UNIT 9 COMMUNICATION PROTOCOLS

AND NETWORK ADDRESSING

Structure
9.0 Objectives
9.1 Introduction
9.2 What are Protocols?
9.3 Computing Protocols
9.4 Communication Protocols: General Concepts
9.5 Common Communication Protocols
9.6 Basic Communication Protocols: IP, UDP, TCP
9.6.1 Internet Protocol (IP)
9.6.2 User Datagram Protocol (UDP)
9.6.3 Transmission Control Protocol (TCP)

9.7 Client-Server Architecture

9.8 Application Level Communication Protocols: FTP, Telnet
9.8.1 File Transfer Protocol (FTP)
9.8.2 Remote Login (Telnet)

9.9 Switching Level Convergence Protocol: ATM

9.10 Multi Protocol Label Switching: MPLS
9.11 Telephone and Mobile Numbering
9.11.1 Landline Telephone Numbering
9.11.2 Mobile Phone Numbering

9.12 Number Portability

9.13 IP Addressing: IPv4, IPv6
9.14 Web Communication Protocols: HTTP, WAP, LTP
9.15 Summary
9.16 Answers to Self-check Exercises
9.17 Keywords
9.18 References and Further Reading

9.0 OBJECTIVES
After going through this Unit, you will be able to understand and appreciate:
 What are protocols;
 Difference between computing and communication protocols;
 Need for communication protocols;
31
Network Fundamentals  Difference between connection-oriented and connectionless protocols;
 Basic packet transfer protocols like IP, TCP and UDP;
 Most widely used network computing architecture: Client-Server;
 File transfer and remote login application protocols like FTP and Telnet;
 Convergent cell switching in ATM networks;
 Fast routing technique using labels: MPLS;
 Numbering schemes used for landline and mobile phones;
 How network computers are addressed world over;
 Details of IPv4 addressing scheme;
 IPv6 features briefly; and
 Web access protocols for both wired and wireless networks.

9.1 INTRODUCTION
Quest for new knowledge is the central theme of human existence. All of us, whether
we realise or not, are in the process of acquiring new knowledge all the time. When we
ask a question, we are seeking knowledge. When we answer a query, we give information
to the person posing the question. When a person assimilates the given information, we
say that the person has acquired knowledge. Knowledge is spread via information that
is communicated from one person to another in some form: oral, writing etc. Thus,
knowledge, information and information communication are three entities that are closely
inter-related.
It is often said that we are in the information age. In the last about six decades, information
in the world has been growing at an exponential rate, i.e. doubling every 10 years.
Information Communication Technology (ICT) has grown leaps and bound in the last
30 - 40 years. Instant transfer of information from one part of the world to any other
part is a reality today. Underlying this development is the convergence of computer and
communication technologies. This convergence process started in late 1960s and has
led to the development of worldwide computer network that is now known popularly
as Internet. A large number of home and office local area networks (LANs) and
innumerable personal computers all over the world have been interconnected to form
Internet. Hence, it aptly said that Internet is a network of networks. Information travels
in the form of data packets on Internet and hence it is also called a data network. Data
packets are of fixed length, say 2048 bytes, i.e. 2" bytes. Long messages are broken
into as many packets as required before transmission. Because of packet-based
transmission, the Internet also carries the nomenclature Packet Data Network (PDN).
Since Internet is an open public network, another related nomenclature that is used
sometimes is Packet Switched Public Data network (PSPDN). Internet is not limited
to its presence only on the land but is also in ships at the seas and in planes in the air.
United Nations today has 192 countries of the world as its members. Almost all these
countries have Internet connection in place. About 200,000 LANs are connected to
the Internet. Over 1.5 billion people, i.e. a quarter of the world population has access
to Internet. With the evolution of Internet, our life-style is changing. A number of our
day-to-day activities are being carried out on the Internet. Clearly, the society is evolving
32
towards a networked community with electronic information as the central commodity.
One might term the society of the 21st century as the Networked Electronic Communication
Information Society (NEIS). It is a society in which activities are centred on networks Protocols and Network
Addressing
and the main commodity on the networks is electronic information in digital form.
It is important to realise that with alt its massive presence, Internet is still evolving.
Today’s Internet services are predominantly text and data oriented with only sprinkles
of graphics, still pictures and slow motion video. Experience shows that Internet is slow
for many network applications. Internet is basically designed for data transport. Real
time services like voice and video transmissions experience serious quality problems.
The key to the solution of current Internet problems lies in building Global Information
Infrastructure (GII) that would have adequate capacity and efficiency to support full-
scale services including high quality audio and motion video and high-resolution graphics
envisioned for NEIS.
Information exchange between computers that are connected to a massive worldwide
network cannot happen without standard procedures and sets of rules that govern such
an exchange. A comprehensive collection of such standard sets of rules and procedures
are called communication protocols. Furthermore, every entity on the Internet needs
to be identified uniquely. This is done by assigning a network address to each entity.
Communication protocols and network addressing are the subject matter of this unit.

9.2 WHAT ARE PROTOCOLS?

Let us start understanding protocols. The word protocol has different connotation under
different circumstances. In governments, protocol means a strict official procedure in
state affairs and diplomatic occasions. For example, in India the President is the Head
of the State and there are protocols that govern his/her participation in state functions.
These protocols specify how and where the President will be seated, who would
accompany him/her, how would the dignitaries be introduced etc. In other words, they
specify the accepted code of behaviour in particular situations. They may cover aspects
about appearance (dress code), ways of greeting, conversation, and eating manners.
All these rules help people successfully communicate and work together.
In inter-governmental dealings, the word protocol is used to denote the original draft of
a diplomatic document containing especially terms of a treaty agreed to in a conference
and signed by the parties concerned. You might have heard of Kyoto Protocol, a
document that spells out the terms to be adhered to by the signatories for controlling
and reducing carbon emissions in the world.
In science, a formal record of scientific experimental observations is often called a
protocol. Procedures for carrying out scientific experiments or a record of the course
of any medical treatment are also known as protocols.
The word protocol is used extensively in computers and communications as well. In
computers, protocols deal with interaction between processes, exchange of messages
etc. A process, as you may know, is a program in execution. In communication, protocols
deal with signalling, switching, routing, forwarding, error control, monitoring, and recovery
procedures in the exchange and transmission of information across entities in a network.
A fundamental difference between the two is as follows. While computing protocols
define rules for communication among processes within a computer, the communication
protocols define rules for communication among computers. Both of them, however,
deal with exchange of information. Protocols are generally software programs that
implement the rules for communication. Some protocol functions are implemented in
hardware, particularly those dealing with the movement of bits and bytes. In Section
33
Network Fundamentals 9.3, we briefly discuss computing protocols and study communication protocols in
greater detail in later sections.

9.3 COMPUTING PROTOCOLS

Computing and communication protocols together define sets of rules and procedures
that govern all the information management functions. With electronic information being
the central commodity in NEIS, information management functions become the core of
technological capability in networks. There are seven functions of electronic information
management that are important:
1) Generation
2) Acquisition
3) Storage
4) Retrieval
5) Processing
6) Transmission
7) Distribution
Computing and communication protocols govern all these functions. In general,
information is generated by human thought process, human acts and by happenings in
nature. Human intellectual activity is creative and intuitive and hence may not be amenable
to protocols. Whether technology generates information is a debatable point. When
data is processed in a computer, the output is considered as information. In that sense,
it may be said that computers generate information. But the basic data comes from
nature or human activity. However, machine generation of information can be governed
by protocols.
Among the other functions, storage, retrieval and processing fall in the realm of computing
protocols. The remaining functions, viz. acquisition, transmission and distribution fall in
the class of communication protocols. Transmission and distribution functions may be
collectively called as information dissemination. Transmission refers to bulk transfer
between two main points. Distribution refers to transfer to end points like user computers
or terminals.
Computing protocols are relatively a recent development. As you are aware, information
processing, storage and retrieval are functions performed by application processes.
You are familiar with applications like word processing, spread sheet, power point
presentation and data base management. Computing protocols deal with information
storage, retrieval and exchange among these applications. For example, how do we
import information from word files into spreadsheets or vice versa? Or how do we
import information from word files to power point presentation slides and vice versa?
Computing protocols are being evolved to make such imports fairly easy. Some of the
well-known computing protocol functions include message passing, process
synchronisation and process switching, simple object access and object communication
and data portability.
The idea of computing protocols is to encourage what are known as open systems
design. Open systems follow industry standards and are capable of running on variety
of platforms. For example, open office is an innovation in computing protocols. Many
34 Java products use open computing protocols. Microsoft has recently announced a
number of open computing protocols and has made them available in public domain. Communication
Open computing protocols offer greater opportunity and choice for freelance developers Protocols and Network
Addressing
as they conform to industry standards. In contrast, closed protocols are proprietary in
nature and are vendor specific.
Interoperability in computer systems is the main goal of computing protocols. By
interoperability we mean the ability of different applications to interwork with each
other using common data. User does not have to reformat and copy data from one
application to another. Interoperability principles include:
 Ensuring open connections
 Promoting data portability
 Enhancing support for industry standards
 Driving open approach across competitors.
Although the open approach is currently limited to application packages from the same
vendor, increasingly computing protocols are addressing issues for interoperability across
different vendor products and platforms. Interoperability concept is also applicable to
networked computers.
Self-Check Exercise
Note: i) Write your answers in the space given below.
ii) Check your answers with the answers given at the end of this Unit.
1) Differentiate between computing and communication protocols.
.....................................................................................................................
.....................................................................................................................
.....................................................................................................................
.....................................................................................................................

9.4 COMMUNICATION PROTOCOLS: GENERAL

CONCEPTS
Communication protocols deal with all aspects of communication functions that are
required for information exchange among computers in a network or across networks.
They are designed especially in the context of Internet. We have already discussed in
Unit 8 the protocols that are used for information exchange in LANs. On the Internet,
communication related functions include:
 Breaking up messages into packets
 Packet sequencing and reassembly
 Synchronisation or handshaking for information exchange
 Signalling: start and end of messages
 Switching: routing or forwarding of messages towards their respective destinations
 Connectionless and connection oriented transfers
35
Network Fundamentals  Message encapsulation and de-capsulation
 Format conversion
 Error detection and correction
 Setting up and termination of sessions
 Recover from unexpected loss of connection.
We discuss each one of the above functions in order in the following paragraphs.
Packet formation: As mentioned earlier, in data networks information is transferred in
the form of packets. Packets are of fixed length with an upper bound on the size. For
example, the maximum size of a packet in Ethernet is 1500 bytes. User messages
longer than the maximum permissible packet size need to be broken into multiple packets.
Packet sequencing: As explained in Unit 8, routers take forwarding decisions
independently for each packet. Even static routing algorithms may have more than one
route defined for the same destination. Depending upon the path taken, the packets
may arrive out of sequence at the destination. If the packets belong to the same message,
they cannot be delivered to the user program unless they are properly sequenced. The
concerned communication protocol needs to perform this function of sequencing the
packets of the same message in proper order. Breaking up a message into multiple
packets at the source and reassembling them at the destination are complementary
functions performed by communication protocols.
Synchronisation: A packet transmission cannot start unless the receiving station is
ready. The sending and the receiving stations exchange handshake signals and synchronise
their transfer process. Handshake signals are like sending a query ‘Are you ready?’
and receiving a response like ‘OK, go ahead’. The synchronisation process includes
agreeing on transfer speeds and the required buffer sizes. When the transfer is in progress,
the receiving station may want a pause for reasons like buffer full. Handshake signals
are exchanged to enforce ‘pause’ and ‘resume’ actions.
Signalling: Once synchronisation is achieved, the actual transmission starts. At this
stage, the sending station must signal to the receiving station the start of the packet. This
is usually done by sending ‘start of text (STX)’ bit pattern. Similarly, the end of the
packet is indicated by ‘end of text (ETX)’ bit pattern.
Routing: We have discussed this function in detail in Unit 8.
Transfer modes: There are two fundamental ways in which information transfer takes
place in our life: connectionless and connection oriented. These transfers are analogous
to postal communication and telephone communication respectively. In postal system,
we write a letter, post the same and expect it to reach the addressee. The postal system
delivers on the best-of-efforts basis. While the letters are delivered most of the time,
some get lost somewhere. In telephone communication a connection is first established
between the parties concerned and then the communication takes place.
Encapsulation: Consider a packet traversing six routers before reaching the destination.
Let the source station and the first four routers be on the same network. The last two
routers and the destination station belong to another incompatible network. The fourth
router will now have to encapsulate the user packet to make it compatible to the
destination network. Encapsulation is like putting one envelope (user packet) into another
and writing the addresses differently on the outer envelop. At the destination, the outer
36
envelope (encapsulated packet) is discarded and the original information obtained. Communication
This is termed as de-capsulation. Protocols and Network
Addressing
Format conversion: Sometimes when moving packets between incompatible networks,
pack formats may have to be changed. An example is moving packets between Ethernet
and Token ring LANs, which calls for format conversion.
Error handling: Errors occur in data transmission. These have to be detected and
corrective action taken. Error detecting codes are used to detect errors. There are two
basic techniques available for error recovery. One is when an error is detected in a
packet, it is discarded and the sending station is requested to retransmit the packet.
This technique is called automatic repeat request (ARQ). Handshake mechanism is
used to request retransmission of the packet. The other is to use forward error correction
codes (FEC) that are capable of both detecting and correcting errors at the receiving
end.
Sessions: A variety of tasks are performed on the networks by establishing sessions
between a server and a client computer. Online search of databases, remote job entry,
remote login to a time sharing system and fie transfer between two systems are examples
of different types of sessions. Different sessions have different requirements. For example,
a dialogue may be two-way simultaneous or one-way alternating. A large file transfer
session may call for establishing roll back points to recover from connection failures.
Session protocols perform functions required to establish, successfully execute and
terminate properly different types of sessions.
Packet loss: It is not unusual to experience unexpected loss of connections in networks.
You might have had this experience while accessing Internet. Some Internet browsers
including Microsoft’s Internet Explorer have provision to resume a session that was
terminated unexpectedly, say due to a power failure. Many communication protocols
have features to recover from unexpected connection failures. This is particularly so in
sessions related protocols.
In this section, we have studied the general features that are required in communication
protocols. In the next section, we look at the details of some of the commonly used
communication protocols.
Self-Check Exercise
Note: i) Write your answers in the space given below.
ii) Check your answers with the answers given at the end of this Unit.
2) We use signalling as a matter of fact in our daily life. Give any four examples of
such signalling.
3) Is SMS a connectionless or connection oriented service?
4) When you are typing on a computer terminal, you make a mistake. Then you
correct it. Which one of the techniques, ARQ or FEC you are using? Give reasons.
5) Many word processing packages have auto correct features. Which one of the
techniques, ARQ or FEC is used there? Give reasons.
.....................................................................................................................
.....................................................................................................................
.....................................................................................................................
.....................................................................................................................
37
Network Fundamentals
9.5 COMMON COMMUNICATION PROTOCOLS
The field of ICT is replete with protocols. Hundreds of protocols have been defined for
various purposes. Many are very specialised, some are rarely used and some are defunct.
You are already familiar with computing and communication protocols. There are other
classes of protocols such as data (bits & bytes) transmission protocols, routing
protocols, access protocols, services protocols and applications protocols. As a
user of networks, you need to be concerned with only about a dozen protocols. This is
much like a language dictionary having over 100,000 words and the average vocabulary
of a person being about 4000 words.
Extensively used communication protocols include:
 Internet Protocol (IP)
 User Datagram Protocol (UDP)
 Transmission Control Protocol (TCP)
 File Transfer Protocol (FTP)
 Remote Login Protocol (Telnet)
 Internet Control Message Protocol (ICMP)
 Dynamic Host Configuration Protocol (DHCP)
 Post Office Protocol 3 (POP3)
 Simple Mail Transfer Protocol (SMTP)
 Internet Message Access Protocol (IMAP)
 Cell Switching Protocols (ATM)
 Muti Protocol Label Switching (MPLS)
 HyperText Transfer Protocol (HTTP)
 Wireless Application Protocol (WAP)
 Lightweight Transport Protocol (LTP)
 General Packet Radio Service (GPRS)
 Simple Network Management Protocol (SNMP)
Of the above, the first three protocols, viz. IP, UDP and TCP are basic protocols used
by a variety of Internet services and applications. We discuss them in Section 9.6. FTP
and Telnet are most extensively used service or application level Internet protocols. A
large number of applications on the Internet use what is known as Client-Server
architecture. FTP and Telnet and web browsers also use this architecture. We present
this architecture in Section 9.7. We discuss FTP and Telnet in Section 9.8.
Routers use ICMP to report any abnormal event on the network. An example of an
abnormal event that a router may discover is the outage of the network in some segment.
Such an event may be reported to all other routers on the network as welt as the to the
network management centre. ICMP is also used to monitor the functioning of Internet.
ICMP, however, is not discussed in this course. DHCP is used for managing IP address
38
allocation in local networks. This is an advanced protocol meant for network
administrators and as such we do not discuss the same. POP3, SMTP and IMAP are Communication
all e-mail related protocols. They are not discussed in this unit. Protocols and Network
Addressing
At the data transmission level, viz. transfer of bits and bytes a new convergent switching
technique has emerged in the 1990s. This technique is known as cell switching and the
associated transfer mode is called Asynchronous Transfer Mode (ATM). There is a set
of protocols associated with ATM. We present an introduction to ATM in Section 9.9.
Multi Protocol Label Switching (MPLS) is router-based technique for routing IP packets
fast. An introduction is given to MPLS in Section 9.10.
HTTP is the widely used web access protocol designed to work with desktop and
laptop computers. WAP and LTP are wireless access protocols designed to work with
small portable devices like mobile phones. These protocols are discussed in Section
9.14.
A related protocol is GPRS that is used to send packets over slow-speed wireless
links. GPRS is not discussed here. SNMP is discussed in Unit 11 that deals with network
management.

9.6 BASIC COMMUNICATION PROTOCOLS: IP,

UDP, TCP
9.6.1 Internet Protocol (IP)
Internet protocol (IP) is fundamental to the operation of Internet. All services on the
Internet use IP for sending or receiving packets. No computer can be connected to the
Internet without the IP running on it. Hence all computer operating systems like Windows
provide IP bundled with them. IP software is usually memory resident. IP specifies
exactly how a packet must be formed and how an Internet router should deal with the
packet.
Packet and packet switching are generic terms used in a variety of contexts in ICT. For
example, a network not conforming to Internet standards may use packet switching
and define its own packet structure. In order to distinguish from other packets and to
uniquely identify IP packets, the term IP datagram or simply datagram is used. We
use the term datagram to mean an IP packet in this course module.
Although IP is a communication protocol defining datagram formats and transfer details,
it serves an important purpose that goes almost unnoticed. Once on the Internet, a user
can create and send datagrams to any computer on the Internet irrespective of where
the destination computer is located. The user is unmindful of the presence of different
component networks and a host of routers that interconnect them. Thus, IP makes a
network of networks appear as one giant seamless data network.
An important aspect of IP is that it delivers datagrams on the best-of-efforts basis. The
delivery is not guaranteed. The central idea in IP design has been internetworking and
fast transfer of datagrams and not aspects like recorded delivery. Such aspects are
taken care of higher-level protocol like TCP that we discuss later.
In general, a packet has the structure shown in Fig. 9.1. We use the term payload to
denote the sum total of data handed over to IP for transmission. The payload may be
pure user data or user data encapsulated with other information by any of the
communication protocols.
39
Network Fundamentals The datagram header has a mandatory fixed length part and an optional variable length
part as shown in Fig. 9.1(b). The fixed length is 20 bytes and the variable length can be
up to 40 bytes making the maximum size of the header as 60 bytes. The different fields
of the fixed part are illustrated in Fig. 9.1 (c) where each row is 32-bit or 4 bytes long.
The source and destination addresses are 32-bit each corresponding to IPv4 address
format.
IP addresses have two versions: Version 4 and Version 6 abbreviated as IPv4 and IPv6
respectively. IPv4 uses 32-bit address and IPv6 128-bit. IPv4 has been in use for a
very long time, over 30 years, and most of the computers on the Internet have IPv4
addresses as of now. IPv6 has been introduced recently. Over the years, IPv6 is expected
to replace IPv4 addresses. IP addresses are discussed in detail in Sections 9.13. The
‘version’ field in the header specifies the version to which the header belongs. Version
information in each datagram permits the coexistence of different versions and smooth
transition from one version to another.

Header Payload
(a) A Generalised Packet format

Mandatory Optional
Fixed length Variable length
(b) Datagram header Darts

Version | Header Service Datagram length

Datagram identifier Fragment identifier

Time to Upper layer Header error control

live Protocol
Source Address
Destination Address

Optional Fields up to 10 32-bit words

(c) Mandatory Fields of IPv4 datagram header
Fig. 9.1: IPv4 datagram formats
The maximum size of IP datagrams can be up to 64 k bytes including the header and the
text part. But rarely such a big size is used. Different networks are allowed to set their
own limit for the maximum size of the datagram well below the theoretical limit of 64 k.
This maximum size set by a network is called the maximum transfer unit (MTU) of
that network. This provision further complicates processing of datagrams. If a datagram
is delivered to a network with a size greater than the MTU of the network, then the
datagram needs to be fragmented for transportation within that network and reassembled
at the exit point of that network. In such a case, we need a provision to identify the
datagram and its fragments. IP header fields, datagram identifier* and ‘fragment identifier’
in Fig. 9.1(c) are provided for this purpose. While reassembling the fragments, IP must
know the original protocol from which the fragments came. This is specified in the field
‘Upper layer protocol’. The one-bit ‘M’ field when set to ‘V implies ‘More fragments
to come’. This bit is set to ‘1’ in all but the last fragment. The last fragment will have this
bit set to ‘0’. There may be certain applications where fragmentation may not be
acceptable. The one-bit ‘D’ field, if set to ‘1’ would mean Do not fragment’. In such
40 cases, the route will be so chosen that no fragmentation occur.
You may recall that sometimes packets may wander indefinitely without getting delivered Communication
to the destination due to routing errors. The field Time to live’ is used to exercise control Protocols and Network
Addressing
over such malfunctioning. The field ‘service type’ addresses issues like priority etc.
Other fields in the header are self-explanatory.

9.6.2 User Datagram Protocol (UDP)

UDP provides connectionless service at the user level. It uses IP for this purpose. In
that sense, UDP is a higher-level protocol when compared to IP. Here, a user submits
his/her entire message to UDP with a request for transfer to the specified destination.
User message is a payload for UDP. In turn, UDP encapsulates this with its own header
and passes the same to IP as payload. User datagrams are different from IP datagrams.
User datagrams do not conform to IP standard. They are just chunks of information of
any size. UDP encapsulates the user datagram with its own header to form UDP
datagram. UDP may split a user datagram into multiple UDP datagrams conforming to
IP standards.
UDP datagram is shown in Fig 9.2. Now let us see as to why UDP adds its own header
to the user data. Many application processes or users on a computer may use UDP
simultaneously. Hence, UDP needs to maintain an identity of individual process and its
corresponding destination process. This information is kept in its header in the form of
source and destination port numbers so that the datagram may be delivered to the
correct destination process along with the source identification. In Fig. 9.2 each row is
4 bytes long. With two rows the UDP header is 8-bytes long. The port fields in the
header identify the source and destination processes or applications.

Source Port Destination Port

UDP Length UDP Error
Control
UDP Data or payload
.

Fig. 9.2: UDP datagram structure

Each port field is 16-bit long. The destination port value is used to deliver the user
datagram to the correct application. The destination application may use the source
port value for sending a response to the source application. The port address feature is
the one that distinguishes UDP from IP. Otherwise, the functional capability of UDP is
the same as that of IP. As in the case of IP, UDP messages may be lost, duplicated and
delivered out of sequence.
The value of the UDP length field specifies the total length of the datagram including
data part and the header. The use of UDP error control field is optional. This field is
used only for the header portion of the PDU, i.e. the error control is done only for the
header. The UDP does not perform error control at the datagram level. This must be
taken care of at the application level. The payload supplied by the user or an application
program follows the header. The entire UDP datagram with its header and user data
becomes the payload for IP.
UDP, being a connectionless service functions on the best-of-efforts basis. There is no
delivery acknowledgement in UDP. There is no guarantee of delivery. But it is used
extensively like the postal system. The protocol is simple, efficient and fast. There are a
large number of applications where occasional non-delivery is acceptable. If the
underlying network is reliable, UDP is very effective.
41
Network Fundamentals 9.6.3 Transmission Control Protocol (TCP)
TCP is a connection-oriented service. It uses IP and is at a higher level. In fact, UDP
and TCP are at the same level. TCP is guaranteed delivery service. TCP provides
reliable and error free communication. It achieves this in four ways:
1) Detects errors in datagrams and uses ARQ technique for error recovery
2) Recognises duplicate datagrams and discards all but one
3) Detect lost datagrams and retransmit the same
4) Sequences datagrams received out of sequence
You are already familiar with error detection and the use of automatic repeat request.
As you aware that some routing algorithms send out multiple copies of datagrams to
achieve robustness. TCP checks for duplicate datagrams and accepts only the error
free copy received first. Detection of lost datagrams is done using acknowledgement
and timer mechanisms. Receipt of every datagram is acknowledged by the destination.
At the time of sending a datagram the source initiates a timer with a value within which
the acknowledgement must be received. If the timer expires and no acknowledgement
has been received, the source concludes that the datagram is lost and despatches another
copy. By adopting the above said four mechanisms TCP is able to provide reliable and
error free transmission.
TCP being a connection-oriented service establishes a connection between two
communicating programs before data transfer begins. The service progresses in four
phases as in the following:
 Source requests TCP for a connection by giving the destination identity. TCP
contacts the destination
 Destination responds with a positive acknowledgement
 Data transfer takes place
 Connection terminated
This is very much like what happens in a telephone conversation. Much as the way we
use both telephone and postal systems extensively in our daily life, both TCP and UDP
are used extensively on the Internet. Since TCP and IP are closely interlinked, vendors
bundle both the software routines as part of the operating system. This is why you
always hear of TCP/IP together.
Self-Check Exercise
Note: i) Write your answers in the space given below.
ii) Check your answers with the answers given at the end of this Unit.
6) How many fields are there in the mandatory portion of IP header?
7) Why is fragmentation required while transferring IP datagrams?
8) What action is expected of a router if the field ‘D’ is set to ‘1’ in the IP header?
9) What purpose the port fields in UDP header serve?
10) How does TCP detect lost datagrams?
42
 ..................................................................................................................... Communication
Protocols and Network
..................................................................................................................... Addressing

.....................................................................................................................
.....................................................................................................................

9.7 CLIENT-SERVER ARCHITECTURE

As mentioned earlier, FTP, Telnet and Web browsers are based on client-server
architecture. Client-Server architecture is the most widely used form of computation on
data networks. It has evolved from interactive computing model that was prevalent in
the 1960s and 1970s. In interactive computing, a user interacts with a mainframe computer
via a terminal that may be dumb or smart. The interaction model follows a master-slave
approach. The mainframe computer acts as the master and the terminal as the slave.
The slave terminal is under the complete control of the master computer.
With the advent of personal computers and data networks, the master-slave model of
interaction has given way to peer-to-peer interaction model. Peer-to-peer interaction
permits arbitrary communication among computers on the network. No distinction is
made among the computers. A PC may contact another PC or a large mainframe as
easily. Similarly, a mainframe computer can contact another mainframe or a PC.
Distributed computing has become the norm. Distributed computing means any form of
computation between two or more computers communicating over network.
Computers called servers that provide different types of services are on the networks.
The services are accessible to other computers that are treated as clients of the service-
providing computers. This model of interaction is known as the client-server
architecture. A computer on the network may act both as a server and a client. When
it provides service, it is a server and when it accesses the services of another computer,
it is a client. We may thus say that the client-server architecture is a form of distributed
computing with peer-to-peer interaction.
The client-server configuration is depicted in Fig 9.3. There are two machines and a
network in the configuration: a server machine, a client machine and a data network.

Fig. 9.3: Client -Server configuration

The server and the client interact via the data network. The server provides a set of
information and computational services that are availed by remote clients. As shown in
Fig. 9.3, usually many clients access one server simultaneously. It must be noted that
the server and client machines do not actually interact. It is the server program and the
client program that interact although we normally speak of server and client interaction.
Support for multiple clients is possible only because of program-to-program interaction.
A server creates as many processes of the same program as there are clients logged on
to it. Use of multiprogramming and time-sharing features of the server operating system
makes this possible. The server machine is one, but the instances of a server application
program are many. This is how many users access one web site simultaneously. Client-
server interaction may take place at one of the following three levels:
43
Network Fundamentals 1) Human - Server Program
2) Human - Human
3) Client Program - Server Program
The first case is the most popular one with human client and a machine server. A typical
example is that of user accessing information from a server, say searching a database.
An example of the second case is student - teacher chat session. In online learning,
student is tutored by a teacher. The student is the client and the teacher is the server.
Timed periodic file transfer or email transfer between two or more machines are examples
of the third case.
In all client-server interactions, it the client that always initiates a session. The server is
ready and waiting without doing anything. When a client request comes, the server
program responds. This is like a shopkeeper ready to sell with his shop open but the
actual transaction takes place only when a customer arrives. The server service must
be available on 24 x 7 (24 hours a day, 7 days a week).
Server systems are generally more powerful than client systems. They fall in one of the
following categories:
 PC servers
 Workstation servers
 Mainframe servers
PC servers typically use standard 32-bit microprocessors. They have large RAM and
hard disk capacity. They are ruggedised for continuous uninterrupted running with backup
power systems and cooling systems where required. PC servers must have an operating
system that can handle multiple users, as many client PCs may connect to the server at
a time. Such operating systems are known as network operating system (NOS). Some
of the popular NOS are MS Windows NT, Windows 2003, Novell Netware and
Linux. All the servers are designed to support simultaneous access from many clients.
Workstation servers use high power or custom designed microprocessors. They are
generally 64-bit or 128-bit microprocessor based systems. Workstation servers run
under Unix like operating system that has a rich set of tools for supporting a wide
variety of applications. Unix is a more reliable and secure operating system when
compared to Windows. Linux is a recent addition to the world of operating systems
and is considered a suitable substitute for both Unix and Windows. Linux is available in
the open software domain. Some predict that in future, both PCs and workstations
may run Linux instead of Windows or Unix. However, experience so far has not shown
this to be true.
Mainframe computer based servers are even more reliable and powerful than Unix
workstation servers. Mainframe based servers are often called enterprise servers to
convey the fact that they are more powerful than PC servers or workstation servers.
Client systems are of two types:
 Desktop personal computers
 Mobile stations
The most popularly used desktop systems are Intel microprocessor based computers
44 running Microsoft’s Windows operating system. Such systems are sometimes called
Wintel systems signifying Windows operating systems and Intel microprocessor. The Communication
other class of desktop personal computer is Apple Macintosh. Mobile stations may be Protocols and Network
Addressing
smart cellular phones (like Blackberry or iPOD), notebook computers and personal
digital assistants or tablet PCs etc.
One of the powerful features of client-server architecture is its scalability. An application
may start on a low-end PC and move in steps to a large PC, workstation and mainframe
as the number of users rises. Interestingly, the upgradatton may happen without the user
ever being aware of it.
In client-server architecture all applications have two program parts: a server program
part and a client program part. The server program part is responsible for providing the
specified services and the client part enables access to the services. Hence, anyone
developing applications that would run on a network needed to develop both server
and client parts of the software. The client software needs to be distributed to all client
machines that may be spread all over the world. For example, you cannot access a
server site that stores PDF (portable document format) files without an Acrobat Reader
that is the client software for accessing PDF databases. In the early days, new server
applications used custom-designed client software. Soon, it was obvious that roiling
out client software to thousands of users all around the world is rather time consuming
and expensive. With the arrival of World Wide Web (WWW), most of the network
applications are designed to be web-enabled so that browsers now available with
most of the PCs can access the applications without having to have special client software.
In other words, the client software is embedded in the browsers. Two of the well-
known browsers are Internet Explorer from Microsoft and Netscape Navigator from
Netscape Communications.
Self-Check Exercise
Note: i) Write your answers in the space given below.
ii) Check your answers with the answers given at the end of this Unit.
11) What are the differences between interactive computing and client-server computing
models?
12) Can we say that Internet uses peer-to-peer communication? Why?
13) What mechanisms are used to support multiple clients on the same server?
.....................................................................................................................
.....................................................................................................................
.....................................................................................................................
.....................................................................................................................

9.8 APPLICATION LEVEL COMMUNICATION

PROTOCOLS: FTP, TELNET
9.8.1 File Transfer Protocol (FTP)
FTP is used to transfer files from one computer to another on the Internet. FTP works
in an interactive mode. A repertoire of FTP commands is available for interaction. A
user invokes FTP client application on his/her computer as the first step. The user then
enters the identity of the remote computer from which files are to be transferred using 45
Network Fundamentals ‘open’ command of FTP. The FTP client then invokes TCP to establish connection with
the remote computer. Once the connection is established, FTP server is activated at the
remote computer.
At the next step, the user is authorised to access the remote computer by inputting a
valid user name and password. A valid user account is required for this purpose. When
the authorisation is successful, the user may examine and select a file on the remote
computer by using the list command ‘Is’. He/She then uses FTP ‘gef command to
transfer the file to his/her own computer. FTP client allows a user to transfer a file to the
remote computer from the local computer as well. For this purpose, user invokes the
‘send’ command of FTP. The FTP client application is closed by ‘bye’ command.
FTP recognises only two types of files: text and binary. Any non-text file is treated as
binary file. Examples include audio, computer programs, spread sheets and graphics
data. Text files have to be strictly according to one of the standard character encoding
schemes like ASCII or EBCDIC. If in doubt about the nature of the fife, it is best to
specify binary format. Binary format will transfer a text file as well successfully. However,
transfer of text files is more efficient and faster. Where known, it is a good idea to
specify ‘text’ as the file type. However, If an incorrect type is specified, the resulting file
may be malformed.
There are server systems on the Internet, which make available files to general public.
Examples include servers providing government circulars or legal judgements. Such
public files can be accessed without the user having an account on the server. FTP
client makes this possible by providing an account called anonymous’ with password
as ‘guest’.
Since FTP application runs on client-server model, the FTP server must run under
multiprogramming and time-sharing operating system to enable multiple clients to access
the server simultaneously.

9.8.2 Remote Login (Telnet)

Telnet allows an Internet user to log into a remote time-sharing computer and access
and execute programs on the remote machine. For this purpose, the user invokes a
Telnet client on his/her machine and specifies the identity of the remote machine. Telnet
client makes a connection to the remote computer using TCP. Once the connection is
established, the remote computer (Telnet server program) takes over the user’s display
and issues a login command. The user follows the regular login procedure by giving his/
her account name and password. From then on the user computer behaves exactly like
a terminal on the remote system. When the user logs out, the remote computer breaks
the Internet connection and the Telnet client on the local machine exits automatically.
Remote login is a general access feature. The generality makes it a powerful tool on the
Internet. It enables the programs on the remote computer accessible without having to
make any changes to the programs themselves. The installation of the Telnet server on
the time-sharing system is ail that is required. The telnet client and server together make
the user computer appear as a standard terminal on the remote system. Hence, no
changes are required as far as the application on the remote system is concerned. In
view of this generality, different arbitrary brand of computers can be connected to the
remote system. In effect, any computer on the Internet can become a Telnet client to
any Telnet server on the Internet. Unlike FTP or e-mail, Telnet allows the user to interact
dynamically with the remote system. Due to this, Telnet service is very popular.

46 Telnet sessions may run into occasional problems. The application program on the
remote computer may malfunction or freeze. The local computer then hangs. We need Communication
a mechanism to come out of this situation. Remember that during a Telnet session, two Protocols and Network
Addressing
programs are running: one the program on the remote computer and the other the
Telnet client on the local machine. Telnet makes a provision to switch between these
two programs. Once a Telnet session is established, every keystroke by the user is
passed on to the remote computer. A special combination keystroke, like Ctrl +], is
reserved to revert to the local program. The Telnet client examines every keystroke of
the user before passing on the same to the remote machine. If the special combination
key is pressed, it stops communication with the remote machine and allows
communication with the local client program. The user can then terminate connection
with the remote computer, close the Telnet client and resume local operations.
Self-Check Exercise
Note: i) Write your answers in the space given below.
ii) Check your answers with the answers given at the end of this Unit.
14) Let the IGNOU LIS course modules be available on a fictitious FTP server called
“cm.lis@Jgnou.ac.in”. Write down the FTP commands and responses to download
this unit to your computer.
.....................................................................................................................
.....................................................................................................................
.....................................................................................................................
.....................................................................................................................

9.9 SWITCHING LEVEL CONVERGENCE

PROTOCOL: ATM
There are three major forms of switching techniques used in telecommunication and
networks:
1) Circuit Switching
2) Packet Switching
3) Cell Switching
Circuit switching is the oldest technique used in telephone networks and has been in
existence for over 120 years. Packet switching is about 50 years old used in data
networks like the Internet. Cell switching is the most recent one evolved during mid
1990s used in new telecommunication infrastructure.
Before we proceed to discuss these techniques, definitions of two terms are in order:
channel and circuit. A channel is defined as an information pipe with some specified
characteristics like bandwidth, capacity, level of attenuation and noise immunity. A
channel is a one-way link. A circuit is a two-way link and comprises two channels that
enable two-way information flow between two entities. The two channels of a circuit
need not have the same characteristics. If they do, then the circuit is said to be symmetric.
Otherwise, the circuit is said to be asymmetric. Some authors tend to use the term
channel to mean a physical medium. This is incorrect. A physical medium like optical
fibre may carry several thousand information channels in a multiplexed mode.
47
Network Fundamentals Circuit switching is connection oriented. A circuit comprising two channels for two-way
communication is established between the two communicating entities before the
information transfer begins. The circuit is established using dedicated physical resources.
The physical resource may be copper wires, optical fibres, radio or satellite links or a
combination of these media. The circuit remains dedicated for the communicating pair
until it is released, it is unavailable to any other communication need while dedicated to
the communicating pair.
The main advantage of circuit switching is that once the circuit is established there is a
direct connection between the communicating entities and the network is transparent to
them. The information flows smoothly over the circuit from one end to another. The
information is delivered in proper sequence and there is no possibility of out of sequence
delivery. There is no delay caused by network elements like routers. The main
disadvantage of circuit switching is that the scarce network resources remain dedicated
for the entire duration of the information transfer phase and are heavily under utilised.
Be it a telephonic conversation or computer interaction, mere are pauses during the
session and the dedicated resources remain idle during the pauses.
In packet switching, messages are split as packets at the source and delivered to the
network. The network transports the packets to the destination. Packet switched
networks adopt two different approaches for transporting packets from the source to
the destination:
 Datagram approach
 Virtual circuit approach
You are already familiar with datagram transport and the associated problems of out-
of-sequence arrival at the destination and datagram losses. Virtual circuit approach
was conceived to take advantage of in-sequence delivery of circuit switching while
better utilizing the physical resources. The virtual circuit approach draws upon the idea
of circuit estabtishment as in circuit switching. Instead of establishing a dedicated circuit,
it establishes a fixed route from the source to destination. Since the packets follow the
fixed route, they are delivered in order to the destination. Need for re-sequencing does
not arise.
Virtual circuit makes routing more efficient and reduce the header overhead. As soon
as a virtual circuit (fixed route) is established between a communicating pair, the same
is given a unique identifier called virtual circuit number (VCN). The VCN defines the
source and destination addresses, the message and the route. Hence, VCN together
with the packet number uniquely identifies the packet. VCN plus packet number is
much smaller in size when compared to the elaborate identification described earlier.
Thus the header size and the transmission overhead are reduced significantly. Routing is
also made simpler as the VCN is used to index a table to find out the outgoing link.
There is no analysis of destination address and route determination.
Although virtual circuit concept is a major step forward, it still suffers from possible loss
of packets and non-smooth flow of information. Since routers are involved, queues
may build up and packets may be discarded. Dynamic variation in queue lengths may
result in packets being delivered with different delay times thus interrupting smooth flow
and affecting real time services.
Cell switching is the most recent switching technique evolved during 1990s. The main
objective of cell switching has been to minimise the problems experienced in virtual
circuit switching. This is done in two ways:
48
 To redefine the packet as a cell that is very small in size Communication
Protocols and Network
 To leap forward in the speeds of virtual circuit switching. Addressing
Cell switching is designed to cause minimal network delay ensuring at the same time
efficient utilisation of network resources. You are aware of MTU and the associated
problem of possible segmentation and reassembly. This problem is completely avoided
in cell switching. The entire infrastructure uses a standard cell size of 53 bytes. The cell
has 48 bytes of payload and 5 bytes of header. Now let us understand the merits of cell
switching. Ceil switching is built on a very reliable and ultra fast network infrastructure.
The reliable technology almost rules out cell tosses. Even if a cell or two is lost very
rarely, the effect is unnoticeable in real time services like voice and video. The small size
of the cell makes the loss imperceptible to hearing or viewing. In data services of course,
recovery is required.
Cell switching uses virtual circuit principle. The cells are guaranteed to be delivered in
sequence. Virtual circuit reduces switching overheads significantly and makes switching
extremely fast. For this reason, cell switching is sometimes called fast packet switching.
The networks that use cell switching are called Asynchronous Transfer Mode (ATM)
networks. The reason for this is that the cells of a particular message are not switched
in a fixed time frame say every millisecond. They are switched as they arrive. The
arrival is a mixed bag of cells from different messages or services. They are switched in
the order in which they arrive. Consecutive cells do not necessarily belong to the same
message or service. In other words, the cells of a message or service are not continuous
or synchronous in time. Hence, the term asynchronous transfer is used. Asynchronous
transfer ensures effective utilisation of the network resources. The resources are not
dedicated to one service.
In contrast, in the conventional circuit switched networks the information transfer is
continuous and synchronous. In synchronous transfers, information pieces arrive in fixed
time frame, say one byte every microsecond. In ATM, the cell arrival is not time
synchronous. The time gap between the arrivals of two consecutive cells of the same
message is not fixed but a variable one. However, the variability is very small because
of the high-speed switching of ATM. For all practical purposes, the services perceive
synchronous arrival. ATM is a technique that marks the convergence of both circuit and
packet switching. Hence, ATM protocols are often referred to as convergence protocols.
Self-Check Exercise
Note: i) Write your answers in the space given below.
ii) Check your answers with the answers given at the end of this Unit.
15) Why is cell switching superior to circuit and packet switching?
.....................................................................................................................
.....................................................................................................................
.....................................................................................................................
.....................................................................................................................

9.10 MULTI PROTOCOL LABEL SWITCHING:

MPLS
As you know, virtual circuit makes routing more efficient and reduce the header overhead
by using VCN. The VCN uniquely defines the source and the destination, the message
and the route. Use of VCN reduces the header size and the transmission overhead. 49
Network Fundamentals Routing is made simpler as the VCN is used to index a table to find out the outgoing
link. Multi Protocol Label Switching (MPLS) is an attempt to bring VCN concept to IP
packets. Here, an IP packet is assigned a label that uniquely identifies the destination.
In fact, the IP packet is encapsulated with the label header. The label is then used to
index into a table to find out the outgoing link to be used for forwarding the packet.
There is no examination of the destination address every time. This greatly simplifies
routing overhead and makes the IP packets move faster through the network. This is
particularly useful where large volume data transfers are involved as in the case FTP
service.
MPLS is a router-based solution to improve the router efficiency. This is not a protocol
that runs on any user machine. User machines run only the conventional communication
protocols like TCP and FTP. We need MPLS-capable routers to implement MPLS.
Only MPLS-capable routers can assign labels and handle MPLS packets. There are
two ways in which the labels are assigned to IP packets: data-driven and control-
driven assignments.
In data-driven assignment, when a packet enters a MPLS-capable router, it contacts
the next MPLS-capable router and asks for a label for the destination address. The
next MPLS-capable router in turn connects to the next one and the process is continued
until the destination router is reached. Thus a fixed route is formed for all packets to the
same destination. The first router now encapsulates the IP packet with the label supplied
by the next router and forwards the packet. From then on, the label is used for routing.
The name mufti protocol is used to signify the fact that the MPLS-capable routers can
forward IP packets from a variety o f protocols like TCP and FTP.
In control-driven assignment, a destination router creates labels for all its host computers
and passes them to its neighbours. The neighbours in turn create labels and contact
other neighbours. The process is continued until all the routers acquire the path.
Thereafter, the label is used for routing.

9.11 TELEPHONE AND MOBILE NUMBERING

Every entity in any network needs to be uniquely identified. Otherwise, the entity cannot
be accessed. In telephone and mobile networks the entity is a phone instrument and it
is number that uniquely identifies the entity. In Internet, the entity is a computer and it is
IP address that uniquely identifies the entity. Although it is called an address, IP address
is also a number. At the user level, the addresses are specified by string of characters on
the Internet, (e.g. ignou.com). In a sense, similar character addressing is also available
in telephone networks by way of directories where one looks up the number
corresponding to a name. The addressing or numbering scheme follows a structure.
We discuss the telephone and mobile addressing schemes in this section and the IP
addresses in Section 9.13.

9.11.1 Landline Telephone Numbering

Telephone numbering worldwide follows an international standard set by International
Telecommunications Union (ITU). The details are specified in the standards E.160 -
E.163 of ITU. In ITU parlance, the numbering scheme is called numbering plan. As
per the plan, the world is divide into 9 zones with each zone being identified by a zone
code as indicated in Table 9.1. The zone names in Table 9.1 are representative. For
exact delineation, one is advised to refer to the standards. Europe is given two codes,
as there are many countries there.
50
 Table 9.1: World zones for telephone numbering Communication
Protocols and Network
Zone Code Zone Code Addressing

North America 1 Australia 6

Africa 2 Russia 7
Europe 1 3 Far East 8
Europe 2 4 South Asia 9
South America 5 - -

The structure of the number is illustrated in Fig. 9.4. The maximum size of the number is
12 digits. The first digit is the zone number. The remaining 11 digits are divided between

(a) International telephone number

(b) National telephone number

EC = Exchange code OC = Operator code
Fig. 9.4: Telephone Number Structure
country code (CC) and the national number (NN). The country code is one or two
digits. With the zone code added, effectively the country code is 3-digit long. In common
usage, zone code is not mentioned separately. It is included as part of country code.
For example, the country code for India is mentioned as ‘91’. But to be precise, one
should say that the zone code for India is ‘9’ and the country code is ‘1’. Together they
make ‘91’. The country code is kept to be of variable length. The general principle
adopted is that the countries with large population are assigned short codes of two
digits (1 zone + 1 country). The countries with smaller population are assigned longer
codes of three digits (1 zone + 2 country). For example, in zone 9, Maldives has a 3-
digit country code ‘960’. With just over 200 countries in the world this coding would
work for millenniums to come, in fact, forever. Countries with large population like
India get 10 digits for national number and smaller countries tike Maldives nine digits.
Ten digits provide for a maximum of 10 billion connections. For India with a population
of 1.2 billion, this is adequate for times to come. You may appreciate that the telephone
numbering plan has been designed with farsightedness.
National telephone number has four parts in it as shown in Fig. 9.4(b). Subscriber
Trunk dialling (STD) code may be further subdivided as one digit region code within
the country and one or more digits of sub area codes within the region. In India, eight
regions have been identified for telephone numbering. These are numbered 1 - 8 as
shown in Table 9.2. The region descriptions are indicative and actual area covered is as
per Department of Telecommunications guidelines.
51
Network Fundamentals Table 9.2: Indian regions for telephone numbering

Region C Region C
Delhi, NCR, od
1 U.P. & Bihar od
5
Haryana, Punjab
Mumbai, 2 Orissa 6
Maharashtra,
Kolkata & North East 3 Central 7
Tamil Nadu & Kerala 4 Karnataka & 8
Andhra
The sub area codes are kept to be of variable length. The general principle adopted is
that the sub areas with large population are assigned short codes of two digits (1 region
+ 1 sub area). The sub areas with smaller population are assigned longer codes of three
or more digits. For example, Delhi has a code ‘11’ and Noida ‘120’. Similarly, Mumbai
has a code ‘22’ whereas Bopal, a town near Ahemadabad has the code ‘2707’.
Operator code (OC) is used when there is more than one service provider in an area.
Until the early 90s, India had only the state operator, the Department of
Telecommunications, providing telecom services in the country. But now telecom is
opened up to private operators. We generally have more than one operator in major
cities and towns. The operator code is used to identify the different service providers.
Every service provider has more than one telephone exchange in a city. Exchange code
(EC) identifies the telephone exchange to which the subscriber is connected. Usually
two or three digits are provided for EC. If the number of exchanges exceeds 99, we
need three digits. This is the case in cities like Delhi and Mumbai. In smaller cities and
towns, only two digits may be used.
The last part of the national telephone number is the line number assigned to the subscriber
in the telephone exchange to which he/she is connected. Exchanges are usually designed
to support 1000 or 10,000 subscribers. Accordingly, the line number may have 3 or 4
digits.

9.11.2 Mobile Phone Numbering

Technically speaking, there is no reason as to why mobile phone numbering could not
follow the same numbering plan as the landline phone numbering. After all, the mobile is
another telephone instrument except that it works on radio technology instead of landline
(electrical or optical cable) technology. In fact, initially mobile phones in the United
States used the same numbering scheme as the landlines. But, commercial considerations
have led to a different scheme of numbering for mobile phones. In the beginning, the
cost of mobile technology was relatively higher when compared to the landline
technology. Mobile service providers needed to charge the customers higher. You may
be aware that in the initial days of mobile communications, the incoming calls to mobile
phones used to attract incoming call charges and the outgoing calls used to cost about
six times the landline charges. The charge differential being so high, users needed to
know whether they are calling a mobile phone or a landline phone. Further, roaming
feature of mobile phones and the associated charging policies made the distinct
identification of mobiles phones necessary. Hence, the need arose to distinguish a mobile
phone from a landline phone. Thus was bom a different numbering scheme for mobile
phones.
In general the series of numbers starting with ‘9’ was reserved all over the world for
52 future use while the landline numbering plan was evolved. When mobile technology
came up and a need arose for distinguishing mobile phone numbering, it was decided to Communication
use the ‘9’ series for mobile numbering. As you have learnt, India has 10-digit national Protocols and Network
Addressing
number with its country code being ‘91’. The 10-digit national number starting with ‘9’
was allotted to mobile phones. The ‘9’ series provides for one billion numbers and it
was considered adequate to meet the needs of mobile users in India, particularly because
the cost being high not much penetration was expected. But, the history proved otherwise.
India has over 500 million mobile users in the country as of June 2010. Close to 750
million mobile numbers have already been allotted. If the rate of growth in mobile users
continues at the present rate, we would run out of mobile number space soon. Hence,
Telecommunications Regulatory Authority of India (TRAI) has opened up unused ‘8’
series numbers for mobile users. An interesting fact is that India has only about 450
landline users and the growth rate here is not very significant. Considering this, it is
likely over a period of time that number space for landline users may be reduced and
the space thus freed may be allotted to mobile users. However, it is important to note
that the potential for high-speed applications is much higher in the case of landline
communications. This is because the radio bandwidth is limited whereas the landline
bandwidth is unlimited.
Self-Check Exercise
Note: i) Write your answers in the space given below.
ii) Check your answers with the answers given at the end of this Unit.
16) What is the maximum number of telephone numbers (both landline and mobile
together) that can be assigned using the 10-digit national number?
17) Identify the different components of the international telephone number
911129534336.
.....................................................................................................................
.....................................................................................................................
.....................................................................................................................
.....................................................................................................................

9.12 NUMBER PORTABILITY

Number portability is a feature that allows a telephone user to retain his telephone
number permanently. It is like the PAN (Permanent Account Number) allotted by the
Income Tax Department or the Social Security Number assigned to individuals in the
United States. Such numbers remain permanent for the lifetime of the individual
irrespective of where the individual lives or works with. The number portability feature
implements a similar concept. Imagine that you are given a telephone number once in
your lifetime and that number remains valid for your complete lifespan. Would that not
be very interesting? Number portability attempts to do just that. However, we have a
long way to go in this regard.
Number portability needs to be considered in three situations from the users’ point of
view:
 Change of location
 Change of operator or service provider
53
Network Fundamentals  Change of service from landline to mobile or vice versa
Accordingly, three kinds of number portability are discussed from the telecom network
point of view:
 Location portability
 Operator portability
 Service portability
Location portability implies that if a user moves his residence or place of work from
one locality to another in the same city or moves from one city to another, his/her
telephone number does not change. Both intra-city and inter-city movements have to
be taken care of.
Operator portability implies that if a user moves from one operator to another, say from
Airtel to Vodafone, his/her telephone number does not change. You may recall that the
national number has a field (OC) that identifies the service provider. When number
portability is introduced, OC may lose significance and may just be used by the old
operator to redirect the call to the new operator.
Service portability implies that a user may move from one form of service to another
and yet retain the original number. At present, three forms of service are available:
landline telephone, mobile phone and voice-over-IP. Service portability must ensure
mobility of the user among all these three services.
So far our discussions were limited to one portability requirement at a time. Portability
requirements may occur in combinations as well. The following combinations may arise:
 Location + operator portability. A user may shift from one city to another and may
want to change the operator also at the same time.
 Location + service portability. A user may shift from one city to another and may
want to change from one form of service to another at the same time.
 Operator + service portability. A user may want to change the operator and the
service as well.
 Location + Operator + Service portability. All aspects being changed at the same
time.
Major changes may be required in the telecommunication equipments for implementing
number portability. Hence, although number portability is being talked about for many
years now, its implementation is not wide spread.
Self-Check Exercise
Note: i) Write your answers in the space given below.
ii) Check your answers with the answers given at the end of this Unit.
18) A mobile user moves from the city of Hyderabad to Mangudi, a village in Tamil
Nadu, which does not have mobile network coverage but has landline connectivity.
What portability aspects would come into picture in this case?
.....................................................................................................................
.....................................................................................................................
.....................................................................................................................
.....................................................................................................................
54
 Communication
9.13 IP ADDRESSING: IPV4, IPV6 Protocols and Network
Addressing
As explained earlier, every entity in a network needs to be uniquely identified. Otherwise,
the entity cannot be accessed. In telephone and mobile networks the entity is a phone
instrument and it is number that uniquely identifies the entity. In Internet, the entity is a
computer and it is IP address that uniquely identifies the entity. Although it is called an
address, IP address is also a number. An IP datagram cannot be delivered to the
destination unless the destination is uniquely and unambiguously identified and IP address
does exactly that. Computers connected to the Internet are called hosts and hence the
term host address is used extensively.
There are two versions of IP addresses under active use. The most widely used one is
defined in Version 4 of Internet Protocol abbreviated as IPv4. IPv4 has been in successful
use for about 30 years and it uses a 32-bit address. With 32 bits we can uniquely
address 232 hosts, i.e. approximately 4 billion (4 x 109) hosts. Remember that the world
has a population of over 6 billion. If everyone were to have a computer in this world,
we would not have enough IP addresses to assign to each one of these. Further, the IP
address is a structured one having different address formats. Structuring has the effect
of reducing the effective address space to a much smaller number than 4 billion. With
the rapid growth of Internet over the last 30 years we are now on the brink of running
out of addresses for new machines. It is in this context, a new version of IP, Version 6
abbreviated as IPv6 has been recently standardised and is being introduced on internet.
Addresses in IPv6 are 128 bits long. With 128 bits we can have approximately 256 x
1036 unique addresses. Such an address space is unlikely to run out in the foreseeable
future and must serve the mankind at least for a few millenniums. The two addresses,
IPv4 and IPv6 are interoperable and would coexist for many decades to come. IP
addresses are assigned and managed by a non-profit corporation called Internet
Corporation for Assigned Names and Numbers (ICANN) to ensure uniqueness in
naming and numbering hosts. In this Unit, we discuss IPv4 in detail and IPv6 briefly.
Readers interested in more details of IPv6 may refer to Further Reading material listed
at the end of this unit.
IPv4 address is structured reflecting the objectives of Internet. You may recall that
Internet is a network of networks. Hence, at the level of ICANN, the main interest
concerns networks rather than hosts. ICANN, through its agents around the world
assigns addresses to networks that contain many hosts. As you are aware, each network
is connected to the Internet via a router or a layer-3 switch. The router has an Internet
network address and is capable of forwarding datagrams towards destination networks.
On its own network, it distributes the datagrams to the respective hosts. The host
addresses are assigned by the respective network owners and maintained on the router.
For the purpose of assigning network addresses by ICANN, the networks are classified
under three categories: large, medium and small signifying the number of hosts on the
network. Corresponding to three network categories, there are three address classes:
Class A, Class B and Class C respectively. The general structure of IPv4 address
has three fields as shown in Fig. 9.5. Class A provides for large, Class B for medium
and Class C for small networks. The ‘Class’ field is of variable length of 1 - 3 bits. A
one-bit class field with a value ‘0’ specifies Class A addresses. The 2-bft field with
value ‘10’ and the 3-bit field with value ‘110’ specify Class B and C addresses
respectively. Often, the class field and the network number field together are called
network address and the host field as host address. We also use the same convention
in this unit.
55
Network Fundamentals
Class Network No. Host

Fig. 9.5: IPv4 Address structure

In Class A address, 7-bit pattern following the first bit specifies the network number.
Seven bits provide for 27 = 128 bit patterns. Two of the 7-bit patterns are reserved for
special purposes. They are all zeros ‘0000000’ and all ones ‘1111111’ patterns. The all
zeros pattern implies the local network in which the host itself is located. The all ones
pattern is used for loop back testing of protocols and applications. That leaves us with
126 Class A networks world over. The remaining 24 bits are used for host addresses
supporting up to 16 million hosts on each network. With 16 million hosts on a single
network, Class A represents the largest possible network on the Internet using IPv4
addresses.
In Class B address, 14-bit pattern following the first two bits specifies the network
number. Leaving out the special patterns mentioned above, this means that up to (214 -
2) = 16, 382 medium sized networks may exist on the Internet. Much as in the case of
network addresses, special patterns are reserved for similar purposes in host addresses
as well. Taking this into consideration, each medium sized network may have up to (216
- 2) = 64 k hosts.
In Class C address, 21-bit pattern following the first three bits specifies the network
number. Leaving out the special patterns mentioned above, this means that up to (221 -
2) H” 2 million small sized networks may exist on the Internet and each such network
may have up to (28 - 2) = 254 hosts. Class C networks are ideally suited for small
organisations. They are used extensively. Some small organisations may have more
than 254 hosts, say ranging from 300 to 1000, but may not have as many as 16 k hosts
to warrant a Class B address. This, in fact is the case with organisations like universities,
research laboratories and large corporate houses. In such cases, the organisation is
allotted as many Class C addresses as needed. For example, three Class C addresses
can support up to 762 (3 × 254) hosts. This approach of using multiple Class C addresses
helps in conserving Class B addresses. If Class B addresses were to be assigned to
such organisations, the address space would remain heavily under utilised.
In addition to the above three primary classes of addresses, there are two special
categories of addresses, Class D & E. Class D address is used for multicasting and
Class E is reserved for future use. Multicasting is the distribution of datagrams to many
hosts that have the same address group. Note that broadcasting is the distribution of
datagrams to all the hosts. Hence, multicasting can be called as ‘limited broadcasting’.
For the sake of convenience and clarity, the 32-bit IPv4 addresses are presented in a
dotted decimal notation. The addresses are viewed as four bytes (32 bits) and the
decimal value of each byte is written with periods separating them. As you know, an 8-
bit pattern can have values ranging from 0 —255. An example address in decimal
notation is 183.41.235.7. The equivalent binary address is 10110111 00101001
11101011 00000111. With experience, it is felt that a hexadecimal representation would
have served the purpose of clarity much better. As you know, hexadecimal representation
uses a base of 16 using symbols 0 through 9 and A, B, C, D, E and F. The hexadecimal
representation of this address is B7.29.EB.07.
Consider the case of an organisation that has 9,000 hosts. We would need to allocate
a Class B address for this organisation to avoid allocating too many Class C addresses.
Recall that one Class B address can support up to 16,382 computers. In this case,
56 7,382 addresses are wasted because the same network address cannot be used for
another organisation. This is yet another example of how address space remains Communication
unutilised. The net result is the loss of address space. It was realised, though rather late, Protocols and Network
Addressing
that a large segment of address space remains unused in IP class based address structure.
While address space remained unused with the existing users, addresses for new users
were becoming unavailable. A class definition with incremental number of hosts would
have been far better. By the time this realisation came, the damage had been done. In
order to contain further damage, Internet management introduced a classless addressing
mechanism known as Classless Inter Domain Routing (CIDR) in late 1990s. The
basic idea behind CIDR is to allocate the remaining IP addresses in blocks of contiguous
addresses without any class consideration. With CIDR, an organisation can seek
provision for hosts in powers of 2 such as 256, 512,1024, 2048 and so on. While
network address length is fixed in Class A, B, C networks as 8, 16, 24 bits respectively,
in CIDR the network address length may lie in the range of 8 - 31 bits theoretically. In
practice, however, the range is 12 - 24. Note that the network address length of 8, 16,
and 24 in CIDR automatically correspond to Class A, B and C networks respectively.
While CIDR makes more efficient utilisation of IP address space, it needed major
changes in the routers all over the Internet, as the algorithm for routing has to undergo
a sea change to handle both classed and classless addresses. Routing was relatively
simpler and the router configuration was small with classed addressing. With CIDR the
situation has changed. Routers have to maintain a much larger database including
information about the length of the network address field. This is kept in a 32-bit field
called subnet mask by setting as many higher order bits of the subnet mask to T as the
length of the network address. For example, if the network address length is 20 bits,
the subnet mask in binary notation is 11111111 11111111 11110000 00000000 and in
decimal notation is 255.255.240.0. In decimal notation 255 means all the 8 bits of the
byte, are set to ‘1’ and 240 implies that four higher order bits of the byte are set to ‘1’.
Therefore, twenty higher order bits of the subnet mask are set to ‘1’ in this case.
IPv4 addressing is a classical example of how ad hoc and non-visionary decisions at
the global level can turn out to be messy. It is worth noting that such a mess has never
occurred in other fields like telephone and ISDN numbering. One of the major drawbacks
of Internet is that many such shortsighted decisions exist and more and more ad hoc
solutions are being found. This is clearly due to the lack of rigorous standardisation
process such as the ones followed in ISO and ITU.
The introduction of CIDR and the concept of subnet mask led to a slight modification
to the decimal notation for the IP address. Since the network address is now of variable
length, its length is indicated by a number at the end of decimal notation after placing a
7. Example of new notation is 183.241.060.000/23. Here the network address is of
length 23 bits. The subnet mask is 255.255.254.0
Let us now briefly discuss IPv6 addresses. As mentioned earlier, IPv6 addresses are
128 bits long. This is a very large address space, i.e. 2128 = 1038, i.e. approximately
1028 times the world population. This number is large enough to probably address
almost every little thing on the earth. Therefore, it is impossible that this number space
will ever get exhausted, certainly not for many millenniums. It is expected that in future
items like refrigerators, air conditioners, cars, buses, ships, airplanes, and even bicycles
will all be assigned IPv6 addresses so that they can be controlled and guided from the
Internet.
IPv6 addresses are structured along the IPv4 addresses. The address space being very
large, over 20 classes of addresses have been proposed. Five significant changes have
been introduced in IPv6: 57
Network Fundamentals 1) Large address space provided by 128-bit addresses
2) Flexible multiple header format. There is one base header that is mandatory and is
of fixed size of 40 bytes. More extension headers can be introduced optionally.
3) Improved control options
4) Permits pre-allocation of resources
5) Provision for protocol extension.
There are only five fields in the fixed header portion other than the source and the
destination addresses. These are version, flow label, payload length, next header
and hop limit. You may recall that Ipv4 header has 11 fields in its header excluding the
source and destination address.
Self-Check Exercise
Note: i) Write your answers in the space given below.
ii) Check your answers with the answers given at the end of this Unit.
19) What are the lengths of network addresses in Class A, B, and C IP addresses?
20) How many hosts can be supported in one Class C address? Explain.
21) Write the following binary IP address in decimal and hexadecimal notations:
00011100 10101000 1100110011 00001100. What class of IP address is this?
What is the network number (decimal) in this address?
22) Write the subnet mask in binary for /22 Classless address suffix?
23) How many hosts can be supported in classless address with suffix /27?
.....................................................................................................................
.....................................................................................................................
.....................................................................................................................
.....................................................................................................................

9.14 WEB COMMUNICATION PROTOCOLS:

HTTP, WAP, LTP
As you are aware, World Wide Web (WWW) or simply Web is very popular on the
Internet. These days, a large number of business houses, government agencies and
many individuals have their own web sites. The number of web sites is growing day by
day. Many business houses are currently upgrading their web sites to facilitate electronic
commerce. In this context, it is necessary for you to learn about the communication
protocol used for web access. Before we discuss the protocol let us briefly review the
basics of web.
The concept of Web started in 1989 in France. In 1994, a consortium called World
Wide Web Consortium (W3C) was established. W3C now has many countries as its
members and is responsible for development and standards for Web and its access.
From the users’ point of view, web is a collection of documents scattered all over the
world that are accessible over the Internet. The documents are often referred to as web
pages. These documents are hypertexts as they contain embedded links to other
58
documents. Embedded links are in the form of Uniform Resource Locator (URL) that Communication
contains three parts: a resource name, the identification of the server in which the resource Protocols and Network
Addressing
is located and the protocol that can be used to access the resource. In effect, URL is
unique identifier for a specific resource on the Internet anywhere in the world. An
embedded URL is called hyperlink.
Web pages are of two types: static and dynamic. Static web pages are designed using
a language called Hypertext Markup Language (HTML) that allows a developer to
place text, graphics, sound, video and hyperlinks in a web page Being a mark up
language, HTML defines how documents are to be formatted. In the process, it mixes
the contents and format information. This poses serious problems while editing the
pages. To overcome this deficiency, two new languages called extensible Mark up
Language (XML) and extensible Style Language (XSL) have been developed. XML
sets standards for structuring the contents and XSL for formatting the pages. Thus, the
content and formatting are separated. These languages are being used increasingly these
days.
As you are aware, web access is based on the client-server architecture that was
discussed in Section 9.7. The web browser that acts as the client sends a web page
request using URL and the server responds by returning the requested document. It is
often necessary for the server to keep track of the user preferences for presenting
information. The server does this by storing what are called cookies on the clients
system. Cookies are short strings of data that a server sends along with a web page and
uses the same later for meeting user preferences. The user, however, has the option of
blocking cookies being stored on his/her system.
The protocol used for communication between the web browser client and the server is
called Hypertext Transfer Protocol (HTTP). For secure applications, Secure
Hypertext Transfer Protocol (HTTPS) is used. HTTPS is discussed in Unit 12 on
Network Security. We discuss HTTP below.
HTTP is used universally to access web services all over the Internet. It specifies how
a client may send requests to servers and how the servers may respond. The requests
are sent in the form of ASCII (American Standard Code for Information Interchange)
strings and the responses are received in the form of Multipurpose Internet Mail Extension
(MIME) format. HTTP establishes a TCP connection on Port 80 of the server and uses
the same for sending requests and receiving responses. More than one request may be
sent without waiting for the responses. This is called pipelining of requests.
The first word of the ASCII string is one of the reserved words that specify the operation
requested. For example, the reserved word GET signifies a read request for a web
page and PUT for storing a page. PUT operation calls for authentication of the user.
The authentication information usually follows the PUT request. The information that
follows the operational request is called request header and need to be specified in a
particular format.
In HTTP parlance, these operational requests are called methods. Other methods
include POST, DELETE, and TRACE etc. POST is used append information to an
existing page. It is used in the case of notice and news boards. POST method requires
authentication so that only authorised users can post notice or news. DELETE is a
request for removing the web page and obviously requires authentication. TRACE is a
request for echoing the message that is being sent. This is used for diagnostic purposes.
The response from the server begins with a 3-digit status word that informs the client
whether the request was successfully processed or not. It also indicates the reason in 59
Network Fundamentals the case of unsuccessful processing. Typical failure messages include ‘no content found’,
‘page not found’, ‘page removed’ and ‘forbidden page’ etc.
With the advent of digital wireless access to Internet, considerable interest was generated
in making small portable devices like mobile phones access web using wireless finks.
Two access protocols Wireless Application Protocol (WAP) and Lightweight
Transport Protocol (LTP) were developed for this purpose. These protocols were
optimally designed to work with low bandwidth wireless links and wireless devices
with a slow CPU, small amount of memory and a small screen. Obviously, such
restrictions do not apply to desktop or laptop PCs. The device and link capability
dictated the design of wireless protocols. In designing HTTP, these restrictions were
not there. Over the time, the handsets have been made very powerful and wireless
communication links have also become faster. The latest example in this category is
iPhone4 from Apple Corporation. Accordingly, the later versions of WAP and LTP are
also more sophisticated.
Self-Check Exercise
Note: i) Write your answers in the space given below.
ii) Check your answers with the answers given at the end of this Unit.
24) How do the design considerations differ for wireless web access protocols when
compared with HTTP?
.....................................................................................................................
.....................................................................................................................
.....................................................................................................................
.....................................................................................................................

9.15 SUMMARY
This unit has dealt with two distinct but closely related aspects: Internet communication
protocols and network addressing. Protocols are generally software programs that
implement the rules and procedures for communication. Some protocol functions are
implemented in hardware too. There are protocols for computing purposes as well.
Computing protocols are relatively a recent development. They define rules for information
exchange among processes within a computer. Communication protocols define rules
for exchange of information among computers. They deal with all aspects of
communication functions that are required for information exchange among computers
in a network or across networks.
An overview discussion of the communication protocols in Section 9.4 brings out their
general functionalities like breaking messages into packets, packet sequencing and
reassembly, message encapsulation and de-capsulation, error detection and correction
and loss recovery. A list of commonly used communication protocols is given in Section
9.5. Then, the basic or fundamental protocols without which Internet cannot function
are discussed in detail. They include IP, TCP and UDP. Internet Protocol (IP) is
responsible for transporting packets from source to destination. Transmission Control
Protocol (TCP) provides assured quality services that ensure errorless and lossless
data transmission. User Datagram Protocol (UDP) is a low overhead, fast and simple
protocol that delivers user messages on best-of-efforts basis.
60
The unit then covers Client-Server architecture that is fundamental to running remote Communication
applications on the Internet. It is the most widely used form of computation model on Protocols and Network
Addressing
data networks. It has evolved from interactive computing model of yesteryears.
Thereafter, two application level protocols that use client-server model for communication
are discussed. File Transfer Protocol (FTP) is used for transferring files from one computer
to another on the Internet. FTP works in an interactive mode using a repertoire of
commands. Remote login protocol (Telnet) allows an Internet user to log into a remote
time-sharing computer and access and execute programs on the remote machine.
The unit then focuses on two switching level communication protocols ATM and MPLS.
Asynchronous Transfer Mode (ATM) is the new communication protocol used in basic
telecommunication infrastructure. It uses the principle of cell switching that combines
the advantages of both circuit and packet switching techniques. ATM is extremely reliable
and fast. Routers use Multi Protocol Label Switching (MPLS) to speed up the process
of routing packets across networks.
The unit then turns its attention to addressing entities uniquely on the networks. First,
the numbering plans for telephone and mobile networks are discussed. Both international
numbering and national numbering in India are elaborated. Another important issue, viz.
number portability is then discussed. Number portability needs to be considered at
three levels: Location portability, Operator portability and Service portability.
Addressing in data networks is then discussed. Version 4 of IP address (IPv4) is
discussed in detail bringing out its limitations and merits. Developments in IPv6 are then
briefly presented.
Finally, the unit discusses the web communication protocols. The universally used
HyperText Transfer Protocol (HTTP) is described in detail. Brief features of wireless
web protocols Wireless Application Protocol (WAP) and Lightweight Transfer Protocol
(LTP) are then presented.

9.16 ANSWERS TO SELF-CHECK EXERCISES

1) Computing protocols define rules for communication among processes within a
computer. Communication protocols define rules for communication among
computers connected to the same or different networks.
Computing protocols are concerned with storage, retrieval and processing functions
of information management. Communication protocols are concerned with
acquisition, transmission and distribution functions of information management.
2) Examples of signalling from our daily life:
 A bus conductor’s whistle to stop and start the bus
 Flagging of a sport event like running race
 Indicator lights in cars
 Caller tunes in mobile phones.
3) Small Messaging Service is a connectionless service. One prepares a message
and sends it across expecting it to be delivered. The service is provided on the
best-of-efforts basis.
4) ARQ is the technique used here. You observe (detect) an error, erase (discard) it
and input the right character (retransmit). 61
Network Fundamentals 5) FEC is the technique here. The package detects the error and corrects it
automatically. There is no retransmission by the user. This is forward correction at
the receiving end.
6) Thirteen including source and destination addresses.
7) Internet is a network of networks. Each component network has its own maximum
transfer unit (MTU) defined. If a datagram is delivered to a network with a size
greater than the MTU of the network, then the datagram needs to be fragmented
for transportation within that network and reassembled at the exit point of that
network.
8) There are certain applications that cannot run with fragmentation. For such
applications, the 1-bit ‘D’ field is set to ‘1’. This would mean ‘Do not fragment’.
If a router finds this bit set to ‘1’, it must route the datagram via such networks that
have MTU equal to or greater than the datagram size. Then no fragmentation will
occur.
9) The port fields in UDP header identify the source and destination processes or
applications. Using the information in these fields, UDP is able to deliver the datagram
to the correct destination application. The source port identifies the source
application that is sending the datagram, tf the destination so desires, it can send a
reply datagram to the source application by using the source port address.
10) In TCP, detection of lost datagrams is done using acknowledgement and timer
mechanisms. Receipt of every datagram is acknowledged by the destination. At
the time of sending a datagram the source initiates a timer with a value within which
the acknowledgement must be received. If the timer expires and no
acknowledgement has been received, the source concludes that the datagram is
lost and dispatches another copy.
11) In interactive computing, a user interacts with a mainframe computer via a terminal
that may be dumb or smart. The interaction model follows a master-slave approach.
The mainframe computer acts as the master and the terminal as the slave. The
slave terminal is under the complete control of the master computer.
12) In client-server architecture, a computer on the network may act both as a server
and a client. When it provides service, it is a server and when it accesses the
services of another computer, it is a client. During interaction, the client is not
under the control of the server. Client can do its own computing. Both the client
and the server act independently and hence share the status of being peers. We
may thus say that the client-server architecture is a form of distributed computing
with peer-to-peer interaction.
13) Yes. Internet uses peer-to-peer communication. Any computer can contact any
other computer. No computer is under the control of another. All computers are
considered autonomous and can function independently. Hence, we say Internet
uses peer-to-peer communication.
14) Machines do not communicate in client-server interaction. The interaction is between
the server program and client program running on the server machine and the
client machine respectively. As many instances of server program are activated as
there are clients accessing the service. One instance is dedicated to one client.
This is made possible by using the multi­programming and time­sharing features of
the server operating system.
62
 ftp>open Communication
Protocols and Network
(to) cm.lis@ignou.ac.in Addressing
Connected to cm.lis@ignou.ac.in
LIS Course Module Services at IGNOU
cm.lis@ignou.ac.in FTP server ready
Name: yourusername
Password:…………..
Login OK
ftp> Is
PORT command successful.
Opening ASCII mode data connection for file list
Block 1: Basics of ICT
Block 2: Middleware Technologies
Block 3: Network Fundamentals
Block 4: Internet Tools and Services
Transfer complete
Xxx bytes received in xx seconds
ftp> Is Block 3: Network Fundamentals
PORT command successful.
Opening ASCII mode data connection for file list
Unit 8: Topology
Unit 9: Communication Protocols and Network Addressing
Unit 10: Protocol Architecture
Unit 11: Network Applications and Management
Unit 12: Network Security
Transfer complete
Xxx bytes received in xx seconds
ftp> get
(remote file): Block 3: Network Fundamentals/ Unit 8: Topology
(local-file): mydocuments/topology
PORT command successful.
Opening ASCII mode data connection for file list
Transfer complete
Local: mydocuments/topology
Xxx bytes received in xx seconds
ftp>bye Goodbye.
63
Network Fundamentals 15) In packet switching, there are problems like out-of-sequence arrival at the
destination and datagram losses. There is also the problem of segmentation and
reassembly due to maximum transfer unit (MTU) limitation. Routing overheads
are high in packet switching. In circuit switching, physical resources remain dedicated
leading to their inefficient use. Cell switching overcomes these problems.
Cell switching is designed to cause minimal network delay ensuring at the same
time efficient utilisation of network resources. The physical resources do not remain
dedicated.
Cell switching is built on a very reliable and ultra fast network infrastructure. The
reliable technology almost rules out cell losses. Even if a cell or two is lost very
rarely, the effect is unnoticeable in real time services like voice and video. The
small size of the cell makes the loss imperceptible to hearing or viewing. In data
services of course, recovery is required.
Cell switching uses virtual circuit principle. The cells are guaranteed to be delivered
in sequence. Virtual circuit reduces switching overheads significantly and makes
switching extremely fast. It is for these reasons that cell switching is superior to
both packet and circuit switching.
16) 10-digit national number can support 1010 i.e. 10 billion numbers.
17) The 12-digit international telephone number given is ‘911129534336’. Here ‘91’
stands for country code for India. The country code may further be subdivided as
zone code ‘9’ and country code within the zone as ‘1’. Next ‘11’ stands for area
code for Delhi. The area code may be further subdivided as region code as ‘1’
and sub area code within the region as ‘1’. The following digit ‘2’ stands for
operator code, which in this case is MTNL. Digits ‘953’ is the exchange code and
‘4336’ is the exchange line number for the subscriber. The combination ‘29534336’
is called the subscriber number.
18) Since the user is moving location, location portability is required. Since the user is
a mobile subscriber and there is no mobile coverage available in the new place,
service portability is required. It is assumed the operator is the same in the new
place. Otherwise, operator portability is also required.
19) The lengths of network addresses in Class A, B, and C IP addresses are 8, 16, 24
bits respectively.
20) Class C address has 8 bits for host address. With 8 bits 28 = 256 hosts can be
supported. But the host addresses of all zeros and all ones are reserved for special
purposes. Hence, a maximum of 254 hosts can be supported in Class C address.
21) The given binary address is ‘00011100 10101000 11101001 00001101’. Its
decimal equivalent is 28.168.233.13. Hexadecimal equivalent is 1C.A8.E9.0D.
This is Class C address as the most significant digit is zero. Seven bits following
the first digit gives the network number which in this case is ‘0011100’ and is 28 in
decimal.
22) The subnet mask in binary for /22 Classless address suffix ‘11111111 11111111
11111100 00000000’
23) Classless address suffix is /27. This means 5 bits are available for host address.
Leaving out the special reserved patterns of all zeros and all ones, we can have (25
- 2) = 30 hosts.
64
24) HTTP is designed for desktop and laptop and other high-end computers like Communication
servers. Here, there are no limitations of memory, computing power and screen Protocols and Network
Addressing
size. It also assumes the availability of high-speed data links of at least 64 kbps.
The emphasis on HTTP design is flexibility and powerful features. On the other
hand, wireless protocols have to work with small portable devices like mobile
phones. Here, the screen size is small, the available memory is very low and the
CPU is not powerful. Data link speeds may be as low as 1.2 kbps or even less.
Hence, the emphasis on WAP and LTP is high efficiency with essential minimal
features only.

9.17 KEYWORDS
Client-Server : A computing and communication model used
extensively in Internet
Connectionless : A service or protocol that commences
information transfer without establishing a
connection with the destination
Connection-oriented : A service or protocol that establishes a
connection between the source and destination
before information transfer commences
DHCP : Dynamic Host Configuration Protocol
Encapsulation : The process of covering a packet with another
layer of header with a different format
Error control : The process of detecting and correcting errors
FTP : File Transfer Protocol
HTTP : HyperText Transfer Protocol
ICT : Information Communication Technology
ICMP : Internet Control Message Protocol
IMAP : Internet Message Access Protocol
Interoperability : Ability of different applications to interwork with
each other using common data
IP : Internet Protocol
IPv4 : IP Version 4 using 32-bit addresses
IPv6 : IP Version 6 using 128-bit addresses
LTP : Lightweight Transport Protocol
NEIS : Networked Electronic Information Society
Open Protocols : Protocols that follow industry standards and are
capable of running on a variety of platforms
POP3 : Post Office Protocol Version 3
Protocol : A set of rules and procedures for information
exchange between computers and applications 65
Network Fundamentals SMTP : Simple Mail Transfer Protocol
SNMP : Simple Network Management Protocol
TCP : Transmission Control Protocol
Telnet : Remote Login Protocol
UDP : User Datagram Protocol
WAP : Wireless Application Protocol.

9.18 REFERENCES AND FURTHER READING

Black, U. Computer Networks: Protocols, Standards and Interfaces. 2nd
Edition. New Delhi: Prentice Hall of India, 1999. Print
Homer, Douglas E. The Internet. New Delhi: Prentice Hall of India, 2000. Print
Homer, Douglas E. Internetworking with TCP/IP, Volume I. 3rd Edition. New Delhi:
Prentice Hall of India, 2001. Print
Lin, Yi-Bing. Wireless and Mobile Network Architectures. Singapore: John Wiley &
Sons (Asia), 2001. Print
Mansfield, Kenneth C and Antonakos, James L. An Introduction to Computer
Networking. New Delhi: Prentice Hall of India, 2002. Print
Panko, R. R. Business Data Networks and Telecommunications. 4th Ed. New Delhi:
Prentice Hall of India, 2002. Print
Stallings, W. Data and Computer Communications. 5th Ed. New Delhi: Prentice
Hall of India, 2000. Print
Tanenbaum, A. S. Computer Networks. 4th Ed. New Delhi: Prentice Hall of India,
2002. Print
Viswanathan, Thiagarajan. Telecommunications Switching Systems and Networks,.
New Delhi: Prentice Hall of India, 2010. Print

66