Faculty:

BCS502 – COMPUTER NETWORKS Prof. Ashok Herur

[Link]@[Link]
Mobile: 91641 01399
Course outline

Module 1:
Introduction: Data Communications, Networks, Network Types, Networks
Models: Protocol Layering, TCP/IP Protocol suite, The OSI model, Introduction to
Physical Layer: Transmission media, Guided Media, Unguided Media: Wireless.
Switching: Packet Switching and its types.
Module 2:
Data Link Layer: Error Detection and Correction: Introduction, Block Coding,
Cyclic Codes. Data link control: DLC Services: Framing, Flow Control, Error
Control, Connectionless and Connection Oriented, Data link layer protocols, High
Level Data Link Control. Media Access Control: Random Access, Controlled
Access. Check Sum and Point to Point Protocol

Computer Networks - Prof Ashok Herur 2

Course outline

Module 3:

Network Layer: Network layer Services, Packet Switching, IPv4 Address, IPv4
Datagram, IPv6 Datagram, Introduction to Routing Algorithms, Unicast Routing
Protocols: DVR, LSR, PVR, Unicast Routing protocols: RIP, OSPF, BGP, Multicasting
Routing-MOSPF

Module 4:

Transport Layer: Introduction, Transport-Layer Protocols: Introduction, User

Datagram Protocol, Transmission Control Protocol: services, features, segments,
TCP connections, flow control, Error control, Congestion control.

Computer Networks - Prof Ashok Herur 3

Course outline

Module 5:

Application Layer: Introduction, Client-Server Programming, Standard

ClientServer Protocols: World Wide Web and HTTP, FTP, Electronic Mail, Domain
Name System (DNS), TELNET, Secure Shell (SSH).

Computer Networks - Prof Ashok Herur 4

Recommended Textbooks

1. Data Communications and Networking

- Behrouz A. Forouzan
Tata McGraw Hill,2013, 5th Edition

2. Computer Networks

- Andrew S. Tanenbaum and David J. Wetherall.

Pearson Education, 5th Edition.

Computer Networks - Prof Ashok Herur 5

Assessment details

• Continuous Internal Evaluation (CIE) : 50% (50 marks)

• Theory part: 25 marks

• Two Internal assessments – 15 marks
• Assignments and Quizzes – 10 marks
• Should score a Minimum of 40% of the above maximum marks (10 marks out of 25)

• Lab part: 25 marks

• Conduction of experiments and Record writing – 15 marks
• Lab test – 10 marks
• Should score a Minimum of 40% of the above maximum marks (10 marks out of 25)

Computer Networks - Prof Ashok Herur 6

Assessment details

• Semester End Exam (SEE) : 50% (50 marks)

• Will have 5 either / or questions – One from each module, to be answered in 3 hours.

• Minimum passing mark for the SEE is 40% of the maximum marks (20 marks)

• For an overall Pass grade, one must score a minimum of 40% of the CIE and
SEE marks taken together.

Computer Networks - Prof Ashok Herur 7

 Module 1

Introduction to Networks,
&
Physical layer

Computer Networks - Prof Ashok Herur 8

Main topics in Module 1

• Data Communications,
• Networks, and Network Types,
• Network Models: Protocol Layering
• The OSI model
• TCP/IP Protocol suite
• Introduction to Physical Layer: Transmission media, Guided Media, Unguided
Media: Wireless.
• Switching: Packet Switching and its types.

Computer Networks - Prof Ashok Herur 9

Introduction to Computer Networks

• A computer network is an interconnection of autonomous computers.

• Benefits:
• Resource sharing
• Data sharing: Consistency, Easy updating, Saving of storage space
• Software sharing: Saving of cost and storage space
• Peripheral sharing: Saving of cost
• Electronic communication: Fast, cheap, secure, convenient
• Higher reliability and easy scalability compared to a single large computer
• Cost-benefit ratio is better compared to a single large computer.

Computer Networks - Prof Ashok Herur 10

Separate network for Data communication ?

• We already have a very extensive telephone network for voice communication.

• Why can’t it be used for data communication?

• In fact, in the late 90’s, it was used in the form of “dial-up” connections
that offered a bandwidth (bit-rate) of 64kbps.

• Later, DSL (Digital Subscriber Link) technology was used on the same
infrastructure to offer a much higher data rate.

Computer Networks - Prof Ashok Herur 11

Voice communication vs Data communication

• Voice is analog, while data is primarily digital

• Voice is continuous, while data is usually discrete

• Latency has to be very low in Voice networks

• Zero tolerance for Jitter, in Voice networks

• Error control is very important in data communication

• Data communication has to be more secure

• Voice needs uniform bandwidth, while data has varying needs

Computer Networks - Prof Ashok Herur 12

Classification of Computer Networks

• Based on geographical area:

• LAN: Local Area Network
• MAN: Metropolitan Area Network
• WAN: Wide Area Network
• Based on transmission technology:
• Point-to-point Network
• Broadcast Network
• Based on ownership:
• Private Network
• Public Network

Computer Networks - Prof Ashok Herur 13

Local Area Network (LAN)

• LAN is a privately owned network that operates within a single building (office
or factory) or a few nearby buildings (university campus).

• LANs connect many PCs and other peripherals (printers, etc) to exchange
information between them.

• The interconnection can be using wired links or it can be wireless or even a

combination of the two.

• Typically, wired LANs run at speeds of 100 Mbps to 1 Gbps, have a very low
delay (milliseconds) and encounter very few transmission errors.

Computer Networks - Prof Ashok Herur 14

Local Area Network (LAN)

• The most popular LAN, popularly called Ethernet, is a wired LAN.

• While the older Bus-based topology enabled a Broadcast transmission, the

newer version, Switched Ethernet, is based on many point-to-point links.

Computer Networks - Prof Ashok Herur 15

Local Area Network (LAN)

• Wireless LANs (WLAN) use an Access Point (Wireless Router) to transfer data
packets between the devices and to the wired network.

Computer Networks - Prof Ashok Herur 16

Metropolitan Area Network (MAN)

• MAN is a privately owned network that operates within a single city.

• The current day MANs evolved from the Cable TV networks that were
established in cities to overcome the poor over-the-air television reception.

• They used some unused part of the cable spectrum to provide 2-way internet
service.

• Data Over Cable Service Interface Specifications (abbreviated as DOCSIS) is a

globally-recognized telecommunications standard that enables high-
bandwidth data transfer via existing coaxial cable systems.

Computer Networks - Prof Ashok Herur 17

Metropolitan Area Network (MAN)

• WiMAX or Worldwide Interoperability for Microwave Access, is another MAN

technology aimed at providing wireless data over longer distances (compared
to WiFi) in a variety of ways.
Computer Networks - Prof Ashok Herur 18
Wide Area Network (WAN)

• A WAN spans a large geographical are, often a country or a continent.

• The Internet is the interconnection of many WANs (who wish to be a part of it)

• The communication links and the switches (routers) that enable

communication between the end devices (called ‘hosts’) together make up
what is termed as sub-net.

• The transmission of messages is in the form of packets, using packet switching.

• Voice messages are typically sent using Circuit switching.

Computer Networks - Prof Ashok Herur 19

Circuit switching

• A end-to-end connection is established before the message is transmitted.

• This is dedicated till the communication is disconnected (like a voice call).

• Results in low latency and zero jitter.

• Drawbacks:
• Set-up time (to establish the connection);
• A huge wastage of Bandwidth when the data transfer is ‘bursty’ (as it
usually is).

Computer Networks - Prof Ashok Herur 20

Circuit switching

Computer Networks - Prof Ashok Herur 21

Packet switching

• There is no dedicated, end-to-end connection.

• The message is divided into small ‘packets’.

• Each packet has information of the message to which it belongs, as well as the
destination address.

• They are received, processed and forwarded by routers in the network.

• At each router, an incoming packet is stored in a buffer (queue), processed

(destination address is seen and analysed) and forwarded on the link that is
seen as part of the best route to the destination.

Computer Networks - Prof Ashok Herur 22

Packet switching

• Advantages:
• Since the bandwidth of a link is not dedicated (but shared with packets
from other hosts), the utilisation of the bandwidth is much higher.
• No set-up time.

• Drawbacks:
• Higher latency (due to waiting time at each of the routers).
• Since different packets of the same message may take different routes,
they may not all take the same time to reach the destination (Jitter).
• Packets may not reach the destination in the proper sequence.

Computer Networks - Prof Ashok Herur 23

Packet switching – Hosts sharing the bandwidth

Computer Networks - Prof Ashok Herur 24

Packet switching – Packets not arriving in sequence

Computer Networks - Prof Ashok Herur 25

Virtual Circuit (VC) Packet switching

• The Datagram Packet Switching explained earlier is used on most WANs,

including the Internet.

• In another method, called Virtual Circuit Packet Switching, all the packets of a
message take the same route (called the Virtual Circuit or VC).

• This route is decided before the first packet is transmitted and the routing
information conveyed to the routers (they are called Switches in this case)
along the route.

• Each router (switch) examines the VCI (VC identifier – a globally unique
number) in the packet and forwards it on the appropriate link.
Computer Networks - Prof Ashok Herur 26
Virtual Circuit (VC) Packet switching

Computer Networks - Prof Ashok Herur 27

Virtual Circuit (VC) Packet switching

• Here, packets will always arrive at the destination in proper sequence.

• However, some time has to be spent in the beginning to establish the VC and
perform the ‘handshake’ with the receiver (informing about the transfer of
message packets with a particular VCI).

• In a variation called Permanent Virtual Circuit (PVC) Packet Switching, the VC

need not be identified for each message between A and B. It is fixed forever,
unless changed.

• This is used, for example, to connect 2 offices with a lot of continuous data
flow between them.
Computer Networks - Prof Ashok Herur 28
Virtual Circuit (VC) Packet switching

• As part of the deal with the network operator, a certain QoS is ensured.

• QoS parameters generally include a minimum Bandwidth and an average

Bandwidth, along with an average transmission time.

• VC networks are relatively costly to implement.

Computer Networks - Prof Ashok Herur 29

Network Reference models

Computer Networks - Prof Ashok Herur 30

OSI Reference model

• The OSI (Open Systems Interconnection) model, created in 1984 by ISO, is a

reference framework for the design of networks.

• The task of transmitting data between computers is divided among seven

‘layers’.

• While the hosts have all the 7 layers implemented in them, other devices may
have less than 7 layers.

• The model adopts a Hierarchical format.

Computer Networks - Prof Ashok Herur 31

OSI Reference model

• Main characteristics:
• Each layer has a well-defined function to accomplish.
• Each layer is implemented with a combination of hardware and software.
• A lower layer provides a service to the one above it (in the hierarchy).
• Each layer has it’s own Protocol, that is transparent to the layers above it.
• The OSI Reference Model only provides a framework and does not define
the actual implementation or the protocols to be used.

Computer Networks - Prof Ashok Herur 32

 A communication model

Protocol L3
Person 1 Person 2

Office 1 Office 2
Protocol L2
Translator 1 Translator 2

Protocol L1
Clerk 1 Clerk 2

Computer Networks - Prof Ashok Herur 33

 A communication model

Protocol L3
Person 1 Person 2

Office 1 Protocol L2 Office 2

Translator 1 Translator 2
Intermediate office

Protocol L1 Pr. L1
Pr. L1
Clerk 1 Clerk 3 Clerk 2

Computer Networks - Prof Ashok Herur 34

OSI Reference model

Computer Networks - Prof Ashok Herur 35

Application layer
• This layer provides the user interface by using various applications (Browsers,
Messengers, etc)
• Functions:
• File management by allowing users to access, retrieve and manage the files
on a remote computer.
• Allows users to send and receive emails.
• Provides distributed database sources and global information about
several objects.
• Protocols used: SMTP, POP3, FTP, HTTP, SNMP.

Computer Networks - Prof Ashok Herur 36

Presentation layer
• This layer is responsible for converting data into a format that is understood
understood by the network components.
• For example, translation of the characters using ASCII or EBCDIC.
• At the transmitter side, it encrypts the data, and decrypts it back into plain text
on the receiver side.
• Protocols used: RSA, AES, DES, etc
• It compresses data and reduces the number of bits needed for network
transmission.
• Protocols used: JPEG, MPEG, GIF, etc

Computer Networks - Prof Ashok Herur 37

Session layer
• Functions of the Session Layer:
• The layer takes care of the establishment, use, and termination of a
connection.
• It adds checkpoints as synchronization points into the data for identifying the
errors easily and thus avoiding data loss at the end points.
• It also decides whether the 2 systems interact with each other in half-duplex
or full-duplex mode.

Computer Networks - Prof Ashok Herur 38

Transport layer
• It takes complete responsibility for the end-to-end delivery of the message.
• It accepts data and breaks it down into smaller units (called ‘segments’) while
sending. It reassembles the segments on the receiving side.
• It follows a service point (Port) address to deliver the message to the correct
process; Multiple processes could be running in the PC at the same time.
• It takes care of Flow Control and thus prevents a fast sender from swamping a
slow receiver.
• This layer provides the reliable Connection-oriented service (TCP protocol) as
well as the not-so-reliable Connectionless service (UDP protocol).

Computer Networks - Prof Ashok Herur 39

Network layer
• Functions of the Network Layer:
• Routing: The network layer protocols determine the ‘best’ route to take the
packets (segment + network layer header) from source to destination.
• These protocols work between 2 adjacent network devices like routers.
• Congestion control: To ensure a smooth and orderly flow of data across the
network.
• Logical Addressing: The network layer defines and implements an addressing
scheme (IPv4, IPv6) to uniquely identify every single device on the Internet.

Computer Networks - Prof Ashok Herur 40

Data link layer
• It frames the data in a way that is meaningful to the receiver using special bit
patterns.
• It adds physical (MAC) addresses of both sender and receiver in every frame.
• This layer controls error by using error-detection and error-correction codes.
• It controls the data flow over a link.
• When a single communication channel is shared by multiple devices, the MAC
sub-layer of the data link layer helps to determine which device has control
over the channel at a given time.

Computer Networks - Prof Ashok Herur 41

Physical layer
• It converts the bits into appropriate signals (electrical, optical, electromagnetic
waves, etc) to be transmitted on the physical transmission medium.
• It enables bit synchronization using a clock to prevent loss of bits.
• It determines the transmission rate / bit rate (bits sent per second).
• It decides the ideal topology for interconnection of the nodes (Bus, Star, Mesh,
etc)
• The physical layer consists of – Hub, Repeater, Modem, and Cables.

Computer Networks - Prof Ashok Herur 42

OSI Reference model

Computer Networks - Prof Ashok Herur 43

OSI Reference model

Computer Networks - Prof Ashok Herur 44

TCP/IP Reference model
• It defines just 4 layers, unlike the 7 layers of OSI model.
• This model, defined in the 1960’s (much before the OSI model), is widely used
(including on the Internet).
• The top 3 layers of the OSI Reference model are combined into one layer here.
• Similarly, the lowest 2 layers of the OSI model are combined into one layer.
• OSI model tried to segregate the various functions as minutely as possible; But
in practice, there were a few ambiguities and overlaps between the functions
of adjacent layers.

Computer Networks - Prof Ashok Herur 45

TCP/IP Reference model

Computer Networks - Prof Ashok Herur 46

Physical layer - Transmission media

Computer Networks - Prof Ashok Herur 47

Transmission media
• The transmission medium carries the message from the transmitter to the
receiver.
• Common transmission media:
• Twisted wire pair (TWP)
• Coaxial cable (Coax)
• Optical fibre
• Space (Wireless transmission)
• The transmission medium is also referred to as Communication Channel or
Link.

Computer Networks - Prof Ashok Herur 48

Transmission media
• Wire pair
• carries message in the form of an electrical signal.
• Coaxial cable
• carries message in the form of an electrical signal.
• Space
• The message travels through space as an electromagnetic signal.
• Optical fiber
• The optical fiber carries the message as a light signal.
• It has a very high bandwidth and very low error rate.

Computer Networks - Prof Ashok Herur 49

Twisted wire pair

• One of the oldest and least expensive

media.
• Extensively used in the last-mile
connectivity in landline telephone
networks.
• Low bandwidth compared to Coax and
fibre.
• Also used in LANs (to connect PCs to the
Hub).

Computer Networks - Prof Ashok Herur 50

Coaxial cable

• Was widely used on the trunk routes in the telephone system.

• Much higher bandwidth and better noise immunity, compared to wire pairs.
• Also used in Cable TV networks, and as a Bus in the older Ethernet LANs.

Computer Networks - Prof Ashok Herur 51

Distortion, noise and errors on copper lines

• The electric signals sent on copper lines (wire pair and coax cables) suffer
from:
• Power loss (attenuation)
• Distortion of the sharp pulses (representing the 1’s and 0’s)
• Electrical Noise from the environment.
• Distortion in a signal is the change of the shape or some other characteristic
of the waveform. It depends on the characteristics of the media.
• Noise is an external random (unpredictable) signal added to the original
signal.

Computer Networks - Prof Ashok Herur 52

Distortion, noise and errors on copper lines

• All the 3 factors increase with the distance travelled.

Original

Distorted

Noisy

Computer Networks - Prof Ashok Herur 53

Distortion, noise and errors on copper lines

• The first 2 parameters can be dealt with by using Regenerative Repeaters

• Repeaters receive weak and noisy signals and retransmit strong and clean
pulses.
• Transmission errors occur mainly because of impulsive noise (that cannot be
predicted or controlled).

T RR

Computer Networks - Prof Ashok Herur 54

Optical fibre

• It consists of a transparent core and a cladding around the core.

• The refractive index of the cladding is higher than that of the core.
• The electrical signal is converted to a light signal by means of a specialised
LED or Laser, and is then transmitted into the core of the fibre.
• The light signal propagates in the core due to “Total Internal Reflection”, even
around bends.
• They offer a high bandwidth and have a very low Error Rate (10-12 and 10-14)
• The power loss is also very low, thereby needing fewer ‘regenerative
repeaters’ along the way.
Computer Networks - Prof Ashok Herur 55
Optical fibre

Computer Networks - Prof Ashok Herur 56

Wire-line versus Wireless transmission

Computer Networks - Prof Ashok Herur 57

Multiplexing

• It involves combining many signals, arriving on as many connections, on to a

single, larger capacity transmission line.
• It results in a huge cost saving and better utilization of the large bandwidth.
• Common types of multiplexing :
• Time Division Multiplexing (TDM)
• Different signals are transmitted in different time-slots
• Frequency Division Multiplexing (FDM)
• Different signals are transmitted using different frequencies
• Wavelength Division Multiplexing (WDM)

Computer Networks - Prof Ashok Herur 58

Multiplexing

A A

Trunk line
B MUX DEMUX B

C C

Computer Networks - Prof Ashok Herur 59

Time Division Multiplexing (TDM)

Computer Networks - Prof Ashok Herur 60

Space

• Transmitting antennas convert the electrical signal to an electromagnetic

signal, of a particular assigned frequency, and is then radiated into space.
• The receiving antenna would convert the e.m. signal to an electrical one.
• Space (wireless) transmission is used in many applications:
• Radio and TV broadcasts
• Microwave transmission
• Satellite communication
• Cellular (phone) networks
• Navigation (airplanes and ships), ….

Computer Networks - Prof Ashok Herur 61

Modulation

• Modulation is the process of varying one or more properties of a periodic,

high-frequency waveform, called the carrier signal, with a separate signal
called the modulation signal that typically contains information to be
transmitted.
• Modulation is done to make the digital signal suitable for long distance
transmission on the media (both wireline and wireless)
• In the case of wireless transmission, this takes care of sending multiple
messages in the same space, as well as to reduce the power loss.
• In the case of wireline transmission, modulation helps implement FDM.

Computer Networks - Prof Ashok Herur 62

Types of Modulation

• Analog modulation techniques (when the message is Analog):

• Amplitude Modulation (AM)
• Frequency Modulation (FM)
• Phase Modulation (PM)
• Digital modulation techniques (when the message is Digital):
• Amplitude Shift Keying (ASK)
• Frequency Shift Keying (FSK)
• Phase Shift Keying (PSK)

Computer Networks - Prof Ashok Herur 63

Digital Modulation techniques

Computer Networks - Prof Ashok Herur 64

Network topologies

• Topology refers to the way the various devices are connected in a network.
• Common topologies are:
• Bus topology;
• Star topology;
• Ring topology;
• Mesh topology.
• While the first 3 are used in a LAN or MAN, the Mesh topology is used in the
WAN.

Computer Networks - Prof Ashok Herur 65

Bus topology

• Every network device is connected to a

single, bi-directional cable
• It is a broadcast in nature.
• Not a robust topology because if the
backbone (Bus) fails the network
crashes.
• Various MAC (Media Access Control)
protocols are followed to allow the
devices to transmit on the Bus.

Computer Networks - Prof Ashok Herur 66

Ring topology

• It forms a ring connecting devices with

exactly two neighbouring devices.
• A repeater is used at every node.
• The data flows in a fixed direction.
• The failure of a single node in the
network can cause the entire network
to fail.
• Various MAC protocols are followed to
allow the devices to transmit on the
Ring.

Computer Networks - Prof Ashok Herur 67

Star topology

• All the devices are connected to a

single hub through a cable.
• This hub forwards the messages to the
other nodes.
• It can be passive (broadcast) or active
(with repeaters) in nature.
• If one link fails only that link will be
affected. If the Hub fails, the network
crashes.
• Easy to identify and isolate faults.

Computer Networks - Prof Ashok Herur 68

Mesh topology

• Every network device is connected to all

the other devices directly.
• Internet backbone uses mesh topology,
where various ISPs are connected to each
other via dedicated channels.
• Very robust, since an alternate route can
be found in case of a link failure.
• Hence, also used in defence applications.

• Fast communication between the nodes.

Computer Networks - Prof Ashok Herur 69

Computer Networks - Prof Ashok Herur 70