IPG III DC Full Slides

8/3/2019 IPG III DC Full Slides

1/188

DATA

COMMUNICAT

ION

Introduction Components

Protocols

Standards Topology


2/188


3/188

What is Computer Network ?

A computer network allows sharing of resources andinformation among devices connected to the network.

The Advanced Research Projects Agency (ARPA) funded thedesign of the "Advanced Research Projects Agency Network"

(ARPANET) for the United States Department of Defense.

It was the first operational computer network in the world.

Network is categorized as PAN (within the personal network

like cells, PDAs, pagers etc. of one person), LAN (within building/plant/campus), MAN (covering an area size of acity) and WAN (Spanning states/countries/whole world),INTERNET (network of networks).


4/188

COMPONENTS

Message Info./data to be communicated.Eg. Text, Numbers, Pictures, Sound, Video,or combination of these

Sender/Transmitter

Receiver

Medium Path

Protocol Set of rules that governs datacommunication. Agreement between

communicating devices.


5/188

a ta C o m m u n ica tio n S y ste mC o m p o n e n ts


6/188

Data transmission occurs betweentransmitter and receiver over some

transmission medium.

Transmission media is classified as:

1.guided - the waves are guided along aphysical path; examples of guided mediaare twisted pair, coaxial cable, andoptical fiber.

2.unguided - provide a means for

transmitting electromagnetic waves butdo not guide them; eg. propagationthrough air, vacuum, and sea water.

3.

In both cases, communication is in the form


7/188

Network Criteria


8/188

Performance

Depends on:1.Number of users In peak load

periods performance decreases, butnetwork responding to load ismeasure of performance.

2.Type of transmission medium Faster medium like fibre optics

cabling.3.Hardware High speed computer withlarge storage,

4.Software Well defined softwares at

sender, receiver and intermediate


9/188

Reliability

1.Frequency of failure

2.

3.Recovery time of a network afterfailure

4.

5.Catastrophe - fire, earthquake,theft.


10/188

Security

Unauthorized Access Protectionis needed at different levels likeidentification codes and passwords

at lowest levels to encryptiontechniques at higher level.

Viruses Illicitly introduced codethat damages the system.


11/188

PROTOCOLS

Protocol defines:

1.What is to be communicated

2.How is it to be communicated

3.When is to be communicated

4.

Key elements of protocols areSyntax, Semantics and Timing.


12/188

Syntax Structure/format of data meaningorder in which they are presented.

Example, a simple protocol expect first 8 bits

of data to be the address of the sender, thesecond 8 bits to be the address of the receiver,and the rest of the stream to be the messageitself.

Semantics - refers to the meaning of eachsection of bits. How is a particular pattern tobe interpreted, and what action is to be takenbased on that interpretation?

Example, an address identify the route to betaken or the final destination of the message?

Timing - refers to two characteristics: whendata should be sent and how fast they can be

sent.


13/188

STANDARDS

Standards are essential in:1.creating and maintaining an open andcompetitive market for equipmentmanufacturers and in

2.guaranteeing national and internationalinteroperability of data andtelecommunications technology andprocesses.

Standards provide guidelines tomanufacturers, vendors, governmentagencies, and other service providers toensure the kind of interconnectivitynecessary in today's marketplace and in


14/188

Categories of Datacommunication standards


15/188

De facto - by fact. Standards thathave not been approved by an

organized body but have beenadopted as standards throughwidespread use are de factostandards. They are establishedoriginally by manufacturers whoseek to define the functionality of anew product or technology.

De jure - by law or by regulation. Those standards that have beenlegislated by an officiallyrecognized body are de jure


16/188

Standards CreationCommittees

International Organization forStandardization (ISO).

International Telecommunication Union-

Telecommunication Standards Sector(ITU-T).

American National Standards Institute(ANSI).

Institute of Electrical and ElectronicsEngineers (IEEE).

Electronic Industries Association (EIA).

Telcordia


17/188


18/188

Line Configuration

Refers to the way two or morecommunication devices attach to alink.

Two configurations:

1.Point to point

2.Multipoint


19/188

Point to Point

Provides dedicated link between twodevices.

Entire capacity of channel is reservedfor the transmission.

Wires/cables/microwave or satellite

links are used to connect.

Eg. Remote control and TV control

system.


20/188

- -o in t to P o in t Lin e C o n fig u ra tio n


21/188



22/188



23/188


24/188

u ltip o in t Lin e C o n fig u ra tio n


25/188

TOPOLOGY

Refers to the way a network is laidout physically or logically.

It is the geometric representation ofthe relationship of all the links andlinking devices (nodes) to eachother.

Two or more devices connect to alinkand two or more links forms a

topology.


26/188

CATEGORIES OFTOPOLOGY


27/188

MESH TOPOLOGY

Every device has a dedicated point-to-point link to every other device.

dedicated means that the link carriestraffic only between the twodevices it connects.


28/188

e sh To p o lo g y


29/188

Advantages The use of dedicated links guarantees that each

connection can carry its own data load, thuseliminating the traffic problems that can occurwhen links must be shared by multiple devices.

Robust. If one link becomes unusable, it does notincapacitate the entire system.

Privacy or security. When every message travelsalong a dedicated line, only the intended recipient

sees it. Physical boundaries prevent other usersfrom gaining access to messages.

Point-to-point links make fault identification andfault isolation easy. Traffic can be routed to avoid

links with suspected problems. This facility enablesthe network manager to discover the precise


30/188

Limitations

The amount of cabling and

the number of I/O portsrequired. Every device mustbe connected to every otherdevice, installation andreconnection are difficult.

The sheer bulk of the wiring canbe greater than the availablespace (in walls, ceilings, or

floors) can accommodate.

The hardware required toconnect each link (I/O portsand cable) is expensive.

USE

Connection oftelephoneregionaloffices inwhich each

regional officeneeds to beconnected toevery otherregionaloffice.


31/188

Star Topology Each device has a dedicated point-to-point

link only to a central controller, called ahub.

The devices are not directly linked to oneanother.

Unlike a mesh topology, a star topologydoes not allow direct traffic betweendevices.

The controller acts as an exchange: If one

device wants to send data to another, itsends the data to the controller, which


32/188

ta r To p o lo g y


33/188

Advantages Less expensive than a mesh topology.

Each device needs only one link and one I/O port toconnect it to any number of others. This makes iteasy to install and reconfigure.

Less cabling needs to be housed, and additions,moves, and deletions involve only one connection:between that device and the hub.

Robustness. If one link fails, only that link is affected.All other links remain active.

As long as the hub is working, it can be used to

monitor link problems and bypass defective links.Therefore easy fault identification and fault


34/188

Limitations

Dependency of the whole topology on onesingle point, the hub. If the hub goesdown, the whole system is dead.

It requires far less cable than a mesh, each

node must be linked to a central hub. Somore cabling is required in a star.

USE: The star topology is used in local-area

networks (LANs). High-speed LANs oftenuse a star topology with a central hub.


35/188

Tree Topology

Variation of Star topology.

Here a central hub (should alwaysbe active)is the main controller andbelow it are different secondaryhubs (active/passive) controlling

different workstations.

Eg: Cable TV technology


36/188

Bus Topology

It is multipoint. One long cable acts as

a backbone to link all the devices in anetwork.

Nodes are connected to the bus cable by

drop lines and taps.

A drop line is a connection running betweenthe device and the main cable. A tap is a

connector.

As a signal travels along the backbone,some of its energy is transformed into

heat. So it becomes weaker and weaker


37/188

u s To p o lo g y


38/188

Advantages

Ease of installation - Backbone cablecan be laid along the most efficientpath, then connected to the nodes bydrop lines of various lengths.

Uses less cabling than mesh or startopologies.

Example - four network devices in thesame room require four lengths of

cable reaching all the way to the hub.In a bus, this redundancy iseliminated. Only the backbone cablestretches through the entire facility.


39/188

Disadvantages

Difficult reconnection and fault isolation. A bus isdesigned to be optimally efficient at installation,therefore difficult to add new devices.

Signal reflection at the taps can cause degradationin quality. This degradation can be controlled by

limiting the number and spacing of devicesconnected to a given length of cable.

A fault or break in the bus cable stops alltransmission, even between devices on the sameside of the problem.

USE:

Ethernet LANs use a bus topology


40/188

Ring Topology

Each device has a dedicated point-to-pointconnection with only the two devices oneither side of it.

A signal is passed along the ring in onedirection, from device to device, until itreaches its destination.

Each device in the ring has a repeater.When a device receives a signal intendedfor another device, its repeater

regenerates the bits and passes themalon


41/188


42/188

Advantages

Easy to install and reconfigure.Each device is linked to only itsimmediate neighbors (eitherphysically or logically). To add ordelete a device requires changingonly two connections.

Fault isolation is simplified. In aring, a signal is circulating at all times.If one device does not receive a signalwithin a specified period, it can issue

an alarm. The alarm alerts the


43/188

Limitations

A break in the ring (such as adisabled station) can disable theentire network.

USE:

Was prevalent when IBM introducedits local-area network Token Ring.

Today, the need for higher-speedLANs has made this topology less

popular.

A t k b h b id


44/188

yb rid To p o lo g yA network can be hybrid

Transmission mode defines the


45/188

Transmission mode defines thedirection of signal flow between two

linked devices.


46/188

Simplex

Communication is unidirectional one way.

One can transmit only and one can

receive only.

Eg. Keyboards, Monitor etc.


47/188

S im p le x

H lf D l


48/188

Half-Duplex

Each station can both transmit aswell as receive, but not at the sametime (when one device sends otherreceive and vice-versa).

Eg. Walkie-Talkie


49/188

-a lf D u p le x

F ll D l


50/188

Full-Duplex

Also called Duplex.

Both stations can transmit andreceive simultaneously.

Eg. Telephone network


51/188

-u ll D u p le x


52/188


53/188

LAN

Links devices in a singleoffice/building/ campus in span offew kms.

LANs allow resources to be shared

between PCs and workstations. Uses only one type of transmission

medium.

Common LAN topologies are Bus,

Ring and Star.

Have data in range 4 to 16 Mbps, butincreasing to 100 Mbps +.


54/188

o ca l A re a N e tw o rk


55/188

o ca l A re a N e tw o rk


56/188

MAN

Designed to extend over entire city.

Can be single network or aconnection of LANs.

Eg. An organization use MAN toconnect LANs at their variousoffices located in the city.


57/188

e tro p o lita n A re a N e tw o rk


58/188

WAN

Provides long distance transmissionof data, voice, image, video overlarge geographical areas.

Wholly owned and used by a singlecompany and so called called

Enterprise Network.

id e A re a N e tw o rk


59/188


60/188

INTERNETWORKS or Internet

Individual networks are joined bydevices like routers and gateways.


61/188

n te rn e tw o rk( )nternet


62/188

SIGNALS

A signal is any time-varying or spatial-varying quantity.

Analog (continuous) and digital(discrete)

Periodic signals (completes a pattern

within a measurable time frame,called a period and repeats thepattern over identical intervals.Completeness of one full pattern is

called a cycle) and Aperiodic


63/188

ransformation of Informationo Signals


64/188

Frequency is No. of cycles per second. F = 1/T

Bandwidth is the range of frequencies the signaloccupies. B = fH fL

Amplitude is the instantaneous height of thesignal

Phase is the shift of wave along the time axis.

Bit rate (bits/sec) and Bit Interval (duration ofone bit) are used to describe a digital signal.


65/188

Communications tasks Transmission system utilization Interfacing Signal generation Synchronization Exchange management Security Error detection and correction Flow control Addressing Routing

Recovery Message formatting Network management

Transmission system


66/188

yutilization

The need to make efficient use oftransmission facilities shared amongof communicating devices.

Techniques (multiplexing) are used toallocate the total capacity of atransmission medium among anumber of users.

Congestion control techniques to assurethat the system is not overwhelmedby excessive demand for transmissionservices.

To communicate a device must interface with the


67/188

To communicate, a device must interface with thetransmission system.

Once an interface is established, signal generation isrequired for communication.

The properties of the signal, such as form and

intensity, must be such that they are: -(1) capable of being propagated through the

transmission system, and(2) interpretable as data at the receiver.

The signals be generated to conform to therequirements of the transmission system andreceiver and there must be some form ofsynchronization between transmitter and receiver.

The receiver must be able to determine when a signal

h i f i i


68/188

There are requirements for communicationbetween two parties under the termexchange management.

If data are to be exchanged in both

directions over a period of time, the twoparties must cooperate.

For data processing devices, more will be

needed than simply establishing aconnection. Like: -

1.whether both devices may transmitsimultaneously or must take turns,

2.the amount of data to be sent at one time,

3.the format of the data, and

4.what to do if certain contingencies, such as


69/188

In all communications systems, there is a potential forerror; transmitted signals are distorted to someextent before reaching their destination.

Error detection and correction are required where

errors cannot be tolerated usually with dataprocessing systems.

Flow control is required to assure that the source

does not overwhelm the destination by sendingdata faster than they can be processed andabsorbed.

Addressing -The transmission system must assure

that the destination system, and only that system,receives the data.

Routing - transmission system may itself be a


70/188

Recovery is a concept distinct from that of errorcorrection by Recovery techniques. The objective iseither to be able to resume activity at the point ofinterruption or at least to restore the state of thesystems involved to the condition prior to thebeginning of the exchange.

Message formatting has to do with an agreement

between two parties as to the form of the data to beexchanged or transmitted. Example, both sidesmust use the same binary code for characters.

Data communications facility is a complex system that

cannot create or run itself. Network managementcapabilities are needed to: -

1. configure the system,

2. monitor its status,

3. react to failures and overloads, and

4. plan intelligently for future growth.

Key Elements of Internet


71/188

Key Elements of Internet Hosts: End systems like PCs, Workstations, servers, mainframes

etc.

Routers: Connects the networks. Each router attaches to two ormore networks.

Hosts breaks the data to be sent into a sequence of packets calledIP Datagram/ packets .

Each packet includes a numeric address of destination host calledIP Address (address is carried in an IP packet).

Hosts are sometimes grouped together in a LAN. Individual hostsand LANs are connected to an ISP (Internet Service

Provider) through a POP (Point of Presence) which is afacility where customer can connect to ISP network.

Connection is made in a series of steps starting with theCustomer Premises Equipments (CPE). CPE is thecommunication equipment located onsite within the host, Eg.

Modem. Network Access Point (NAP) - Facility that provide

infrastructure to move data between connected networks. It isone of the several major internet interconnection points thatserve to tie several ISPs together.

NAPs are owned by NSPs (Network Service Providers). An NSPcan also be ISP but not always.

Frame: group of bits containing data, addresses & control info.

Flow Control and


72/188

Flow Control andCongestion

When sender is sending at higher speedthan receiver can receive.

Occurs in error free as well as error pronetransmission.

Flow control throttles the sender in

sending no faster than receiver canhandle. Requires feedback mechanism to make

aware sender if receiver is keeping upwith it or not.

Flow control prevents congestion. Flow control is avoiding sending data

faster. Congestion control is avoiding congestion

within the subnet.


73/188

Congestion occurs when users ofnetwork collectively demand moreresources than network has tooffer.

Congestion can be solved by goodrouting and buffer management.

Controlling the flow of packets in thenetwork is flow control.

NEED FOR PROTOCOL


74/188

ARCHITECTURE When computers, terminals, and/or other dataprocessing devices exchange data, there

must be a data path between the two eitherdirectly or via a communication network.

Typical tasks to be performed are: -

1.The source system must either activate thedirect data communication path or informthe communication network of the identityof the desired destination system.

2.The source system must ascertain that the

destination system is prepared to receivedata.

3.The file transfer application on the sourcesystem must ascertain that the filemanagement program on the destinationsystem is prepared to accept and store thefile for this particular user.

Architecture for file


75/188

Architecture for filetransfer


76/188

Communications involve threeagents: applications, computers,

and networks.

Communication task is divided into

three relatively independent layers:

1.Network access layer

2.Transport layer

3.Application layer

NETWORK ACCESS LAYER


77/188

NETWORK ACCESS LAYER Concerned with the exchange of data between a

computer and the network to which it is attached. The sending computer provide the network with the

address of the destination computer.

The sending computer may invoke certain services,such as priority, that might be provided by thenetwork.

The specific software used depends on1. the type of network to be used;

2. different standards have been developed for circuitswitching, packet switching, local area networks,and others.

TRANSPORT LAYER


78/188

TRANSPORT LAYER

All of the data arrive at the destinationapplication and in the same order inwhich they were sent.

Mechanisms for providing reliabilityindependent of the nature of the

applications.

APPLICATION LAYER

Contains the logic needed to supportthe various user applications like filetransfer, a separate module is neededthat is peculiar to that application.

Example


79/188

Example Computer A, wishes to send a message to another

application, associated with computer B.

The application at A hands the message over to its

transport layer with instructions to send it to SAP oncomputer B.

The transport layer hands the message over to the

network access layer, which instructs the networkto send the message to computer B.

Network need not be told the identity of the

destination service access point. All that it needs toknow is that the data are intended for computer B.

To control this operation, control information, as well

Th di li ti t bl k f d t d


80/188

The sending application generates a block of data andpasses this to the transport layer.

The transport layer break this block into two smaller piecesto make it more manageable.

To each of these pieces the transport layer appends a

transport header, containing protocol control information.

The combination of data from the next higher layer and

control information is known as a protocol data unit(PDU).

The header in each transport PDU contains controlinformation to be used by the peer transport protocol atcomputer B.

Header include:

Next the transport layer hand each protocol


81/188

Next the transport layer hand each protocoldata unit over to the network layer, withinstructions to transmit it to the

destination computer.

Network access protocol must present thedata to the network with a request for

transmission.

Network access protocol appends a networkaccess header to the data it receives fromthe transport layer, creating a networkaccess PDU.

Header includes:

1.Destination computer address.

TCP/IP Protocol


82/188

TCP/IP ProtocolArchitecture

TCPI-IP protocol suit is a result of protocol research anddevelopment conducted on the experimental packet-switched network, ARPANET (Advanced Research ProjectsAgency Network), funded by the Defense AdvancedResearch Projects Agency (DARPA).

It consists of a large collection of protocols that have been

issued as Internet standards by the Internet ArchitectureBoard (IAB).

Communication task for TCPIIP into five relativelyindependent layers:

1. Application layer

2. Host-to-host, or transport layer

3. Internet layer

4. Network access layer

PHYSICAL LAYER


83/188

PHYSICAL LAYER

It covers the physical interface betweena data transmission device (e.g.,workstation, computer) and a

transmission medium or network.

This layer is concerned with specifying:-

1.the characteristics of the transmissionmedium,

2.the nature of the signals,

NETWORK l


84/188

NETWORK access layer It is concerned with the exchange of

data between an end system and thenetwork to which it is attached.

Provide the network with the address of thedestination computer, invoke certainservices, such as priority.

Specific software used at this layer dependson the type of network to be used;

different standards have been developedfor circuit-switching, packet-switching(e.g., X.25), local area networks (e.g.,

Ethernet), and others.

INTERNET LAYER


85/188

INTERNET LAYER

In cases where two devices are attached to differentnetworks, procedures are needed to allow data totraverse multiple interconnected networks. This isthe function of the internet layer.

The internet protocol (IP) is used at this layer toprovide the routing function across multiplenetworks.

This protocol is implemented in the end systems andin routers.

A router is a processor that connects two networks

and whose primary function is to relay data from

one network to the other on its route from thesource to the destination end s stem.

HOST-TO-HOST LAYER/


86/188

HOST TO HOST LAYER/TRANSPORT LAYER

Data be exchanged reliably and

All of the data arrive at thedestination application and

Data arrive in the same order inwhich they were sent.

APPLICATION LAYER


87/188

APPLICATION LAYER

It contains the logic needed tosupport the various userapplications.

Eg. file transfer, a separate module isneeded that is peculiar to that

application.

TCP/IP Protocol


88/188

/architecture model

TCP/IP Protocol


89/188

/Architecture

TCPI-IP protocol suit is a result of protocol research anddevelopment conducted on the experimental packet-switched network, ARPANET (Advanced Research ProjectsAgency Network), funded by the Defense AdvancedResearch Projects Agency (DARPA).

It consists of a large collection of protocols that have been

issued as Internet standards by the Internet ArchitectureBoard (IAB).

Communication task for TCPIIP into five relativelyindependent layers:

1. Application layer

2. Host-to-host, or transport layer

3. Internet layer

4. Network access layer

PHYSICAL LAYER


90/188

PHYSICAL LAYER

It covers the physical interface betweena data transmission device (e.g.,workstation, computer) and a

transmission medium or network.

This layer is concerned with specifying:-

1.the characteristics of the transmissionmedium,

2.the nature of the signals,

NETWORK access layer


91/188

NETWORK access layer It is concerned with the exchange of

data between an end system and thenetwork to which it is attached.

Provide the network with the address of thedestination computer, invoke certainservices, such as priority.

Specific software used at this layer dependson the type of network to be used;

different standards have been developedfor circuit-switching, packet-switching(e.g., X.25), local area networks (e.g.,

Ethernet), and others.

INTERNET LAYER


92/188

INTERNET LAYER

In cases where two devices are attached to differentnetworks, procedures are needed to allow data totraverse multiple interconnected networks. This isthe function of the internet layer.

The internet protocol (IP) is used at this layer toprovide the routing function across multiplenetworks.

This protocol is implemented in the end systems and

in routers.

A router is a processor that connects two networks

and whose primary function is to relay data from

one network to the other on its route from thesource to the destination end s stem.

HOST-TO-HOST LAYER/


93/188

TRANSPORT LAYER

Data be exchanged reliably and

All of the data arrive at thedestination application and

Data arrive in the same order inwhich they were sent.

APPLICATION LAYER


94/188

APPLICATION LAYER

It contains the logic needed tosupport the various userapplications.

Eg. file transfer, a separate module isneeded that is peculiar to that

application.

TCP/IP Protocol


95/188

architecture model

OSI (Open Systems


96/188

p yInterconnection ) Model

Developed by the International Organization forStandardization (ISO) as a model for acomputer communications architecture andas a framework for developing protocolstandards.

It consists of seven layers:1.Application

2.Presentation

3.Session

4.Transport5.Network

6.Data Link

7.Physical

Seven layers of the OSI


97/188

ymodel

Interaction between layers in the OSImodel


98/188

model

Peer-to-Peer Processes


99/188

Device A sends a stream of bits to device B (throughintermediate nodes).

At the higher layers, communication move down

through the layers on device A, over to device B,and then back up through the layers.

Each layer in the sending device adds its owninformation to the message it receives from thelayer just above it and passes the whole package tothe layer just below it.

At the receiving machine, the message is unwrapped

layer by layer, with each process receiving andremoving the data meant for it.

Exam le, la er 2 removes the data meant for it, then


100/188

Organization of the Layers


101/188

g y The seven layers belongs to three subgroups.

NETWORK SUPPORT LAYERS:

Layers I, 2, and 3-physical, data link, and network-are thenetwork support layers.

Deal with the physical aspects of moving data from one device toanother (such as electrical specifications, physical connections,

physical addressing, and transport timing and reliability).

USER SUPPORT LAYERS

Layers 5, 6, and 7-session, presentation, and application are theuser support layers;

TRANSPORT LAYER

Layer 4, the transport layer, links the two subgroups and ensuresthat what the lower layers have transmitted is in a form thatthe upper layers can use.

The upper OSI layers are almost always implemented in software;lower la ers are a combination of hardware and software

OSI MODEL


102/188

OSI MODEL

A packet (header and data) at level 7 isl t d i k t t l l 6 Th


103/188

encapsulated in a packet at level 6. Thewhole packet at level 6 is encapsulated ina packet at level 5, and so on.

Data portion of a packet at level N - 1carries the whole packet (data and headerand maybe trailer) from level N.

The concept is called encapsulation; level N- 1 is not aware of which part of theencapsulated packet is data and which

part is the header or trailer.

For level N - 1, the whole packet comingfrom level N is treated as one integral

unit.

PHYSICAL LAYER


104/188

PHYSICAL LAYER

Coordinates the functions required tocarry a bit stream over a physicalmedium.

Deals with the mechanical andelectrical specifications of theinterface and transmission

medium. Defines the procedures and functions

that physical devices and interfaceshave to perform for transmission to

Movements ofindividual bits from one hop (node) to


105/188

individual bits from one hop (node) tothe next.


106/188

Physical layer defines the characteristics ofthe interface between the devices and the

transmission medium.

Representation of bits. The physical layerdata consists of a stream of bits

(sequence of Os or 1s) with nointerpretation. Bits must be encoded intosignals--electrical or optical. The physicallayer defines the type of encoding (howOs and I s are changed to signals).

Data rate.The transmission rate-thenumber of bits sent each second-is alsodefined by the physical layer. The physicallayer defines the duration of a bit, whichis how lon it lasts.

Synchronization of bits The sender


107/188

Synchronization of bits.The senderand receiver not only must use the

same bit rate but also must besynchronized at the bit level. Thesender and the receiver clocks mustbe synchronized.

Line configuration.The physical layeris concerned with the connection ofdevices to the media. In a point-to-point configuration, two devices areconnected through a dedicated link. In

a multipoint configuration a link is

Physical topology The physical


108/188

Physical topology. The physicaltopology defines how devices are

connected to make a network.Devices can be connected by usinga mesh topology, a star topology, aring topology , a bus topology or a

hybrid topology.

Transmission mode.The physical

layer also defines the direction oftransmission between two devices:simplex, half-duplex, or full-duplex.

Data Link Layer


109/188

The data linklayertransformsthe

physicallayer, arawtransmissi

on facility,to areliablelink.

Framing. The data link layer divides the stream ofbits received from the network layer into manageable


110/188

y gdata units called frames.

Physical addressing. If frames are to be distributed todifferent systems on the network, the data link layeradds a header to the frame to define the sender and/orreceiver of the frame. If the frame is intended for asystem outside the senders network, the receiveraddress is the address of the device that connects thenetwork to the next one.

Flow control.

Error control.The data link layer adds reliability to thephysical layer by adding mechanisms to detect andretransmit damaged or lost frames. It also uses amechanism to recognize duplicate frames. Error controlis normally achieved through a trailer added to the endof the frame.

Network LayerTh k l i ibl f h


111/188

The network layer is responsible for the source-to-destination delivery of a packet across multiple networks(links).

Whereas the data link layer oversees the delivery of thepacket between two systems on the same network(links), the network layer ensures that each packet getsfrom its point of origin to its final destination.

If two systems are connected to the same link, there is noneed for a network layer.

If the two systems are attached to different networks (links)with connecting devices between the networks (links),there is a need for the network layer to accomplishsource-to-destination delivery.

RESPONSIBILITIES OF THE NETWORK


112/188

LAYER Logical addressing.The physical addressing

implemented by the data link layer handles theaddressing problem locally. If a packet passes thenetwork boundary, logical addressing system helpdistinguish the source and destination systems.

The network layer adds a header to the packet coming fromthe upper layer that includes the logical addresses of thesender and receiver.

Routing. When independent networks or links are

connected to create intemetworks (network of networks)or a large network, the connecting devices (calledrouters or switches) route or switch the packets to theirfinal destination.

Transport Layer


113/188

Transport Layer Responsible for process-to-process delivery of the

entire message. A process is an applicationprogram running on a host.

Network layer oversees source-to-destination

delivery of individual packets, it does notrecognize any relationship between thosepackets. It treats each one independently, aseach belonged to a separate message,whether or not it does.

The transport layer, ensures that the wholemessage arrives intact and in order, overseeingboth error control and flow control at the

source-to-destination level.

Transport layer is responsible for thedelivery of a message from one


114/188

delivery of a message from oneprocess to another.

Responsibilities of the transport layer


115/188

Service-point addressing. Computers run severalprograms at the same time. So the source-to-destination

delivery means delivery not only from one computer tothe next but also from a specific process (runningprogram) on one computer to a specific process (runningprogram) on the other.

The network layer gets each packet to the correctcomputer; the transport layer gets the entire message tothe correct process on that computer.

Segmentation and reassembly. A message is dividedinto transmittable segments, with each segmentcontaining a sequence number. These numbers enablethe transport layer to reassemble the message correctlyupon arriving at the destination and to identify andreplace packets that were lost in transmission.

Connection control.The transport layer can be eitherconnectionless or connection-oriented


116/188

connectionless or connection oriented.

A connectionless transport layer treats each segment as anindependent packet and delivers it to the transport layer

at the destination machine.

A connection-oriented transport layer makes a connectionwith the transport layer at the destination machine firstbefore delivering the packets. After all the data aretransferred, the connection is terminated.

Flow control. Like the data link layer, the transport layer isresponsible for flow control.

Error control. Like the data link layer, the transport layeris responsible for error control.

The sending transport layer makes sure that the entiremessage arrives at the receiving transport layer withouterror (damage, loss, or duplication). Error correction isusuall achieved throu h retransmission.


117/188

Session Layer

The session layer is the network dialogcontroller.

It establishes, maintains, and synchronizesthe interaction among communicatingsystems.


118/188

Dialog control.The session layer allows

two systems to enter into a dialog. It allows the communication between two

processes to take place in either half-duplex (one way at a time) or full-duplex

(two ways at a time) mode.

Synchronization. The session layer

allows a process to add checkpoints, orsynchronization points, to a stream ofdata.

Presentation Layer


119/188

The presentation layer is concerned

with the syntax and semantics ofthe information exchangedbetween two systems.

Responsible for translation,compression, and encryption.

Translation.The processes (running programs)in two systems are exchanging information in


120/188

in two systems are exchanging information inthe form of character strings, numbers, and soon.

The information must be changed to bit streamsbefore being transmitted.

Different computers use different encodingsystems, the presentation layer is responsiblefor interoperability between these differentencoding methods.

The presentation layer at the sender changes theinformation from its sender-dependent formatinto a common format. The presentation layerat the receiving machine changes the common

format into its receiver-dependent format.

Encryption.To carry sensitive information, a systemmust be able to ensure privacy.


121/188

p y

Encryption means that the sender transforms theoriginal information to another form and sends theresulting message out over the network.

Decryption reverses the original process to transform

the message back to its original form.

Compression. Data compression reduces the numberof bits contained in the information.

Data compression becomes important in thetransmission of multimedia such as text, audio, andvideo.

Application Layer


122/188

Application Layer

The application layer enables the user, whetherhuman or software, to access the network.

It provides user interfaces and support for

services such as electronic mail, remote fileaccess and transfer, shared databasemanagement, and other types of distributedinformation services.

Responsible for providing services to the user

Services provided by theli ti l


123/188

application layer Network virtual terminal. A network virtual

terminal is a software version of a physicalterminal, and it allows a user to log on to aremote host.

The user's computer talks to the softwareterminal which, in turn, talks to the host, andvice versa.

The remote host believes it is communicatingwith one of its own terminals and allows theuser to log on.

File transfer, access, and management.This


124/188

, , gapplication allows a user

1. to access files in a remote host,2. to retrieve files from a remote computer for use inthe local computer, and

3. to manage or control files in a remote computerlocally.

Mail services.This application provides thebasis for e-mail forwarding and storage.

Directory services.This application providesdistributed database sources and access forglobal information about various objects andservices.

Summary of OSI layers functions


125/188

Layers of the TCPIIP and OSI architectures


126/188


127/188

Advantages of standards

1. Assures that there will be a large market for a particular piece of

Equipment or software.2. Encourages mass production and, the use of large-scale-

integration (LSI) or very-large-scale-integration (VLSI)Techniques, resulting in lower costs.

3. A standard allows products from multiple vendors to

communicate, giving the Purchaser more flexibility inequipment selection and use.

Disadvantages :

A standard tends to freeze the technology. By the time a standardis developed, Subjected to review and compromise, and

promulgated, more efficient Techniques are possible. There are multiple standards for the same thing which gives rise

to multiple conflicting of Standards.

Service Primitives


128/188

Services between adjacent layers inOSI model are expressed inprimitives and parameters.

Primitive specifies function to beperformed

Parameters are used to pass dataand control information.

Service Primitives Types Request: issued by service user to invoke some service


129/188

Request: issued by service user to invoke some serviceand pass parameters to specify fully the requestedservice.

Indication: Issued by service provider either to 1. Indicate that a procedure has been invoked by

the peer service user on the connection and to provide the

associated parameters, or 2. notify the service user of a provider-initiated

action

Response: Issued by service user to acknowledge or

complete some procedure previously invoked by anindication of that user.

Confirm: Issued by service provider to acknowledge orcomplete some procedure previously invoked by a

request by the service user.

SWITCHING NETWORK Switching nodes provide facility that will move data


130/188

Switching nodes provide facility that will move datafrom node to node until they reach destination.

They are connected in some topology. End devices wishing to communicate are called

stations.

Collection of nodes is called communication network.

Some nodes have sole task of connecting to othernodes (internal switching of data).

Some nodes have one or more stations attached.

Network is not fully connected usually.

To enhance reliability it is desirable to have more than

one possible path. Switched networks are classified as:

1. Circuit Switched

2. Packet switched.


131/188

CIRCUIT

SWITCHING

CIRCUIT SWITCHINGD di t d i ti th b t t


132/188

Dedicated communication path between twostations.

Path is a connected sequence of links betweennetwork nodes.

On each physical link, a logical channel isdedicated to the connection.

Communication involves three phases:

1. Circuit establishment

2.Data transfer

3. Circuit disconnect

Examples : Public Telephone Network, PrivateBranch Exchange (PBX)

Circuit Establishment Before any signals can be transmitted, an end-to-end

(station-to-station) circuit must be established.


133/188

(station to station) circuit must be established.

Example: station A sends a request to node 4 requesting aconnection to station E.

The link from A to 4 is a dedicated line, so that part of theconnection already exists.

Node 4 must find the next leg in a route leading to node 6.Based on routing information and measures ofavailability node 4 selects the link to node 5, allocates a

free channel on that link and sends a messagerequesting connection to E.

Dedicated path has been established from A through 4 to 5.

Number of stations may attach to 4, it must be able toestablish internal paths from multiple stations to multiplenodes.

Node 5 dedicates a channel to node 6 and internally tiesthat channel to the channel from node 4.

Node 6 completes the connection to E. In completing theconnection, a test is made to determine if E is busy or isprepared to accept the connection.

Data transfer

Information can now be transmitted from A through the


134/188

gnetwork to E.

The data may be analog or digital, depending on the nature

of the network.

As the carriers evolve to fully integrated digital networks,the use of digital (binary) transmission for both voice anddata is becoming the dominant method.

The path is A-4 link, internal switching through 4,4-5

channel, internal switching through 5, 5-6 channel, andinternal switching through 6, 6 -E link. Generally, theconnection is full-duplex (FDX).

Circuit disconnect

After some period of data transfer, the connection isterminated by the action of one of the two stations.

Signals must be propagated to nodes 4, 5, and 6 todeallocate the dedicated resources.

LIMITATIONS


135/188

Connection path is established before data transmissionbegins. Thus,

1. channel capacity must be reserved between each pair ofnodes in the path, and

2. each node must have available internal switching capacityto handle the requested connection.

The switches must have the intelligence to make theseallocations and to devise a route through the network.

Channel capacity is dedicated for the duration of aconnection, even if no data are being transferred.

For a voice connection, utilization may be rather high, but itstill does not approach 100 percent.

For a terminal-to-computer connection, the capacity may beidle during most of the time of the connection.

Performance: there is a delay priort i l t f f ll


136/188

to signal transfer for call

establishment. But once the circuitis established, the network iseffectively transparent to the users.

Information is transmitted at a fixeddata rate with no delay other thanthat required for propagationthrough the transmission links. Thedelay at each node is negligible.

TechnologyLimitations:


137/188

1.Getting dedicated path itself is a

problem.2.For end to end connection all

intermediate nodes and links must

be free.3.When any node/link is unavailable

request gets a busy response.

4. EconomicLimitations:Providing dedicated path is a costly

affair

Blocking


138/188

oc g Blocking occurs when network is

unable to connect two stations asall possible paths are already inuse.

Switches provide no intermediatestorage capacity.

So blockage is possible on a circuit

switch.

Limited value addedf ti


139/188

function Line protocols which can regulate the

traffic (flow control) are notavailable.

Additional softwares and microcodemust be added to provide these

value added functions.

Constant data rate


140/188

Connection provides for transmissionat a constant data rate.

Each of the two devices must

transmit and receive at same andconstant data rate.

Thus limits utility of network in

interconnecting a variety of hostcomputers and terminals.

Reordering not possible


141/188

When two or more packets aretransmitted, there is possibility thatthey reach destination in wrong

order.

Reordering of packets

Examples Circuit switching was developed to handle voice traffic


142/188

Circuit switching was developed to handle voice trafficbut is now also used for data traffic.

Public telephone networkis actually a collection ofnational networks interconnected to form theinternational service.

Originally designed and implemented to serviceanalog telephone subscribers, the network nowhandles substantial data traffic via modem and isgradually being converted to a digital network.

Private branch exchange (PBX), used tointerconnect telephones within a building or office.Circuit switching is also used in private networks-corporations or other large organizationsinterconnecting their various sites; these usuallyconsist of PBX system.

Public telephone network Architectural components of a public telecommunications


143/188

network: -

1. Subscribers:The devices that attach to the network. Mostsubscriber devices to public telecommunicationsnetworks are telephones, but the percentage of datatraffic increases year by year.

2.

3. Local loop: It is the link between the subscriber and the

network, also called the subscriber loop. Almost all localloop connections used twisted-pair wire. The length of alocal loop is typically in a range from a few kilometersto a few tens of kilometers.

4.

5. Exchanges:The switching centers in the network. Aswitching center that directly supports subscribers isknown as an end office. An end office will support manythousands of subscribers in a localized area.

6.

Key requirements for voice traffic is that there must bevirtually no transmission delay and certainly novariation in delay


144/188

variation in delay.

A constant signal transmission rate must bemaintained, as transmission and reception occur atthe same signal rate.

These requirements are necessary to allow normalhuman conversation.

Also the quality of the received signal must besufficiently high to provide, at a minimum,

intelligibility. Circuit-switching technology has been driven by those

applications that handle voice traffic.

Public circuit-switchingnetwork.


145/188

Subscribers connectdirectly to an endoffice, whichswitches trafficb


146/188

betweensubscribers andbetween asubscriber andother exchanges.

The other exchangesare responsible

for routing andswitching trafficbetween endoffices.

To connect two

subscribersattached to thesame end office,a circuit is set upbetween.

If two subscribers

Digital Switch

Provide a transparent (thatthere is a directconnection between

Elements of a circuit-switch node


147/188

connection betweenattached pair of devices)signal path between any

pair of attached devices.

The connection must allowfull-duplex transmission.

The network-interfaceelement represents thefunctions and hardwareneeded to connect digitaldevices, such as dataprocessing devices anddigital telephones, to the

network.

Analog telephones can also

be attached if thenetwork interfacecontains the logic for

converting to digital

Functions of Control Unit


148/188

Establishes connections: done on demand-

that is, at the request of an attached device.To establish the connection, the control unitmust handle and acknowledge the request,determine if the intended destination is free,and construct a path through the switch.

Maintain the connection: The digital switch

uses time-division principles, this may requireongoing manipulation of the switchingelements. The bits of the communication are

transferred transparently (from the point ofview of the attached devices).

Tear down the connection: either in response

to a request from one of the parties or for its

Circuit-switching device is blocking or non-blocking?


149/188

Blocking occurs when the network is unable to connect twostations because all possible paths between them are

already in use.

A blocking network is one in which such blocking is

possible. A non-blocking network permits all stations tobe connected (in pairs) at once and grants all possible

connection requests as long as the called party is free. When a network is supporting only voice traffic, a blocking

configuration is generally acceptable, as it is expectedthat most phone calls are of short duration and thattherefore only a fraction of the telephones will beengaged at any time.

When data processing devices are involved, these

assumptions may be invalid. For example, for a data-entry application, a terminal may be continuously

connected to a com uter for hours at a time

Switching techniques internal to asingle

circuit-switching node: SPACE-


150/188

DIVISION SWITCHING Space-division switching was originally developed

for the analog environment and has been carriedover into the digital.

The fundamental principles are the same,

A space-division switch is one in which the signal

paths are physically separate from one another(divided in space).

Each connection requires the establishment of a

physical path through the switch that is dedicatedsolely to the transfer of signals between the twoendpoints.

Space-division switch


151/188

A simple crossbar matrix

with 10 full-duplex I/Olines.

The matrix has 10 inputs

and 10 outputs; each

station attaches to thematrix via one inputand one output line.

Interconnection is

possible between anytwo lines by enablingthe appropriatecrosspoint.

A total of 100

Limitations of crossbarswitch


152/188

switch

The number of crosspoints grows withthe square of the number of attachedstations. This is costly for a largeswitch.

The loss of a crosspoint preventsconnection between the two deviceswhose lines intersect at thatcrosspoint.

The crosspoints are inefficientlyutilized; even when all of theattached devices are active, only a

Multiple-stage Switches


153/188

Three-stage space-division switch

ADVANTAGES


154/188

The number of crosspoints isreduced, increasing crossbarutilization. The total number of

crosspoints for 10 stations isreduced from 100 to 48.

There is more than one path throughthe network to connect twoendpoints, increasing reliability.

LimitationsR i l t l


155/188

Requires a more complex control

scheme.

To establish a path in a single-stagenetwork, it is only necessary to enable

a single gate. In a multistage network,a free path through the stages mustbe determined and the appropriategates enabled.

It may be blocking. But a single-stagecrossbar matrix is nonblocking; thatis, a path is always available toconnect an input to an output.

Time-division Switching All modern circuit switches use digital time-division

techniques for establishing and maintaining


156/188

techniques for establishing and maintaining"circuits."

Time-division switching involves the partitioning of alower-speed bit stream into pieces that share ahigher-speed stream with other bit streams.

The individual pieces, or slots, are manipulated bycontrol logic to route data from input to output.

TDM bus switching, and all digital switchingtechniques, are based on the use of synchronous

time-division multiplexing (TDM).

Synchronous TDM permits multiple low-speed bitstreams to share a high-speed line. A set of inputs issampled in turn. The samples are organized serially

into slots (channels) to form a recurring frame ofslots, with the number of slots per frame equal tothe number of inputs. A slot may be a bit, a byte, orsome longer block.

In synchronous TDM, the source and destination of thedata in each time slot are known. Hence, there is no

need for address bits in each slot

Each device attaches to the switchthrough a full-duplex line.

These lines are connected throughcontrolled gates to a high-speed


157/188

g g pdigital bus.

Each line is assigned a time slot forproviding input. For the durationof the slot, that line's gate isenabled, allowing a small burst ofdata onto the bus. For that sametime slot, one of the other linegates is enabled for output. Thus,

during that time slot, data areswitched from the enabled inputline to the enabled output line.

During successive time slots,different input/output pairings areenabled, allowing a number ofconnections to be carried over

the shared bus. An attached device achieves full-

duplex operation by transmittingduring one assigned time slot andreceiving during another.

Advantages of CircuitSwitching

Transparency: Sender and Receiver can use any


158/188

Transparency: Sender and Receiver can use anybit rate, format or framing method they

want to. Carrier/trunks does not know orcare. This transparency allows voice, dataand fax to co-exist in one telephone system.

Charging is based on distance and time and noton traffic.

Reliability increases by the use of Multistageswitches.

TDM Bus switching can accommodate lines ofvarying data rates.

Once a circuit is established, it appears asdirect connection of two attached station, nospecial networking logic is needed at either


159/188

special networking logic is needed at eitherpoints.

Delivery is guaranteed and sequential.

Delays are small and constant.

Communication takes place in real time i.e. canhandle real time traffic.

Is not store and forward type of switching.

(Switches provide no intermediate storagecapacity.)


160/188

MESSAGE

SWITCHING

MESSAGE SWITCHING Based on store and forward technology.


161/188

gy

No dedicated physical path is established in advancebetween sender and receiver.

When a node has a message block to send it is stored in thefirst switching office and then forwarded later to nextnode.

Message hops from node to node.

At each hop the entire message:1. Is received

2. Is inspected for errors

3. Temporarily stored in the secondary memory until a nextnode is available.

Example User A sends a message to user B.


162/188

User A sends a message to user B. Links 1 and 2 are unavailable (either down or in

use) for transmission. Message must be retransmitted through over

links 3 and 4.

Working:1.Message is transmitted from user A to the DTE

(Data Terminal Equipment) at node 1 whereit is stored.

2.When link 3 is free, message is read fromstorage at node 1 and is sent to node 3,where is stored.

3.When link 4 is free, message is read fromstorage at node 3 and sent to node 2.

4 Finally message is read from storage at node 2

Message Switch


163/188

Is a specialized computer.

Functions: -Accepts traffic from terminals and

computers attached to it through dialup/leased telephone lines.

Examines address in the header of themessage and switches (routes) thetraffic to receiving DTE.

Can store the information due tostorage capability of switch.

Do not support interactive (real time)


164/188

pptraffic. But message can be sent athigh speed by establishing priorityfor different types of traffic.

High priority traffic is queued forshorter periods. This supportsinteractive (real time) traffic.

Use tape files to perform back up of disk files. Operates with Master-slave Relationship.

Switches perform Polling and selection to


165/188

Switches perform Polling and selection tomanage traffic coming into and going out.

Example: User DTE A has data for user DTEB.

1)Switches perform Polling and upon polling A

2)Message is transmitted to the switch.3)Based on priority data are stored in one of

several disk queues.

4)Depending on overall traffic condition and

message priority, switch dequeues themessage (message goes out of waitingqueue)

5)Sends select command to C.

6)At site C an acknowledgement is sent to switch.

Advantages Queuing onto the disk provides a method of smooth traffic


166/188

Queuing onto the disk provides a method of smooth traffic(by queuing a low priority traffic during peak traffic

periods).

No real end to end path is created. This makes good use ofcommunication channels as no resources need to bereserved until the channels are available.

Overcomes limitations of circuit switching in which linkefficiency was reduced due to non-storage capacity. Inmessage switching efficiency is increased as less numberof switches can serve more number of users.

Supports interactive applications by using priority in

queuing. Queuing method also smoothens traffic (flowcontrol) as we can queue up low priority traffic at peaktraffic period.

Limitations


167/188

It is master-slave structure so if switches fails,entire network can be lost. It is only the

switch where all network goes in and out.Note- To overcome this limitations many

organizations provide a duplicate message

switch, which can resume working if firstone fails.

Due to arrangement of many message switchesthere is degradation of response time and

decreased throughput.

Does not utilize communication lines efficientlycompared to other techniques.

Message blocks can be very long so largestorage capacities must be available at each


168/188

g pnode to buffer.

When long messages are in processing shorter

message have to wait for long, thusstarvation of short messages.

Note Hence is not used in LANs.

Waiting delay at nodes.

Data is stored so do not support real time(interactive) communication.

Due to unequal message lengths equal

distribution of traffic is not ossible

By priority and queuing, certainportion of the network are busy and


169/188

portion of the network are busy and

can cause blocking of thetraffic/certain messages.

These low priority blockedmessages/traffic are stored andlater routed to sites when they arefree to accept.


170/188

PACKET

SWITCHING

Introduction Much of the time the line is idle in a circuit-switching

approach.


171/188

approach.

In a circuit-switching network, the connection providesfor transmission at constant data rate. Thus the twodevices that are connected must transmit andreceive at the same data rate as the other; thislimits the utility of the network in interconnecting a

variety of host computers and terminals. Packet switching is similar to message switching.

System accepts packets from an information source,stores them in buffer memory and forward to next

packet switch where same store and forwardoperation occurs.

In a packet-switched network, there is no resource

reservation; resources are allocated on demand.

WORKING Data are transmitted inshort packets.


172/188

A upper bound on packetlength is 1000 bytes.

If a source has a longer

message to send, themessage is broken upinto a series of packets.

Each packet contains aportion (or all for ashort message) of theuser's data plus some

Header contains the control information that includesthe information like source and destination addressand packet sequence etc. to route the packet


173/188

and packet sequence etc. to route the packetthrough the network and deliver it to the intended

destination.

Trailer is a checksum used for error control.

At each node en route, the packet is received, storedbriefly, and passed on to the next node.

There is acknowledgement scheme between adjacent

packet switches so that packets can beretransmitted if not acknowledged (when lost intransit) or negatively acknowledged (when error isdetected).

Example Consider a packet to be sent

from station A to stationE.


174/188

The packet will include

control information thatindicates that theintended destination is E.

The packet is sent from A tonode 4. Node 4 stores the

packet, determines thenext leg of the route (say5), and queues the packetto go out on that link (the4-5 link).

When the link is available,the packet is transmittedto node 5, which willforward the packet tonode 6, and finally to E.

ADVANTAGES Line efficiency is greater, as a single node-to-

node link can be dynamically shared by manyk t ti Th k t d


175/188

packets over time. The packets are queued up

and transmitted as rapidly as possible over the link.(In circuit switching, time on a node-to-node link ispreallocated then much of the time, such a link maybe idle because a portion of its time is dedicated toa connection which is idle.)

A packet-switching network can perform data-rate

conversion. Two stations of different data ratescan exchange packets because each connects to itsnode at its proper data rate.

No Blocking: When traffic becomes heavy on a

circuit-switching network, some calls are blocked;the network refuses to accept additional connectionrequests until the load on the network decreases.On a packet-switching network, packets are still

accepted but delivery delay increases

Packet SwitchingTechniques


176/188

Datagram Packet Switching

Virtual Circuit Packet Switching


177/188

Datagram

PacketSwitching

DatagramPacket

Each packet is treatedindependently, with noreference to packets thathave gone before.


178/188

Switching

Station A has a three-packet message to sendto E. It transmits thepackets, 1-2-3, to node4.

On each packet, node 4must make a routingdecision.

Packet 1 arrives for deliveryto E. Node 4 forward thispacket to either node 5or node 7 as the nextstep in the route.

But for packet 3, node 4 finds that its queue fornode 7 is now shorter and so queues packet 3for that node.


179/188

So the packets, each with the same destinationaddress, do not all follow the same route.

Note:The destination address in the header of apacket in a datagram network remains thesame during the entire journey of the packet.

Thus it is possible that packet 3 will beat packet2 to node 6.

Also possible that the packets will be deliveredto E in a different sequence from the one inwhich they were sent.

It is possible for a packet to bedestroyed in the network Example if


180/188

destroyed in the network. Example, if

a packet switching node crashesmomentarily, all of its queued packetsmay be lost.

If this happen node 6 has no way ofknowing that one of the packets in thesequence of packets has been lost.

It is up to E to detect the loss of apacket and figure out how to recoverit.

In this technique, each packet, treated independently, andso called as a datagram (unit of data).


181/188

Datagram switching is a connectionless network. It is

implemented in the network layer.Note - A connectionless network goes directly from an idle

connection (two DTEs are not logically connected to eachother) into a data transfer mode, followed directly by idleconnection.

Thus no connection establishment phase and connectionrelease phase in datagram approach.

It does not have network-wide acknowledgements, flow

control, error recovery. But these services are providedon a link by link basis.

Also called Dynamic Routing.

RoutingTable There are no setup or

teardown phases, soeach switch (or packetswitch) has a routingtable which is based


182/188

table which is basedon the destinationaddress.

The routing tables are

dynamic and are

updated periodically. The destination

addresses and thecorresponding

forwarding outputports are recorded inthe tables.

This is different from the

table of a circuit Datagram packetswitching

Circuit switching


183/188

A datagram network withfour switches (routers)


184/188

ADVANTAGES Call setup phase is avoided. If a station wishes to send


185/188

Call setup phase is avoided. If a station wishes to sendonly one or a few packets, datagram delivery will bequicker.

Faster than Message Switching so can be used in LANsfor supporting interactive traffic.

The efficiency of a datagram network is better than

that of a circuit-switched network as the resourcesare allocated only when there are packets to betransferred.

Flexible If congestion develops in one part of thenetwork, incoming datagrams will be routed tosome other part away from congestion.

Reliable If few nodes fails, packets find alternativeroute bypassing damaged node.


186/188

Enables dynamic use of bandwidth and providehigh robust services.

An upper limit is placed on packet size so that no

user can monopolize any transmission linemore than a few 10s of microsec.

In case of multipacket message, each node can

forward any packet of the multipacket message assoon as the packet arrives, it does not have to wait

for the next packet. This decreases delay andincreases throughput.

Small switching nodes as compared to

electromechanical or computerized switching nodes

is Circuit Switchin

LIMITATIONS Each independent datagram is routed independently, so

processing for each datagrams at each node is longer


187/188

processing for each datagrams at each node is longer.Thus long message datagrams does not hold good.

Congestion avoidance is difficult, as resources are not reserved

in advance and hence necessary BW and routing capacity is notavailable to arriving packets.

Dynamics of traffic patterns in network and lack of timelyknowledge about those patters makes itdifficult for WANs.

Overhead bits in each packet.

Greater delayin a datagram network than in a virtual-circuit

network. Although there are no setup and teardown phases,each packet may experience a wait at a switch before it isforwarded. Not all packets in a message necessarily travelthrough the same switches, the delay is not uniform for the

ackets of a messa e

Delay in a datagramnetwork


188/188

The packet travelsthrough twoswitches.

There are threetransmission

times (3T),threepropagationdelays (slopes3't of the lines),and two waiting

ti (W1 +

Documents

IPG III DC Full Slides