[Notes on Unit-I]: Computer Network

[By: ROHIT KUMAR]

Question No-(1) What do you mean by Computer Network and Distributed Systems?
Answer: Computer Network:-Computer Networks refers to Collection of Computers which are
• Interconnected By wire or wire-less media.
• Able to Exchange Information among themselves on Mutual agreement.
• Are Autonomous i.e. Individual Existence is Visible to user.

In a Computer Network a Lot of other devices (along with Computer) are also needed e.g. Switches, Hubs, Routers,
Bridges, Connectors etc.Each device on Network has its own Identification.

Distributed System & Networked System are two Completely Different Concepts However A Network may support both
type of System-Concepts Simultaneously.

Networked System:-
In Networked Systems Each CPU has its individual existence. These are connected via wire or wireless media.
These share Information & Resources on Mutual agreement. Individual Existence of Computers is visible to the user.
e.g. if user want to access data stored on some other Computer he has to Log on to that Computer.
Examples of Networked Services are FTP, Telnet etc

Distributed System:-
In Distributed System All CPUs are logically assumed as Single Coherent System in spite of their Locations.
Individual Existence of C.P.Us. is not meaningful for the user. Thus Difference is basically in Software & Not in
Hardware. Distributed Concept is usually realized by a Layer (Called Middleware) of Softwares above the O.S.
Example of Distributed Services is WWW

Question No-(2) What are Goals and Applications of Computer Network?

Answer: Goals & Application of Networks:-

FOLLOWING ARE THE COMMON GOALS OF COMPUTER NETWORKS

• Resource Sharing
• Scalability.
• Reliability
• Communication Media
• Centralized Control.
• Better Performance/Cost Ratio.

Resource Sharing:-
With the Help of Network, Peripheral Devices like Printers etc may be shared among Computers
Concept Behind all Such Shared Services is Spooling & Buffering.

Spooling means: - Simultaneous Peripheral operation online

All the requests from Different machines on Network are first stored in some specific files (Called Spooled file) on Hard-
disk. Then as Per the Capability of Shared device requests are send to Device Buffer after Scheduling etc.
From the Device-Buffer Requests are executed by Device. The event of Storing requests First on Hard-disk is Called
Spooling and The event of Sending requests to Device-buffer is Called Buffering.

Scalability:-
With the Help of Network, Performance of Particular System may be increased as Per Load. This is Possible by using
resources of other System which is offered low load at that time. in the network by logically adding other processors to
Highly Load-offered Machine, Performance may be increased.

Reliability:-
With the Help of Network, Reliability of the Data & all work is increased. since their exist a lot of Hard-disks on Network.
Data may be replicated on various disks, If one System is Crashed, work may be continued with others

Communication Media:
Computer Network Provides a Strong Communication Media Among Local People as well as people from all over the
world.

Centralized Management:-
In the Network, Some Computer may be made Server Which may be made responsible for Central Supervision.

Performance By Cost Ratio:-

With the help of Network, Equivalent Work of a Big Mainframe Computer May be done with a large number of small
Networked Computers.

FOLLOWING ARE THE COMMON APPLICATIONS OF COMPUTER NETWORKS

1) World-Wide-Web
2) E-Commerce.
3) Video-on-Demand.
4) Video-Conferencing.
5) E-governance.
6) Business Process Outsourcing.
7) Electronic Money/Data Transfer

Question No-(3): What do you mean by Network Structure? What are the devices used in Computer Network?
Answer: Network Structure:- Network structure refers to the hardware details of the computer network.
Network may be formed in following steps:-
Network may be classified into following groups
1) LAN 2)MAN 3)WAN 4)VPN 5)Wireless-Network
This Categorization is based on the Area, Devices-used & ownership etc.

Local Area Network:-> This is the first level of networking, Computers with Network Interface cards are connected with
the help of switch and hubs. This is restricted to a Single Building usually. Within 10-500 meter usually
Only Needs are Computers, Switches, and Hubs etc.

Switch Provides Point-to-Point Connection. Whereas Hub Broadcast Packet to all Members. Whosoever is detinged
accepts packet, rest ignore.

LAN RESOURCES(COMPUTERS,SWITCHES,HUBS) MAY BE ARRANGED BY VARIOUS WAYS;

Each arrangement follow specific standards & called Topology. E.g. Bus, Star, Ring, Tree etc

Metropolitan Area Network:-> MAN is built by connecting two or more LANs. Restricted within 10 KM usually.
Additional Devices needed are Bridges & Routers. Established usually among different offices of same Organization in
a City. Bridge can connect only Two Same-Type of Networks. i.e. Two-VLANs. Router can connect More-than Two
Different-Type of Networks. i.e. LAN-MANs or MANs-WAN. The topology used in MAN is DQDB.

Wide Area Network:->

• *Built by Connecting various MANs,LANs etc.
• *Resides in a world-wide Large geographically area.

Virtual Private Network:->

• *It may be big-MAN or WAN which is dedicated only for Single Organization.
• *If Separate Resources Like Leased Lines are used among them.
• *than it will be a Private-Network.
• *If Resources of Public Network are used among them,
• *than it will be a Virtual Private-Network.

Question No-(4):What do you understand by Network topologies? Illustrate common type of topologies used in
Computer Networking?
Answer: Topology: Network topologies refer to the diagrammatic representation of the arrangement of network
devices and computers. Such representation illustrates the actual organization of the devices:-
Following are the main types of network topologies used in LAN:-
• Star Topology
• Bus Topology
• Ring Topology
• Mesh Topology
• Tree Topology

a) Mesh Topology: The advantages of a mesh topology are that the two nodes are using the entire capacity of
the link exclusively. In terms of robustness, a mesh topology network can withstand the destruction of anyone of
its components without incapacitating the entire network. Due to its nature, the mesh network physically
prevents any intrusion on the information sent. Total number links required:- n(n-1)/2

To connect all its nodes together, a mesh topology would require n(n-1)/2 wires, or channels. This number
increases exponentially when new nodes are added to the network. Each node itself would, then, be equipped
with n-1 I/O ports to connect itself to the network. The disadvantage of the mesh topology is, simply, the amount
of cabling itself.

(b) Bus Topology: The alternative to the simple network topology is by using a shared media. On such
topology is the bus topology which makes use of a common shared media. In this case, a linear bus network
would dramatically reduce the amount of wiring needed.

Each node is connected to the backbone via a combination of a drop line and a tap. This configuration reduces
the amount of wires when compared to the mesh topology. The backbone itself can be stretched to be
physically closer to the individual nodes themselves, thus reducing the length of the drop line.

The bus topology fails in terms of fault detection and isolation. Should there be a break in the backbone; a
network administrator would have to test the link physically to find the problem. Secondly, since a tap must
physically be connected to the backbone, it would have to undergo constant change in order to incorporate new
nodes.

(c) Tree Topology:

Addition of new networks to a bus topology would be hard. A tree topology, however, is an extension of the
bus. In a tree, the trunk is usually a high capacity device or connection which facilitates data transmission
amongst the branches.

The advantage of a tree topology is usually that the privacy of transmissions is isolated in a particular branch.
Also, if a particular branch is incapacitated, the trunk and the other branches may still function.

The disadvantage is naturally the fault isolation. Should a transmission fails to be sent, a fault detection
exercise may have to cover the entire distance of the cabling. The trunk of the tree must also be properly
configured to be robust.

(d) Star Topology: The star topology is logically similar to a bus topology, though physically different. In a star
topology, every node is connected to a hub, which acts as a central facilitator of transmissions. Each node has
a dedicated link to the hub.
The advantage of the star topology is that fault isolation and detection can be done centrally. The hub can act
as a monitor to network conditions. In some more intelligent hubs, multiple transmission channels can be
directly be re-routed to the destination, simultaneously

While the hub is the star topology’s chief advantage, it is also the main disadvantage of the topology. The
network is fully dependent on the hub for it to work. Cabling sometimes poses a problem to the star topology
too, as each node must have a dedicated link to the hub making for a potentially messy infrastructure.

(e) Ring Topology: The final structure to link a network is the ring topology. In the ring topology, a node has
direct connections only to the two nodes on either sides. This “chain” would then provide a unidirectional
transmission channel for all the nodes connected to it.

The ring topology provides a relatively easy way installing and reconfiguring a network. The process of adding
and removing a node would have to only consider the two nodes beside it. Fault isolation can be easily done
when a station has not received any signals for a time. It would then issue an alarm to the administrator. The
disadvantage of a ring topology is often the unidirectional traffic it can handle. If a node is down, signals can no
longer circulate amongst the network, thus preventing the network from operating further.

Question No-(5): What do you understand by Network Architecture? What is protocol? Explain
Functions of different OSI Layers?
Answer: Network Architecture: - In Network Architecture we are supposed to understand software details
necessary to realize the network. This is based on divide and conquer concept. All the functions & activities of
Computer Network are classified into different groups. Each such Logical group is called Layer in Network
Terminology. All such layers are arranged in some scientific & Hierarical order.
 Based on the functions specified for each layer programs are developed to realize them.
Programs are also developed to Co-ordinate among different layers

Protocol
Protocol refers to Set of Rules, Regulations & Specifications
• Based on which Programs are developed to realize Network
• Based on which Hardware devices are arranged in different Layers.
• Which specifies Format of Data-tranfer, Method of Computer-identification etc.
• Which have to be followed by all devices of that Network.
Each protocol corresponds to specific Layer. Different Protocols resides in different layers.

All layers with Protocols, which are arranged in some hieratical order, is called as Protocol-model or Network
Architecture.Communication between two Computers may be assumed as Communication between Layers of one
computer with layers of other.

Each layer of one computer communicates with same layer on other Computer
Layers are arranged in such manner so that at cursory glance Communications seems only between uppermost layer.
But if as we consider things upto more refined level. Communication seems between uppermost-layer to second-layer
of one-Computer to second-layer of Second-Computer to uppermost-layer of second-Computer.
if we continue considering things more refined level this will goes-on similarly upto last-layer.

There are following common terms used in Any Protocol Model

Services:-refers to set of operation made available by one layer to other
Interface:-Through which Services offered.
SAPs:-Actual Transfer Points at Interfaces.
Protocol:- Single set of Ruls & Regulations, e.g. SMTP
Protocol-Model:-Hierarical Collection of Protocols of all Layer. e.g. TCP/IP
Protocol-Reference-Model:- Guidelines not Actual e.g. OSI
Protocol-Stack: - Protocol model with at most one Protocol at each Layer.

OSI Protocol-Reference Model:

The most common example of Protocol-model is OSI-Reference Model. These are the Guidelines Laid down by ISO for
developing Protocol-models. This is Called OSI-Open System Interconnection because OSI-Based Systems are open to
connect any other System.

There are total Seven Layer in it. At cursory glance Communication seems between Uppermost-Layers. But in actual,
Communication take place through More Refined layers.
Packet is transmitted after Passing through all Layers.
Main Responsibilities for Each layer may be described as below:-

Application Layer:-
• Enable user to access network.
• Provide user-Interface.
• Support for various services as E-mail, File-transfer, remote-login etc.
• Network Management.

Examples of Application layer Protocols are FTP, SNMP, SMTP etc.

Presentation Layer:-
• Compatibility between Computers with different internal Data-format/O.S.
• Data Encryption & Decryption for Security.
• Compression & Decompression.

In actual no Protocol is available separately to support only Presentation layer tasks, Application layer Protocols
Performs Presentation layer tasks.

Session Layer:-
• Establish, maintain & synchronize the connections.
• Validation & Authorization of users.
• Ensure Direction of flow i.e. Half or full-duplex.
Tasks of this layer are also performed by application layer Protocols.

Transport Layer:-
• Host-to-Host i.e. Source to Destination delivery of complete message
• Segmenting message into Packets at source.
• Assigning Sequence number to packets.
• Reassembly packets into message at destination.
• Reliable delivery of complete message in right order.
Examples of Transport layer Protocols are TCP, SPX, NCP etc.

Network Layer:-
• Receive Packets from Transport layer
• Routing through optimized path.
• Congession Control of traffic.
Examples of Network layer Protocols are IP, IPX etc.

Data-Link Layer:-
• Segmenting Packets into frames
• Error detection & Correction
• Flow control.

Physical Layer:-
• Bit-by-bit reliable Data-transfer.
• Electrical & Mechanical Specifications of Link.

Question No-(6): What are main differences between TCP/IP and OSI-Reference Model?
Answer: Following are the differences between TCP/IP and OSI-Reference Model.

SN Issue OSI-Reference Model TCP/IP

1 Implementation This is guideline only This is actually implementation developed
based on OSI-Reference model.
2 Layers Total 7 Layers Total 5 Layers
3 Names of the Application, Presentation, Session, Transport, Application, Transport, Internet, Data Link
Layers Network, Data Link and Physical. and Physical
4 Definitions Meaning and specifications for services, Meaning and specifications for services,
protocols and interfaces are not clearly protocols and interfaces are clearly defined
defined for each layer for each layer.
5 Transport Layer Only Connection-Oriented Services are Both Connection-less and Connection-
services supported by OSI-Reference model in Oriented Services are supported by TCP/IP
Transport services. model in Transport layer.
6 Network Layer Both Connection-less and Connection- Only Connection-Oriented Services are
Services Oriented Services are supported by OSI- supported by TCP/IP model in Network
Reference model in Network layer. layer.

Question No-(7):- What do you understand by Delay-Analysis in Network Topology Design?

Answer:- Concept of Delay:- In Network Terminology, Delay refers to total Time Consumed during Message
Transferring from sender to receiver.

There are three types of Delay.

1) Propagation-Delay:->>Due to distance
Tp=d/V, where d=Distance V=velocity.

2) Transmission-Delay:->>Time taken by Frame to come out fro Transmitter.

Tt=L/U where L=Frame-size U=Data-rate.

3) Node-Delay or Processing-Delay :->> Time taken at each node.

Tn=Time consumed at each node for Processing etc.
Delay-Consideration in Topological Design.

*Propagation Delay can not be reduced, as it depends upon distance between transmitter-receiver & velocity of Signal.

*Transmission Delay can be reduced, by increasing the data-rate of the line. It can also be reduced by limiting the
size of frame.
*Node delay can be reduced, by more dedicated paths. i.e. in Circuit-Switching, However Call-Set-up time increased
here.

Question No-(8): What do you understand by Switching? Explain three kind of switching?
Answer: - Switching: - Refers to the concept of establishing path from source to destination to transfer data
between these two.

Switching is of three types

(1)Circuit Switching.
(2)Packet Switching.
(3)Message Switching.

Circuit Switching
In circuit switching first of all optimized path is established, complete data is transferred then through that path. Main
advantage of this kind of switching is reliability; time-consumption in path establishment at beginning is the main
disadvantage of circuit-switching.

Packet Switching
In packet switching instantaneous optimized path is established between two nodes. One packet of data is transferred
then through that path. Similarly further Optimized Paths are decided and packet transfers.
Main problem of packet switching is un-reliability of packets; paths may lost in between of the path. As far as
advantages concern is better time-response multiple paths in parallel may be used by different packets of the same
message. Time is not consumed at beginning of the communication in setting up the path.
Message Switching
Instantaneous Optimized Path is established between two nodes. Complete message is then transfered to Connected
Path. In Packet-Switching different packets of message may go through different paths.
But in message-switching all packets of message will go through same Path.
Message switching is the middle choice between circuit and packet switching in terms of reliability and time-
consumption.

Question No-(9): What are various kinds of transmission media? Illustrate with examples?
Answer: Transmission media can be classified into two categories:-
(a) Guided Media—(i.e. Wired media) (b) Un-Guided Media—(i.e. Wireless media)
(a) Guided Media:- Guided Media covered following three kind of media:-
– Twisted Pair Cables (Unshielded, Shielded)
– Coaxial Cables
– Optical Fiber Cables

Twisted Pair:- Twisted pair refers to the collection of two insolated wires that are twisted with one another.
Twisted pair can be of two types:-

UTP (Unshielded Twisted Pair) :

• UTP cables are the most common telecommunications medium.
• The frequency range of the twisted pair cables enable both voice and data transmission.
• Usually consists of two copper wires wrapped in individual plastic insulation.
• Previously, two flat wires were used for communications but interference caused extremely high levels of
damaged signals.

Noise in parallel transmission

Noise in Twisted Pair Transmission

 Advantages of UTP cables:
– Cost
– Ease of Use
– Flexibility
– Ease of Installation

The Electronic Industries Association (EIA) has developed standards to gauge the quality of cables.
• Category 1 – basic twisted pair cabling used in telecommunications system
• Category 2 – transmission up to 4Mbps
• Category 3 – transmission up to 10Mbps
• Category 4 – transmission up to 16Mbps
• Category 5 – transmission up to 100Mbps
• UTP cables come with connectors which is similar to that of telephone jacks. Male connectors snap into female
connectors with a lock to hold them in place.

STP (Shielded Twisted Pair):-

• An alternative to UTP
• The metal mesh around the insulated wires eliminates a phenomenon called crosstalk.
• Crosstalk occurs when one line picks up some of the other signals traveling down another line.
• Shielding each pair individually can eliminate most crosstalk.
• STP shares the same standards and considerations as UTP except that its more expensive.

Coaxial Cable
• Coax cables operate at a higher frequency range.
• Due to the construction which is more noise resistant and is very durable
• Due to their ability to transfer more information than standard telephone cables.
• Was primarily used to connect cable television. Now, coaxial cables are used as backbones of bus topologies.
• Connectors for coax cables are T-connectors and terminators.
Optical Fiber Cable
• Optical fiber technology differs from the previous conductive cables in that it uses light instead of currents.
• Before discussing about optical fiber, we need to explore the rationale behind its technology, which is light
Nature of Light
• The main criteria of light which matters toward the context of optical fiber technology are as follows:
– Refraction
• Angle of incidence
• Angle of refraction
– Critical Angle
– Reflection
Refraction
• Describes how light changes direction when it enters a different medium, due to its sudden change in speed.
• The angle of incidence describes the degree of deviation from the vertical axis before the source enters the new
medium
• The angle of refraction refers to the deviation degree from the vertical axis after the change of medium
• The angle of incidence will always be larger than the angle of refraction when light is traveling into a medium
which is more dense.
• And the angle of refraction will be larger when light travels into a medium less dense.

Critical Angle
• If we increase the angle of incidence when light is traveling to a less dense medium, we know that the refracted
angle will stray towards the horizontal
• At some point in the process, the refracted angle will be traveling along the horizontal.
• The angle of incidence will then be known as the critical angle.

Optical Fiber
• Optical fiber utilizes reflection to guide light along.
• Optical fiber cables consists of a core surrounded by a less dense cladding.
• Both core and cladding can be made out of glass or plastic
• The difference of density is calculated to a point that light is reflected along the wire instead of refracted into it.
Advantage of Fiber Optic
• Noise resistance
– Because optical fiber uses light, it is completely immune to the effects of noise
• Less signal attenuation
– Transmissions can travel large distances before requiring regeneration
• Higher bandwidth
– Fiber cables can transport much higher data rates than twisted pair or coax cables. Currently, data
rates are limited by transmission and reception technology, not the medium.

Disadvantage of Fiber Optic

• Cost
– Due to their need for perfection in the cable’s core, manufacturers have to undergo very precise
procedures.
– Also, a laser transmitter costs significantly more than electrical generators
• Installation
– The entire length of the fiber cable must be crack-free and all splices must be perfectly fused. Also, the
entire network must be light-proof.
• Fragility
– Because of its nature, fiber cables are not very portable as shocks could crack the fiber and render it
useless.

Question No-(10): What are the characteristics of ISDN ? Illustrate various reference points and devices used in
ISDN network?
Answer: ISDN (Integrated Service Digital): ISDN is a WAN technology based on digital communication, media is
usually optical fiber. ISDN connections can support multiple services such as voice, internet and video conferencing
services through same telephone line.

ISDN Connections can be of two types:-

(a) BRI (Basic Rate ISDN):- In BRI two B channels are used to carry data and one D channel is used to carry control
information. Every B channel is of 64 Kbps and one D channel is of 16 Kbps.

(b) PRI (Primary Rate ISDN):- In PRI 23(T1) or 30(E1) B channels are used to carry data and one D channel is used to
carry control information. Every B channel is of 64 Kbps and one D channel is of 64 Kbps.
ISDN-Network Structure:- A typical scenario for ISDN-Network may be as follows:-
The Meaning of various ISDN reference points and special devices are mentioned below:-
ISDN-Reference Point:-

R: The connection between a non-ISDN Terminal Equipment (TE2) and a Terminal Adapter (TA), e.g. an
RS-232 serial interface.

S: References the points that connect into the customer switching device Network Termination type 2
(NT2) and enables calls between the various types of customer premises equipment.

T: Electrically identical to the S interface, it references the outbound connection from the NT2 to the
ISDN network or Network Termination type 1 (NT1).

U: The connection between the NT1 and the ISDN network owned by the telephone company.

ISDN-Devices:-