead
bytes in the frame structure for use in monitoring the signal quality, detecting faults
and signaling among SONET equipment to identify rapid recovery from faults.

The SONET standard uses a 51.84 Mbps electrical signal, known as Synchronous
transport signal level – 1 (STS-1), as a building block to extend the digital
transmission. Hierarchy into the multigigabit/ second stage.

A SONET system is divided into three layers

Section , lines and Paths.
A Section refers to the span of fiber between two adjacent devices, such as two
regenerators. The section layer deals with the transmission of an STS-n signal across
the physical medium. A line refers to the span between two adjacent multiplexers and
therefore in general encompasses several sections. Lines deals with the transport of an
aggregate multiplexed stream and the associated overhead.

A path refers to the span between the two SONET terminals at the endpoints of the
system and in general encompasses one or more lines. Adjacent multiplexers
exchange information using frames.

A frame consisting of a rectangular array of bytes arranged in 9 rows by 90 columns

is repeated 8000 times a second. Thus each byte in the array corresponds to a bit rate
of 64 kbps and the overall bit rate of the STS-1 is
• 8 * 9 * 90 * 8000 = 51.84 Mbps
The first three columns of the array are allocated to section and line overhead. The
overhead provide framing, error monitoring and other section related management
functions. For example A1 and A2 indicates the beginning of the frame. The fourth
byte B1 carries parity checks of the transmitted signal and is used to monitor the bit
error rate in a section.

The last three bytes of the section overhead are used to provide a data
communications channel between regenerators that can be used to exchange alarms,
control, monitoring and other administrative messages. The line overhead is used for
synchronization and multiplexing, performance monitoring, line maintenance, as well
as protection-switching capability in case of faults. The byte B2 is used to monitor the
bit rate in a line. The K1 and K2 bytes are used to trigger recovery procedure in case
of faults.

The remaining 87 columns of the frame constitute the information payload that
carries the path layer of user information.
The bit rate of the information payload is
8 * 9 * 87 * 8000 = 50.122 Mbps.

Let us see how information is organized at the path level.

The SONET terminal equipment takes the user data and the path overhead and maps
into a synchronous payload envelope (SPE), which consists of nine rows and 87
columns.

The path overhead uses the first columns of this array. This SPE is then inserted into
the STS-1 frame. The first two bytes of the line overhead are used as a pointer that
indicates the byte within the information payload where the SPE begins. The use of
pointer makes it possible to extract a selected signal from the multiplexed signal. This
feature gives SONET multiplexers an add-drop capability, which represent the
selected signal can be dropped and individual signal can be added without to
demultiplex the entire signal.

The pointer structure consisting of the H1, H2 and H3 bytes maintains

synchronization of frames and SPEs is situation where their clock frequencies differ
slightly. If the payload stream falls behind the payload stream. To allow the frame to
catch up, an extra SPE byte is transmitted in a frame from time to time. This extra
byte, which is represented by H3 clears the backlog. When the payload stream is
slower than the frame stream, the number of SPE bytes transmitted in a frame needs
to be reduced by one byte from time to time.

TRANSPORT NETWORKS

A transport network provides high bit rate connections to clients at different

locations. The clients of a transport network are routers, large telephone switches or
even other networks.

Routers are interconnected for form an ISP backbone network and the telephone
switches are interconnected to form a telephone network. The failure of a single
connection in a transport network can be disastrous because each connection carries
heavy traffic.

SONET Networks

In the beginning, asynchronous multiplexing concept was used. In that bit stuffing
was used to deal with slips and hence the complete multiplexed stream to be
demultiplexed to access a single tributary.
Transit tributaries would then have to be remultiplexed onto the next hop. Back-to-
back multiplexers are an unnecessary expense in situations where most of the traffic
is transit and only a few tributaries need to be dropped. SONET which enables in
reduction in cost by enabling add-drop multiplexers (ADM) that can insert and extract
tributary streams without disturbing tributary streams that are in transit.

ADMs are arranged in linear fashion to interconnect various SONET terminals. These
terminals can be part of other equipment, they could be interfaces to a large router.
From the point of view of the terminals, they are arranged in the fully connected
mesh. The SONET standards also defines automatic protection switching (APS)
schemes that provide linear protection against failures at the line layer.

Here, protection at the line level applies to a multiplexed signal while it traverses the
line between two multiplexers. Two SONET multiplexers connected in a 1+1 (one
plus one ) configuration using two SONET lines, a working line and protection line.
Two lines carries the same information. A selector in the downstream node picks the
better signal based on the information provided by the two monitors and does not
need to coordinate with the upstream node. Recovery from failures are fast because it
is done by the hardware.

A 1:1 ( one for one) APS arrangement is shown (refer ppt), Here, signal is
transmitted in the working line during normal operation. The optical signal that is
received is monitored and in case if the signal is degraded, then a request is sent.
Once again, upstream node switches the signal to the protection line.
The 1:1 takes more time to recover from failure than 1 + 1 APS.
The 1:1 scheme is efficient in terms of the bandwidth.

In the other arrangement 1:1 APS scheme can be generalized to a 1:n APS scheme
where one protection line protects n working lines. The scheme assumes that the
working lines are unlikely to fail at the same time.

The SONET linear APS specification says that fault recovery be completed within 50
milliseconds in all of the schemes discussed previously.

Ring Networks
In the figure above, a, b and c represents the three sites and are connected by three
add-drop multiplexers. At node b, two STS-n tributaries are inserted destined for node
c and for node a, respectively. The first tributary ends at node c and second tributary
flows across node c and terminates at node a.

SONET cross-connect and mesh networks

SONET ring networks provide good service even in case of faults. SONET networks
suffers from the drawback where scalability issue is concerned. As networks carries
more traffic , certain spans in a ring will become congested. To increase the capacity
of a single span in a single ring network, all the ADMs need to be upgraded at the
same time. This upgrade involves very high expense and effort.
The drawback can be eliminated by building multiple parallel ring networks.
This approach is compatible with WDM transmission system, which provides parallel
rings using a single fiber channel. At each node, the signal carried in each wavelength
must be connected to a separate ADM. To exchange traffic between ADMs at a
node, the traffic must first be dropped from a source ADM, connected to a switch and
then added into the destination ADM.
In SONNET cross-connect system can take SONNET OC-n optical signals at its
inputs, decompose these signals into its components STS-1 or other tributaries, switch
these component signals to output ports and combine the component signals into
outgoing SONET OC-n signals. Mesh networks are constructed by interconnecting
cross-connect systems with SONNET optical lines .

Mesh networks
When certain segments become congested, mesh networks have the advantage that
only the capacity of the affected segments have to be upgraded. Mesh networks
require less protection bandwidth than ring networks.

Optical transport networks

Optical transport networks provide optical wavelength connections between attached

clients. Optical add-drop multiplexers (OADM) have been used for WDM systems.
An OADM takes a multiwavelength signal arriving in an input fiber, drops one or
more pre-selected wavelength at a site and adds one or more pre-selected wavelengths
into multiwavelngth signal that exists in an output fiber.
OADM can be arranged in linear and ring topologies.
chain of optical add-drop multiplexers in which a single fiber connects adjacent
multiplexers. Each fiber contains a set of four wavelengths that are removed and
added to provide a one-directional communication link from source to destination.

The advantage of all optical networks is that they can carry signals transparently. In
WDM each wavelength is modulated separately, so each wavelength need not carry
information in the same format. Optical mesh networks based on optical cross-
connect and optical fiber switching systems are used as backbone of transport
networks. The use of fiber switch is to transfer entire multiwavelength signals from
input ports to output ports without demultiplexing.
Figure shown above, optical signals that carry WDM signals arrive in individual
fibers to a node. Optical signals that carry signals destined for the node are switched
to the local wavelength cross-connect. The wavelength signals are destined for the
node are then dropped. Locally generated wavelengths signals are added through the
cross-connect. Next the outputs from the cross-connect are fed to WDM multiplexers
that produce composite optical signals that are then inserted into the fiber switch and
transferred out in an outgoing fiber.

Optical transport networks

The previous mentioned arrangement makes the overall cost incurred is low with
respect to increasing the traffic.

Circuit switches
The function of a circuit switch is to transfer the signal that arrives at a given input to
an appropriate output. The interconnection of a sequence of transmission links and
circuit switches enables the flow of information between inputs and outputs in the
network.
Space-division switches

Space-division switches provide a separate physical connection between inputs and

outputs so the different signals are separated in space. Crossbar switch is an example
for this category.

Crossbar switch
The crossbar switch consists of an N x N array of crosspoints that can connect any
input to any available output. When a request comes in from an incoming line for an
outgoing line, the corresponding crosspoint is closed to enable information to flow
from the input to the output. The crossbar switch is nonblocking switch means
requests are never denied because of lack of connectivity resources that is
crosspoints. Connections requests are denied when only outgoing line is already in
use.

The complexity is cross bar switch is N * N. Huge number of crosspoints cannot be

implemented.
In the figure shows a multistage switch that consists of three stages of smaller space
division switches. The N inputs are grouped into N/n groups of n input lines. Each
group of n input lines enters a small switch in the first stage that consists of n x k
array of crosspoints. Each input switch has one line connecting it to each of k
intermediate stage N/n x N/n switches. Each intermediate switch has one line
connecting to each N/n switches in the third stage. The next switches are k x n. Each
set of n input lines shares k possible paths to any one of the switches at the last stage.

CLOS Nonblocking switching fabric

Here , we determine when a multistage switch becomes nonblocking. The worst case
for the desired input is when all the other inputs in its group have already connected.
Similarly it is same for the output , where all the other outputs in its group have
already been connected. The set of routes which uses the intermediate switches
already in use shown in the next figure.
Each existing connection uses a different intermediate switch. The maximum number
of intermediate switches not available to connect the desired input to the desired
output is 2 (n-1). Suppose k = 2n -1 , then k paths are available from any input group
to any output group, because 2 (n -1) of these paths are already in use, it follows a
single path remains available to connect the desired input to the desired output.
Hence, the multistage switch k = 2n -1 is nonblocking.

The number of crosspoints required in a three stage switch is the sum of the following
components
• N/n input switches x nk crosspoints/input switch
• K intermediate switches x (N/n)^2 crosspoints/intermediate switch
• N/n output switches x nk crosspoints/output switch.
Total number of crosspoints is 2 Nk + k (N/n)^2. The number of crosspoints required
to make the switch nonblocking is
2 N(2n -1) + (2n-1)(N/n)^2.

Time division switches

We know in TDM a slot within a frame refers to a single connection. The time-slot
interchange (TSI) technique replaces the crosspoints in a space switch with the
reading and writing of a slot into a memory. Assume we have many speakers want to
have conversation. The speech of the speaker are sampled at 8000 bytes/second and
also bytes are placed on T-1 carrier shown in the next figure.
In the figure above , first pair of speakers are assigned slots 1 and 23. For the
speakers to hear each other, we need to route slots 1 and 23 in the incoming frames to
slots 23 and 1 in the outgoing frames. The octets in each incoming frame are stored
into a register. The call setup procedure has a set a permutation table that controls the
order in which the contents of the register are read out. Hence the outgoing the frame
begins by reading the contents of slot 23, followed by slot 24 and so on until slots 1
and 2 are read.

A hybrid switch, TSI switches are used at the input and output stages and a crossbar
space switch is used at the intermediate stage. These switches are called time-space-
time switches. We replace n x k switch in the first stage of a multistage space switch
by an n x k TSI switch , shown in the next figure.
Each input line to the switch corresponds to a slot, so the TSI switch has input frames
of size n slots. The output frame from the TSI switch has k slots. In essence, n slots
from the incoming frame and reading them out in a frame of size k, according to
some preset permutation table. The first slot corresponds to the first output line out of
each of the first stage switches. First line out of the first stage switch is connected to
the first intermediate switch. Hence, first slot in each intermediate frame will be
directed to intermediate switch 1 shown in the next figure.

The second slot in a frame is directed to crossbar switch 2 and all other intermediate
switches are idle. It thus becomes apparent that only one of the crossbar switches is
active during any given timeslot. This makes it possible, to replace the k intermediate
crossbar switches with a single crossbar switch that is time-shared among the k slots
in a frame shown in the next figure.
 To replace the k intermediate original crossbar switches, the time-
shared crossbar switch must be reconfigured to the interconnection pattern of the
corresponding original switch at every time slot. This approach to sharing a space
switch is called time-division multiplexing.