16657572 Lecture Slides on Networking (1)

8/3/2019 16657572 Lecture Slides on Networking (1)

1/126

1

DS126111:FUNDAMENTALS OF COMPUTERNETWORKS

By:John Mpenzi

BSc in CompEng & IT

Msc in Software Engineering

KIGALI INDEPENDENTUNIVERSITY-ULK


2/126

2

CONTENT Communcation network Architecture

Communication protocols Data security and integrity on network Data comm through network and network

mngt

Client-server computing Network concepts,components,protocols OSI Models and Layers LAN and WAN Technologies and Topologies etc


3/126

3

Lesson 1: Networking Basic

This lesson covers the very basics of networking.Well start with a little history that describes howthe networking industry evolved. Well then moveon to a section that describes how a LAN is built:essentially the necessary components (like NICcards and cables). We then cover LAN topologies.And finally well discuss the key networking devices:

hubs, bridges, switches, and routers


4/126

4

Networking History

Early networks

From a historical perspective, electronic communication hasactually been around a long time, beginning with Samuel Morseand the telegraph. He sent the first telegraph message May 24,1844 from Washington DC to Baltimore MD, 37 miles away. Themessage? What hath God wrought? or shaped

Less than 25 years later, Alexander Graham Bell invented thetelephone beating out a competitor to the patent office only by acouple of hours on Valentines Day in 1867. This led to thedevelopment of the ultimate analog network the telephonesystem.

The first bit-oriented language device was developed by EmileBaudot the printing telegraph. By bit-oriented we mean thedevice sent pulses of electricity which were either positive or had

no voltage at all. These machines did not use Morse code.Baudots five-level code sent five pulses down the wire for eachcharacter transmitted. The machines did the encoding anddecoding, eliminating the need for operators at both ends of thewires. For the first time, electronic messages could be sent byanyone


5/126

5

Telephone Network

But its really the telephone network that has had the greatestimpact on how businesses communicate and connect today.Until 1985, the Bell Telephone Company, now known as AT&T,

owned the telephone network from end to end. It representeda phenomenal network, the largest then and still the largesttoday


6/126

6

Developments inCommunication

In 1966, an individual named Carter invented a specialdevice that attached to a telephone receiver that would allowconstruction workers to talk over the telephone from a two-way radio.

Bell telephone had a problem with this and sued andeventually lost.

As a result, in 1975, the Federal Communications Commissionruled that devices could attach to the phone system, if theymet certain specifications. Those specifications were approved

in 1977 and became known as FCC Part 68. In fact, years agoyou could look at the underside of a telephone notmanufactured by Bell, and see the Part 68 stamp of

approval.


7/126

7

1960's - 1970's Communication

In the 1960s and 1970s, traditional computercommunications centered around themainframe host. The mainframe contained allthe applications needed by the users, as well

as file management, and even printing. Thiscentralized computing environment used low-speed access lines that tied terminals to thehost.These large mainframes used digital signals pulses of electricity or zeros and ones, what iscalled binary -- to pass information from theterminals to the host. The informationprocessing in the host was also all digital.


8/126

8

Problems faced in communication

This brought about a problem. The telephone industry wanted to

use computers to switch calls faster and the computer industrywanted to connect remote users to the mainframe using thetelephone service. But the telephone networks speak analog andcomputers speak digital. Lets take a closer look at this problem.Digital signals are seen as ones and zeros. The signal is either onor off. Whereas analog signals are like audio tones for example,

the high-pitched squeal you hear when you accidentally call a faxmachine. So, in order for the computer world to use the servicesof the telephone system, a conversion of the signal had to occur.


9/126

9

The solution

a modulator/demodulator or modem. The modem takes thedigital signals from the computer and modulates ortransforms the signal into analog format. In sendinginformation from a desktop computer to a host using POTSor plain old telephone service, the modem takes the digitalsignals from the computer and modulates the signal into

analog format to go through the telephone system. Fromthe telephone system, the analog signal goes throughanother modem which converts the signal to digital formatto be processed by the host computer.This helped solve some of the distance problems, at leastto a certain extent.


10/126

10


11/126

11

Birth of the personal computer

The birth of the personal computer in 1981 really fueled theexplosion of the networking marketplace. No longer werepeople dependent on a mainframe for applications, file

storage, processing, or printing. The PC gave users incrediblefreedom and power.


12/126

12

The Internet 1970's - 1980's The 70s and 80s saw the beginnings of the Internet. The Internet as we know it today

began as the ARPANET The Advanced Research Projects Agency Network built by adivision of the Department of Defense essentially in the mid 60's through grant-fundedresearch by universities and companies.. Many local area networks connected to theARPANET with TCP/IP. TCP/IP was developed in 1974 and stands for TransmissionControl Protocol / Internet Protocol. The ARPANET was shut down in 1990 due to newernetwork technology and the need for greater bandwidth on the backbone.In the late 70s the NSFNET, the National Science Foundation Network was developed.

This network relied on super computers in San Diego; Boulder; Champaign; Pittsburgh;Ithaca; and Princeton. Each of these six super computers had a microcomputer tied to

it which spoke TCP/IP. The microcomputer really handled all of the access to thebackbone of the Internet. Essentially this network was overloaded from the word "go".

Further developments in networking lead to the design of the ANSNET -- AdvancedNetworks and Services Network. ANSNET was a joint effort by MCI, Merit and IBMspecifically for commercial purposes. This large network was sold to AOL in 1995. TheNational Science Foundation then awarded contracts to four major network accessproviders: Pacific Bell in San Francisco, Ameritech in Chicago, MFS in Washington DCand Sprint in New York City.

By the mid 80's the collection of networks began to be known as the Internet inuniversity circles. TCP/IP remains the glue that holds it together.In January 1992 the Internet Society was formed a misleading name since theInternet is really a place of anarchy. It is controlled by those who have the fastest linesand can give customers the greatest service today.The primary Internet-related applications used today include: Email, News retrieval,Remote Login, File Transfer and World Wide Web access and development.


13/126

13

1990's Global Internetworking

With the growth and development of the Internet came theneed for speed and bandwidth. Companies want to takeadvantage of the ability to move information around theworld quickly. This information comes in the form of voice,data and video large files which increase the demands on

the network. In the future, global internetworking will providean environment for emerging applications that will requireeven greater amounts of bandwidth. If you doubt the futureof global internetworking consider this the Internet isdoubling in size about every 11 months.


14/126

14


15/126

15

LAN building

In this section some of the basic elements neededto build local area networks (LANs) will bedescribed.

The term local-area network, or LAN, describes ofall the devices that communicate togetherprinters, file server, computers, and perhaps evena host computer. However, the LAN is constrainedby distance. The transmission technologies usedin LAN applications do not operate at speed over

long distances. LAN distances are in the range of100 meters (m) to 3 kilometers (km). This range

can change as new technologies emerge


16/126

16

For systems from different manufacturers to

interoperatebe it a printer, PC, and file servertheymust be developed and manufactured according toindustry-wide protocols and standards.

More details about protocols and standards will be givenlater, but for now, just keep in mind they representrules that govern how devices on a network exchangeinformation. These rules are developed by industry-widespecial interest groups (SIGs) and standardscommittees such as the Institute of Electrical andElectronics Engineers (IEEE).


17/126

17

Major characteristics of LANs

The network operates within a building or floor of a

building. The geographic scope for ever more powerfulLAN desktop devices running more powerful applicationsis for less area per LAN.

- LANs provide multiple connected desktop devices(usually PCs) with access to high-bandwidth media.

- An enterprise purchases the media and connections usedin the LAN; the enterprise can privately control the LANas it chooses.

- LANs rarely shut down or restrict access to connected

workstations; local services are usually always available.

- By definition, the LAN connects physically adjacentdevices on the media.


18/126

18

Components of LAN

- Network operating system (NOS)In order for computers to be able to communicate with each

other, they must first have the networking software thattells them how to do so. Without the software, the

system will function simply as a standalone, unable toutilize any of the resources on the network.Network operating software may by installed by thefactory, eliminating the need for you to purchase it, (for

example AppleTalk), or you may install it yourself


19/126

19

Network interface card(NIC)-each network device must also have a networkinterface card. These cards today are alsoreferred to as adapters, as in Ethernet adaptercard or Token Ring adapter card.

-The NIC card amplifies or intensifies electronicsignals which are generally very weak within thecomputer system itself.

-The NIC is also responsible for packaging datafor transmission, and for controlling access tothe network cable.

-The NIC also provides the physical connectionbetween the computer and the transmissioncable (also called media).


20/126

20

Wiring Hub

-In order to have anetwork, you must haveat least two devices thatcommunicate with eachother

-In this simple model, itis a computer and aprinter. The printer alsohas an NIC installed

-which in turn is pluggedinto a wiring hub

-The computersystem is alsoplugged intothe hub, whichfacilitatescommunicationbetween thetwo devices.

-The benefit of buildingthis network is that bysharing resources acompany can affordhigher quality

components. Forexample, instead ofproviding an inkjetprinter for every PC, acompany may purchasea laser printer


21/126

21

Cables or Transmission MediaThe wires connecting the variousdevices together are referred to as

cables.

- Cable prices range frominexpensive to very costly and cancomprise of a significant cost of thenetwork itself.

- Cables are one example oftransmission media. Media arevarious physical environmentsthrough which transmission signalspass. Common networkmedia include twisted-pair, coaxialcable, fiber-optic cable, and theatmosphere (through whichmicrowave, laser, and infraredtransmission occurs). Another termfor this is physical media.

-As their nameimplies, theconnector is thephysical locationwhere the NICcard and thecabling connect.

-Registered jack (RJ)connectors were originallyused to connect telephonelines. RJ connectors arenow used for telephoneconnections and for10BaseT and other typesof network connections.

Different connectors areable supportdifferent speeds oftransmission because oftheir design and thematerials used in

their manufacture.


22/126

22

Network Cabling

Cable is the actual physical path upon which an electricalsignal travels as it moves from one component to

another.

Transmission protocols determine how NIC cards taketurns transmitting data onto the cable

There are three primary cable types:

- Twisted-pair (or copper)

- Coaxial cable and

- Fiber-optic cable


23/126

23

Twisted-pair (or copper)

-Unshielded twisted-pair(UTP) is a four-pair wire

medium used in avariety of networks. UTPdoes not require thefixed spacing betweenconnections that isnecessary with coaxial-

type connections

There are five types of UTP cablingcommonly used as shown below:

- Category 1: Used for telephonecommunications. It is not suitablefor transmitting data.

- Category 2: Capable oftransmitting data at speeds up to 4Mbps.

- Category 3: Used in 10BaseTnetworks and can transmit data atspeeds up to 10 Mbps.

- Category 4: Used in Token Ringnetworks. Can transmit data at

speeds up to 16 Mbps.

- Category 5: Can transmit data atspeeds up to 100 Mbps.


24/126

24

Coaxial cable

-Coaxial cable consists of a

solid copper coresurrounded by an insulator,a combination shield andground wire, and an outerprotective jacket.The shielding on coaxial

cable makes it lesssusceptible to interferencefrom outside sources. Itrequires termination at eachend of the cable, as well asa single ground connection

-Coax supports 10/100 Mbps

and is relatively inexpensive,although more costly than UTP.Coaxial can be cabled overlonger distances than twisted-pair cable. For example,Ethernet can run at speed over

approximately 100 m (300feet) of twisted pair. Using

coaxial cable increases thisdistance to 500 m.


25/126

25

Fiber-optic cable

-Fiber-opticcableconsists ofglass fibersurroundedby shielding

protection:a plasticshield,kevlarreinforcing,and anouter jacket

-Fiber-opticcable is themostexpensive ofthe threetypes

discussed inthis section,but itsupports100+ Mbps

line speeds


26/126

26

There are two types of fiber cable

Single or mono-modeAllows only one mode (orwavelength) of light topropagate throughthe fiber; is capable ofhigher bandwidth and

greater distances thanmultimode. Often usedfor campus backbones.Uses lasers as the lightgenerating method. Singlemode is muchmore expensive thanmultimode cable. Maximumcable length is 100 km.

-MultimodeAllowsmultiple modes of light topropagate through thefiber. Often usedfor workgroupapplications. Uses light-emitting diodes (LEDs)as light generatingdevice. Maximum cablelength is 2 km

-Super servers,high-capacityworkstations, andmultimediaapplications havealso fueled theneed for highercapacity

bandwidths.


27/126

27

Throughput Needs....!!

The examples on above

image shows that theneed for throughputcapacity grows as aresult of a desire totransmit more voice,video, and graphics. The

rate at which thisinformation may be sent(transmission speed) isdependent on how datais transmitted and themedium used fortransmission.

-The throughput rateis the rate ofinformation arriving at,

and possibly passingthrough, a particularpoint in a network.

The term bandwidthmeans the total capacity of agiven network medium(twisted pair, coaxial, orfiber-optic cable) or protocol.

- Bandwidth is also used to

describe the differencebetween the highest and thelowestfrequencies available fornetwork signals. Thisquantity is measured inMegahertz (MHz).

-The bandwidth of a givennetwork medium orprotocol is measured in bitsper second (bps).


28/126

28


29/126

29

Questions Qn1.Write short notes on the following:

a) LAN, its characteristics and why its implementation

b) Discuss different types of soft wares and giveexamples

c) Give different types of operating systems

d) Discuss different types of LAN components

e) Explain the term Network cabling

f) Discuss different kinds of cabling

g) What do you understand by the term Topology

h) Explain different modes of Topologies


30/126

30

LAN Topologies

Topology refers to thephysical arrangementof network componentsand media within anenterprise networking

structure. There arefour primary kinds ofLAN topologies: bus,tree, star, and ring.

-Bus and Treetopology

-Star topology

-Ring topology-


31/126

31

Bus topology

-Is A linear LAN architecturein which transmissions fromnetwork componentspropagate the length of

the medium and arereceived by all othercomponents.- The bus portion is the

common physical signalpath composed of wires orother media across whichsignals can be sent from

one part of a network toanother. Sometimes called ahighway.- Ethernet/IEEE 802.3networks commonlyimplement a bus topology

Tree topology is

- Similar to bus topology,except that tree networks can

contain branches with multiplenodes. As in bus topology,

transmissions from onecomponent propagate thelength of the medium andare received by all othercomponents.

The disadvantage of bustopology is that if theconnection to any one user isbroken, the entire networkgoes down, disruptingcommunication between allusers. Because of thisproblem, bus topology israrely used today.The advantage of bus topology

is that it requires less cabling(therefore, lower cost) thanstar topology.


32/126

32

Star topology

-Star topology is aLAN topology in whichendpoints on anetwork areconnected to acommon centralswitch or hub bypoint-to-point links.

-The benefit of startopology is that even ifthe connection to any

one user is broken, thenetwork staysfunctioning, andcommunication betweenthe remaining users isnot disrupted

The disadvantage ofstar topology is thatit requires more

cabling (therefore,higher cost)than bus topology.


33/126

33

Ring topology

-Ringtopologyconsists of aseries ofrepeatersconnected toone another

byunidirectionaltransmissionlinks to forma singleclosed loop.

-Each station on

the networkconnects to thenetwork at arepeater.-A ringtopologies aremost often

organized in a

closed-loop star


34/126

34

LAN/WAN Devices

Hub

Bridges

Switches

Routers


35/126

35

Hub

-Star topology networks generally havea hub in the center of the network thatconnects all of the devices together

using cabling. When bits hit a networkingdevice, be they hubs, switches, or routers,the devices will strengthen the signal andthen send it on its way.A hub is simple a multi-port repeater. There isusually no software to load, and noconfiguration required (i.e. network

administrators dont have to tell the device

what to do).

-Devices on thenetwork are constantly

listening for data. Whendevices sense a frameof information that isaddressed (and we willtalk more aboutaddressing later) for it,

then it will accept thatinformation into

memory found on thenetwork interface card(NIC) and beginprocessing the data.

-In fairly small networks,hubs work very well.However, in large

networks the limitationsof hubs creates

problems for network

managers. In thisexample, Ethernet is thestandard being used.The network is also baseband, only one station

can use the network at atime. If the applicationsand files being used onthis network are large,and there are morenodes on the network,contention for bandwidthwill slow the

responsiveness of thenetwork down


36/126

36

Bridges

-Bridges improvenetwork throughput

and operate at a moreintelligent level thando hubs. A bridge isconsidered to be astore and forwarddevice that usesunique hardwareaddresses to filter

traffic that wouldotherwise travel fromone segment toanother.

Bridge functions:

Reads data frame headers and

records source address/port(segment) pairs- Reads the destination address ofincoming frames and usesrecorded addresses to determinethe appropriate outbound portfor the frame.- Uses memory buffers to store

frames during periods of heavytransmission, and forwards

them when the medium is ready

-When station 124transmits to station 125,the frame goes into thehub (who repeats it andsends it out allconnected ports) andthen on to the bridge.The bridge will notforward the framebecause it recognizesthat stations 124 and125 are on the samesegment. Only trafficbetween segmentspasses through thebridge

NB:When one station

transmits, all otherstations must wait untilthe line is silent againbefore transmitting


37/126

37

Switches

-Switches use bridgingtechnology to forwardtraffic between ports. Theyprovide full dedicatedtransmission rates betweentwo stations that are

directly connected to theswitch ports. Switches alsobuild and maintain addresstables just like bridges do.These address tables areknown as contentaddressable memory.

-Replacing the two hubsand the bridge with an

Ethernet switch providesthe users with dedicatedbandwidth. Each stationhas a full 10Mbps pipe tothe switch. With a switchat the center of thenetwork, combined with

the 100Mbps links, usershave greater access to the

network


38/126

38

Routers A router has two basic functions, path determination using a

variety of metrics, and forwarding packets from one networkto another. Routing metrics can include load on the link between devices,

delay, bandwidth, and reliability, or even hop count (i.e. thenumber of devices a packet must go through in order to reachits destination).

In essence, routers will do all that bridges and switches will do,plus more. Routers have the capability of looking deeper intothe data frame and applying network services based on thedestination IP address. Destination and Source IP addressesare a part of the network header added to a packetencapsulation at the network layer.


39/126

39

Lesson 2: OSI Reference Model

This lesson covers the OSI reference model. It is sometimes

also called ISO or 7 layer reference model. The model was developed by the International Standards

Organization in the early 1980's.

It describes the principles for interconnection of computersystems in an Open System Interconnection environment


40/126

40

The Layered Model The concept of

layeredcommunication

is essential toensuringinteroperabilityof all the piecesof a network

In this image, the

goal is to get amessage fromLocation A toLocation B. Thesender doesntknow whatlanguage the

receiver speaks

so the senderpasses themessage on to a

translator.

The translator, while notconcerned with thecontent of the message,

will translate it into alanguage that may beglobally understood bymost, if not alltranslators thus itdoesnt matter whatlanguage the finalrecipient speaks. In thisexample, the language

is Dutch. The translatoralso indicates what thelanguage type is, andthen passes themessage to anadministrativeassistant.


41/126

41

Upon closer study of the processemployed to communicate, you willnotice that communication took placeat different layers. At layer 1, theadministrative assistantscommunicated with each other. Atlayer 2, the translators communicated

with each other. And, at layer 3 thesender was able to communicate withthe recipient.


42/126

42

Why a Layered Network Model.........?

-A layerednetwork

model doesa number ofthings.

1. It reduces thecomplexity ofthe problems

from one large

one to sevensmaller ones

2. It allows thestandardizationof interfacesamong devices

3. It also facilitatesmodular engineering soengineers can work on onelayer of the network modelwithout being concerned

with what happens atanother layer


43/126

43

Devices Function at Layers

-Each of thesedevices operates ata different level ofthe OSI Model

-NIC cards receiveinformation fromupper level

applications andproperly packagedata for

transmission on to

the network media

-NIC cards liveat the lowerfour layers ofthe OSI Model.

-Hubs, whetherEthernet live atthe physical layer.

They are onlyconcerned withpassing bits fromone station toother connectedstations on thenetwork

-Bridges andswitches onthe otherhand, will filtertraffic andbuild bridgingand switchingtables in order

to keep trackof what device

is connectedto what port

Routers, or thetechnology of

routing, lives atlayer 3.


44/126

44

A closer look at Host& Media Layers

Host Layers :-

The upper four layers,Application, Presentation,Session, and Transport, are

responsible for accuratedata delivery betweencomputers. The tasks orfunctions of these upper

four layers mustinteroperate with theupper four layers in thesystem being

communicated with.

Media Layers :-

The lower three layers Network, Data Link and

Physical -- are called themedia layers. The medialayers are responsible forseeing that theinformation does indeedarrive at the destinationfor which it was intended


45/126

45

Layer Functions


46/126

46

Layers in a Peer-to-Peer Communications Network

In this exercise wepackage information andmove it from Host A,across network lines to

Host B.

Each layer uses its ownlayer protocol tocommunicate with itspeer layer in the othersystem. Each layersprotocol exchanges

information, calledprotocol data units(PDUs), between peerlayers.

-This peer-layerprotocolcommunication isachieved by usingthe services of thelayers below it. Thelayer below anycurrent or active

layer provides itsservices to the

current layer

-The transportlayer will ensurethat data is keptsegmented orseparated fromone other data.

-At the networklayer we getpackets thatbegin to beassembled. Atthe data linklayer thosepackets becomeframes and then

at the physicallayer thoseframes go out onthe wires fromone host to theother host asbits


47/126

47

Data Encapsulation

-This whole process of moving data fromhost A to host B is known as dataencapsulation the data is being wrapped

in the appropriate protocol header so it canbe properly received.

Lets say we compose an email that we wishto send from system A to system B. Theapplication we are using is Eudora. We writethe letter and then hit send. Now, thecomputer translates the numbers into ASCIIand then into binary (1s and 0s). If theemail is a long one, then it is broken up andmailed in pieces. This all happens by thetime the data reaches the Transport layer.

-At thenetworklayer, anetworkheader isadded to the

data. Thisheader

containsinformationrequired tocomplete thetransfer, suchas source and

destinationlogicaladdresses

-The packet from thenetwork layer is thenpassed to the datalink layer where aframe header and aframe trailer areadded thus creating adata link frame.


48/126

48

Layers 1 & 2: Physical &Data Link Layers

-Locatingcomputer

systems on aninternet workis an essentialcomponent ofany networksystem

-Every NIC card onthe network has its

own MAC address. Inthis example we havea computer with theMAC address000.0C12.3456. TheMAC address is a

hexadecimal numberso the numbers in

this address heredont go just fromzero to nine, but gofrom zero to nine andthen start at "A" andgo through "F".

-Every type ofdevice on anetwork has a

MAC address


49/126

49

Layer 3: Network Layer

Network Layer: Path Determination

-Path determination occursat Layer 3. The path

determination functionenables a router to evaluatethe available paths to adestination and to establishthe preferred handling of apacket

-Data can take differentpaths to get from a source to

a destination.

-At layer 3, routersreally help determine

which path. Thenetwork administratorconfigures the routerenabling it to make anintelligent decision asto where the routershould sendinformation through

the cloud.

-The network layersends packets fromsource network todestination network.After the router

determines which pathto use, it can proceedwith switching thepacket: taking thepacket it accepted onone interface andforwarding it to

another interface orport that reflects thebest path to thepackets destination.


50/126

50

AddressingNetwork and Node

-Each device in a localarea network is given alogical address. The first

part is the networknumber

-The second part is a nodenumber, in this example wehave nodes 1, 2, and 3. Therouter uses the networknumber to forward

information from onenetwork to another.

-This host addressderives automatically

from information inhardware of thespecific LAN device.


51/126

51

Network Layer Protocol Operations

Take an exampleThe message willexit Station X andtravels through thecorporate internalnetwork until it getsto a point where itneeds the services ofan Internetservice provider

-The messagewill bounce or

jump throughtheir network andeventually arriveat MomsInternet provider

in Dearborn

-As information

travels fromStation X itreaches thenetwork levelwhere anetworkaddress is

added to thepacket.

At the datalink layer, theinformation is

encapsulatedor summed upan Ethernetframe. Then itgoes to therouter hereit is Router A


52/126

52

Multi-protocol Routing

-Routers arecapable ofunderstandingaddressinformationcoming from

many differenttypes of networksand maintainingassociated routingtables for severalrouted protocolsconcurrently

-As the routerreceives packetsfrom the users onthe networks usingIP, it builds a routingtable containing the

addresses of thenetwork of these IPusers.

-Routing tables cancontain addressinformation frommultiple protocolnetworks.

-Each uses routingtables to determinepaths


53/126

53

Routed Versus Routing Protocol

Routed protocols -They are any network

protocol suite that providesenough information in itsnetwork layer address toallow a packet to direct usertraffic

Routing protocol

support a routedprotocol by providingmechanisms for sharingrouting information.Routing protocolmessages movebetween the routers.

-A routing protocolallows the routers tocommunicate with otherrouters to update andmaintain tables

-Routing protocol

messages do notcarry end-usertraffic from networkto network.

-A routing protocoluses the routedprotocol to passinformationbetween routers

Static Versus Dynamic Routes


54/126

54

Static Versus Dynamic RoutesRouters must be awareof what links, or lines,on the network are upand running, which onesare overloaded, or whichones may even be downand unusable. There aretwo primary methodsrouters use to determinethe best path to adestination: static anddynamic

Static knowledge isadministeredmanually: a networkadministrator enters

it into the routersconfiguration

Dynamic knowledgeworks differently. Afterthe networkadministrator entersconfigurationcommands to startdynamic routing, routeknowledge isupdatedautomatically by arouting processwhenever new topologyinformation is receivedfrom the internetwork

Dynamic routingtends to revealeverything knownabout aninternetwork. For

security reasons, itmight be appropriateto conceal parts of aninternetwork. Staticrouting allows aninternetworkadministrator to

specify what isadvertised aboutrestricted partitions.


55/126

55

Layers 47: Transport, Session, Presentation, andApplication Layers

Transport Layer

It also establishes theend-to-end connection,

from your host to anotherhost. As the transportlayer sends its segments,it can also ensure dataintegrity. Essentially thetransport layer opens upthe connection from your

system through a networkand then through a widearea cloud to thereceiving system at theother end

Transport Layer Segments Upper-Layer Applications

it segments upper layer applicationinformation. You might have more

than one application running onyour desktop at a time. You mightbe sending electronic mail openwhile transferring a file from theWeb, and opening a terminalsession. The transport layer helpskeep straight all of the information

coming from these differentapplications.


56/126

56

Transport Layer Establishes Connection

-Another function of thetransport layer is toestablish the connection

from your system toanother system

-When you arebrowsing the Weband double-click on alink your system tries

to establish aconnection with thathost

-Once the connectionhas been established,there is somenegotiation that

happens between yoursystem and the systemthat you are connectedto in terms of how datawill be transferred

-Once the negotiations

are completed, data willbegin to transfer. As soonas the data transfer iscomplete, the receivingstation will send you theend message and yourbrowser will say done.

-Essentially, thetransport layer isresponsible then forconnecting andterminatingsessions from yourhost to anotherhost

Transport Layer Sends Segments with Flow


57/126

57

Transport Layer Sends Segments with FlowControl

Another important functionof the transport layer is tosend segments andmaintain the sending and

receiving of informationwith flow control.

When a connection isestablished, the hostwill begin to sendframes to the receiver.When frames arrivetoo quickly for a hostto process, it stores

them in memorytemporarily. If theframes are part of asmall burst, thisbuffering solves theproblem. If the trafficcontinues, the host or

gateway eventuallyexhausts its memoryand must discardadditional frames thatarrive.

Instead of losing data,the transport functioncan issue a not ready

indicator to thesender. Acting like astop sign, thisindicator signals thesender to discontinuesending segmenttraffic to its peer

After the receiver hasprocessed sufficientsegments that itsbuffers can handleadditional segments,the receiver sends aready transportindicator, which is likea go signal. When itreceives this indicator,the sender canresume segment

transmission.


58/126

58

Lesson 3: Introduction to TCP/IPTCP/IP isshorthandfor a suite ofprotocolsthat run ontop of IP

IP-is theInternet

Protocol, andTCP is the

mostimportantprotocol thatruns on topof IP. Anyapplicationthat cancommunicat

e over theInternet isusing IP,

-Protocols that run ontop of IP include:

TCP, UDP and ICMP

-TCP/IP protocols worktogether to break data intopackets that can be routedefficiently by the network. In

addition to the data, packetscontain addressing,sequencing, and errorchecking information. Thisallows TCP/IP to accuratelyreconstruct the data at theother end.

Heres an analogy of what TCP/IPdoes. Say youre moving across thecountry. You pack your boxes and

put your new address on them. Themoving company picks them up,makes a list of the boxes, and shipsthem across the country using themost efficient route. That mighteven mean putting different boxeson different trucks. When the boxesarrive at your new home, you check

the list to make sure everything hasarrived (and in good shape), andthen you unpack the boxes andreassemble your house.

A suite of protocols- Rules that dictate howpackets of information aresent across - multiplenetworks- Addressing- Error checking

IP


59/126

59

IP-Every computer on theInternet has at least one

address that uniquelyidentifies it from all othercomputers on the Internet(aptly called its IP address!).

-When you send or receive

datasay an emailmessage or web pagethe message gets dividedinto little chunks calledpackets or data grams.

-Each of these packets

contains both the source IPaddress and the destinationIP address.

-IP looks at thedestination address

to decide what to donext. If thedestination is on thelocal network, IPdelivers the packetdirectly

-If the destinationis not on the localnetwork, then IPpasses the packetto a gateway

usually a router

-Computers usuallyhave a single defaultgateway. Routers

frequently haveseveral gatewaysfrom which tochoose. A packetmay get passedthrough several

gateways beforereaching one that ison a local networkwith the destination.

TCP/IP Transport Layer


60/126

60

TCP/IPTransport Layer

-After TCP/IP was invented anddeployed, the OSI layered networkmodel was accepted as a standard.OSI neatly divides network protocolsinto seven layers; the bottom fourlayers are shown in this diagram. The

idea was that TCP/IP was aninteresting experiment, but that itwould be replaced by protocols basedon the OSI model

TCP/IP Applications

- Application layer

- File Transfer Protocol (FTP)- Remote Login (Telnet)- E-mail (SMTP)

- Transport layer

- Transport Control Protocol (TCP)- User Datagram Protocol (UDP)

- Network layer

- Internet Protocol (IP)

- Data link & physical layer

- LAN Ethernet, Token Ring, FDDI, etc.- WAN Serial lines, Frame Relay, X.25, etc

TCP is the most important of all the IPprotocols. Most Internet applications youcan think of use TCP, including: Telnet,HTTP (Web), POP & SMTP (email) andFTP (file transfer).


61/126

61

TCPTransmission Control Protocol

TCP stands for Transmission Control Protocol.

-TCP establishes a reliableconnection between twoapplications over the network.

This means that TCP guaranteesaccurate, sequential delivery of

your data. If something goeswrong, TCP reports an error, soyou always know whether yourdata arrived at the other end.

Heres how it works:Every TCP connection is

uniquely identified by fournumbers:

- source IP address- source port- destination IP address- destination port

-TCP packets also includea checksum to verify theintegrity of the data.Packets that fail

checksum getretransmitted


62/126

62

UDP User Datagram Protocol

Unreliable- Fast

- Assumes applicationwill retransmit onerror- Often used indiskless workstations

-UDP is a fast, unreliableprotocol that is suitable forsome applications.Unreliable means there is nosequencing, no guaranteeddelivery (no automatic

retransmission of lost packets)and sometimes no checksums.Fast means there is noconnection setup time, unlikeTCP

ICMPPing

-Ping is an example of aprogram that uses ICMPrather than TCP or UDP. Pingsends an ICMP echo requestfrom one system to another,then waits for an ICMP echo

reply. It is mostly used fortesting.


63/126

63

IPv4 Addressing

-IPv4 addresses are 32bits long and are usually

written in dot notation.An example would be192.1.1.17.

-The Internet is

actually a lot of smalllocal networksconnected together.Part of an IP addressidentifies which localnetwork, and part ofan IP addressidentifies a specificsystem or host on thatlocal network.

-What part of an IP

address is for thenetwork andwhat part is for the

host isdetermined by theclass or the subnet

11000000

IP AddressingThree Classes


64/126

64

IP AddressingThree Classes

-Class A: NET.HOST.HOST.HOST- Class B: NET.NET.HOST.HOST

- Class C: NET.NET.NET.HOST

To summarize:

-IPv4 addresses are 32 bits with anetwork part and a host part.

-The network part of an address isused for routing packets over theInternet. The host part is used forfinal delivery on the local net.

IP Add i Cl A


65/126

65

IP AddressingClass A

Heres an example ofa class A address.Any IPv4 address in

which the first octetis less than 128 is bydefinition a class Aaddress.This address is forhost #222.135.17 onnetwork #10,

although the host isalways referred to byits full address.

Example:-10.222.135.17- Network # 10

- Host # 222.135.17- Range of class Anetwork IDs: 1126- Number ofavailable hosts:16,777,214

IP AddressingClass BHeres an example

of a class Baddress. Any IPv4address in whichthe first octet isbetween 128 and191 is by definitiona class B address

Example: -128.128.141.245

- Network #128.128- Host # 141.245

- Range of class Bnetwork IDs:128.1191.254- Number ofavailable hosts:65,534

IP AddressingClass C

Heres an example of aclass C address. Most IPv4addresses in which thefirst octet is 192 or higherare class C addresses, butsome of the higher rangesare reserved for multicast

applications.

Example: - 192.150.12.1

-Network # 192.150.12-Host # 1-Range of class C network

IDs: 192.0.1223.255.254-Number of availablehosts: 254


66/126

66

140.1

79.220.200

Written in binary form:

140 .179 .220 .200

10001100.10110011.11011100.11001000

We see the address in the decimal form Your computer sees it in the binary form

example

Binary Octet:


67/126

67

Binary Octet:

An octet is made up of eight 1s and/or 0s, representing

the following values:

128 64 32 16 8 4 2 1

So the value of140 (the first octet of our example) looks likethis:

1 0 0 0 1 1 0 0


68/126

68

Contd

1 0 0 0 1 1 0 0

128 + 0 + 0 + 0 + 8 + 4 + 0 + 0 = 140


69/126

69

Lesson 4: LAN Basics

covers thefundamentals ofLAN technologies.Well look atEthernet, TokenRing, and FDDI.For each one, welllook at the

technology as wellas its operations.

The most popular local area networkingprotocol today is Ethernet. Most networkadministrators building a network from scratchuse Ethernet as a fundamental technology.

Token Ring technology is widely used in IBM

networks.

FDDI networks are popular for campus LANs and are usually built to support highbandwidth needs for backbone connectivity.

Ethernet


70/126

70

Ethernet

Ethernet and IEEE 802.3

Ethernet was initially developedby Xerox. They were laterjoined by Digital EquipmentCorporation (DEC) and Intel todefine the Ethernet 1specification in 1980.

There have been furtherrevisions including the Ethernetstandard (IEEE Standard802.3) which defines rules forconfiguring Ethernet as well asspecifying how elements in anEthernet network interact withone another.

Ethernet is the most popularphysical layer LAN technologybecause it strikes a good balancebetween speed, cost, and ease ofinstallation.

These strong points, combinedwith wide acceptance in thecomputer marketplace and the

ability to support virtually allpopular network protocols, make

Ethernet an ideal networkingtechnology for most computer users

today

The Fast Ethernet standard (IEEE802.3u) has been established fornetworks that need highertransmission speeds. It raises theEthernet speed limit from 10Mbps to 100 Mbps with onlyminimal changes to the existing

cable structure

Gigabit Ethernet is an

extension of the IEEE 802.3Ethernet standard.

It increases speed tenfoldover Fast Ethernet, to 1000Mbps, or 1 Gbps.


71/126

71

Benefits andbackground

- Ethernet is the

most popularphysical layer LANtechnology becauseit strikes a goodbalance betweenspeed, cost, andease of installation

- Supports virtuallyall networkprotocols- Xerox initiated,then joined by DEC& Intel in 1980

Revisions ofEthernetspecification

- Fast Ethernet (IEEE 802.3u)raises speed from 10 Mbps to100 Mbps- Gigabit Ethernet is anextension of IEEE 802.3 whichincreases speeds to 1000 Mbps,

or 1 Gbps

Ethernet Protocol Names

Ethernet and Fast Ethernet


72/126

72

Ethernet and Fast Ethernet

-This chart gives you an idea of therange of Ethernet protocols includingtheir data rate, maximum segmentlength, and medium

-Ethernet has survived as an essential

media technology because of itstremendous flexibility and its relative

simplicity to implement and understand

Ethernet Operation

-Lets say in our example herethat station A is going to send

information to station D. StationA will listen through its NIC cardto the network. If no otherusers are using the network,station A will go ahead andsend its message out on to thenetwork. Stations B and C and

D will all receive thecommunication.


73/126

73

-At the data link layer it willinspect the MAC address.Upon inspection station Dwill see that the MAC

address matches its ownand then will process theinformation up through therest of the layers of theseven layer model.

As for stations B & C, they toowill pull this packet up to theirdata link layers and inspectthe MAC addresses. Upon

inspection they will see thatthere is no match between thedata link layer MAC addressfor which it is intended andtheir own MAC address andwill proceed to dump the

packet.

Ethernet Reliability


74/126

74

Ethernet Reliability

Ethernet is known as being a

very reliable local areanetworking protocol. In thisexample, A is transmittinginformation and B also hasinformation to transmit. Letssay that A & B listen to thenetwork, hear no traffic and

broadcast at the same time. A

collision occurs when these twopackets crash into one anotheron the network. Bothtransmissions are corrupted andunusable.

-When a collision

occurs on thenetwork, the NICcard sensing thecollision, in this case,station C sends out a

jam signal that jamsthe entire network for

a designated amountof time


75/126

75

-Once the jam signal hasbeen received and

recognized by all of thestations on the network,stations A and D will bothback off or pull back fordifferent amounts of timebefore they try to retransmit.This type of technology isknown as Carrier Sense

Multiple Access WithCollision Detection CSMA/CD.

High-Speed Ethernet


76/126

76

Options

Fast Ethernet- Fast EtherChannel- Gigabit Ethernet

- Gigabit EtherChannel

-Weve mentioned that Ethernetalso has high speed options that

are currently available. FastEthernet is used widely at thispoint and provides customers with100 Mbps performance, a ten-foldincrease. Fast EtherChannel is aCisco value-added feature thatprovides bandwidth up to 800

Mbps. There is now a standard forGigabit Ethernet as well and Ciscoprovides Gigabit Ethernet solutionswith 1000 Mbps performance.

What Is Fast Ether Channel?


77/126

77

-Grouping of multiple FastEthernet interfaces intoone logical transmission

path

-Scalable

bandwidth upto 800+ Mbps- Usingindustry-standard FastEthernet- Load

balancingacross parallellinks- Extendable toGigabitEthernet

What Is GigabitEthernet?

-In some cases, Fast EtherChannel technology may not beenough

-The rule of thumb today is toplan for 80 percent of the traffic

going over the backbone

-Gigabit networking is important to accommodatethese evolving needs.Gigabit Ethernet builds on the Ethernet protocol butincreases speed tenfold over Fast Ethernet, to 1000Mbps, or 1 Gbps

S T l i


78/126

78

Summary Topologies- Summary -- LAN technologies include Ethernet, Token Ring, and FDDI

- Ethernet

- Most widely used- Good balance between speed, cost, and ease of

installation

- 10 Mbps to 1000 Mbps

- Token Ring

- Primarily used with IBM networks- 4 Mbps to 16 Mbps

- FDDI

- Primarily used for corporate backbones

- Supports longer distances- 100 Mbps

Assignment

Go and read about

Token RingArchitecture/technologies and FDDI

Ref:1.Wesley

2.Williamstallings(Businessdata communication.

Prepare forpresentation

Lesson 5: Understanding LAN Switching


79/126

79

Lesson 5: Understanding LAN Switching

Shared LAN Technology

LAN Switching Basics

- Key Switching Technologies

Sh d LAN T h l


80/126

80

Shared LAN Technology

-As you can see indicated in the

diagram on the left, Ethernet isfundamentally what we call ashared technology

-All users of a given LANsegment are fighting for thesame amount of bandwidth

-Frames, or packets, do in

our network as we're tryingto make transmissions onour Ethernet (framesfighting for the samebandwidth

Oth B d idth C


81/126

81

Other Bandwidth Consumers

we basically haveone transmitterthat's trying toreach one receiver,

which is by far themost common, orhopefully the mostcommon form ofcommunication inour network.

Another way tocommunicate is with

a mechanism knownas a broadcast. Andthat is when onetransmitter is tryingto reach all receiversin the network.So, asyou can see in thediagram, in themiddle, our serverstation is sending outone message, and it'sbeing received byeveryone on thatparticular segment.

And a multicast iswhen onetransmitter is

trying to reach, noteveryone, but asubset or a groupof the entiresegment

Hub-Based LANsBridges


82/126

82

Hubs are introduced into

the network as a betterway to scale our thin andthick Ethernet networks

-Basically what we haveis an individual desktopconnection for eachindividual workstation or

server in the network,and this allows us tocentralize all of ourcabling back to a wiringcloset

Bridges

Another way is to add bridges.In order to scale our networks

we need to do somethingknown as segmentation

-And bridges provide acertain level ofsegmentation in ournetwork. And bridges dothis by adding a certain

amount of intelligence intothe network

Bridges operate at Layer2, while hubs operate atLayer 1. So operating atLayer 2 gives us more

intelligence in order tomake an intelligent

forwarding decision.

-bridges are moreintelligent than a hub,

because they canactually listen in, oreavesdrop on thetraffic going throughthe bridge, they canlook at source anddestination addresses,

and they can build atable that allows themto make intelligentforwarding decisions.

SwitchesLayer 2


83/126

83

switches to provide themost control in ournetwork, at least at

Layer 2

-So that when we look atour network we see thatphysically each stationhas its own cable intothe network, well,conceptually we can

think of this as eachworkstation having theirown lane through thehighway

-Theres -micro-segmentation. That's

a fancy way simply tosay that eachworkstation gets itsown dedicatedsegment through thenetwork.

Switches versus Hubs


84/126

84

-If we compare that with ahub or with a bridge, we'relimited on the number ofsimultaneous conversationswe can have at a time

-Remember, if two stationstried to communicate in ahubbed environment, thatcollisions

-In a switched environment,not going to expectcollisions because eachworkstation has its owndedicated path through thenetwork.

-In terms ofbandwidth, and

scalability, is we havedramatically morebandwidth in thenetwork.

-Each station now willhave a dedicated 10megabits per second

worth of bandwidth.

The Need for Speed: Early Warning Signs


85/126

85

The Need for Speed: Early Warning Signs

-include increased delay on our filetransfers

-print jobs that take a very long timeto print out. (From the time wequeue them from our workstation, tillthe time they actually get printed, ifthat's increasing, that's an indicationthat we may have some LAN

congestion problems. )

Typical Causes of NetworkCongestion

-if we have too many users on ashared LAN segment. (Rememberthat shared LAN segments have a fixedamount of bandwidth )

-As we add users, proportionally, we'redegrading the amount of bandwidth per

user

-So we gonna get to a certain numberof users and gonna be to be too muchcongestion, too many collisions, toomany simultaneous conversations tryingto occur all at the same time

Network Traffic Impact from Centralization ofServers


86/126

86

-Also, the way thetraffic is distributed onour network can have

an impact as well. Avery common thing todo in many networks isto build what's knownas a server farm

-In a server farm

effectively what we'redoing is centralizing allof the resources on ournetwork that need to beaccessed by all of theworkstations in ournetwork.

-So what happens,we cause congestionon those centralizedsegments within thenetwork

-Servers are gradually moving into acentral area (data center) versus

being located throughout thecompany to:

- Ensure company data integrity- Maintain the network and ensureoperability- Maintain security

- Perform configuration andadministrative functions

Todays LANs


87/126

87

Today s LANs

Mostly switched resources;few shared- Routers provide scalability- Groups of usersdetermined by physicallocation

Chapter 6: WAN Basics


88/126

88

In this chap, Wegonna at ,what a WAN

is, to talking aboutbasic technology suchas WAN devices ,circuit and packetswitching also covertransmission options

from POTS (plain oldtelephone service) toFrame Relay, to leasedlines etc

So, what is a WAN? A WAN isa data communicationsnetwork that serves users

across a broad geographicarea and often usestransmission facilitiesprovided by common carrierssuch as telephone companies

-These providers arecompanies like

MTN,Rwandatel,ISPA,Artel,KIST etc

- Telephone serviceis commonlyreferred to as plain

old telephoneservice (POTS).

- WAN technologiesfunction at thelower three layersof the OSI referencemodel: thephysical layer, thedata link layer, andthe network layer.

Common WAN networkcomponents include WANswitches, access servers,modems, CSU/DSUs, and ISDNTerminals

WAN Devices


89/126

89

-A WAN switch is a multiportinternetworking device used incarrier networks

-operate at the data link layerof the OSI reference model.

-These WAN switches can sharebandwidth among allocatedservice priorities, recover fromoutages, and provide networkdesign and management

systems.

A modem is a device thatinterprets digital andanalog signals, enabling

data to be transmitted overvoice-grade telephonelines. At the source, digitalsignals are converted toanalog. At the destination,these analog signals arereturned to their digit

An access server is a concentrationpoint for dial-in and dial-outconnections.

A channel service unit/digitalservice unit (CSU/DSU) is a digital

interface device that adapts thephysical interface on a dataterminal equipment device (suchas a terminal) to the interface of adata circuit terminating (DCE)device (such as a switch) in aswitched-carrier network

-The CSU/DSUalso providessignal timingforcommunicationbetween thesedevices.

An ISDN terminalWAN TerminatingEquipment

DTE is the


90/126

90

An ISDN terminalis a device usedto connect ISDNBasic RateInterface (BRI)connections toother interfaces,such as EIA/TIA-232.

- A terminaladapter is

essentially anISDN modem

Equipment

The WAN physical layerdescribes the interfacebetween the data terminalequipment (DTE) and thedata circuit-terminatingequipment (DCE).

-Typically, the DCE isthe service provider,

attacheddevice (thecustomers

device).

In this model, theservices offered tothe DTE are madeavailable through amodem or channelservice unit/dataservice unit(CSU/DSU).

CSU/DSU (ChannelCircuitSwitching


91/126

91

U/ U ( aService Unit / DataService Unit) Devicethat connects the end-user equipment to thelocal digital telephoneloop or to the serviceproviders datatransmission loop.

-The DSU adapts thephysical interface on a

DTE device to atransmission facilitysuch as T1 or E1.

-Also responsible forsuch functions assignal timing for

synchronous serialTransmission

Circuit switching is a WANswitching method in which adedicated physical circuit orpath is established,maintained, and terminated

through a carrier networkfor each communicationsession

-circuit switchingoperates much like anormal telephone call.Integrated Services

Digital Network (ISDN)is an example of acircuit-switched WANtechnology.

Packet Switching


92/126

92

Packet switchingis a WAN

switching methodin which networkdevices share asingle point-to-point link totransport packetsfrom a source to

a destinationacross a carriernetwork

Statistical multiplexing is usedto enable devices to sharethese circuits or routes.Asynchronous Transfer Mode(ATM), Frame Relay, SwitchedMulti-megabit Data Service(SMDS), and X.25 areexamples of packet-switched

WAN technologies.

Network devicesshare a point-to-point

link to transportpackets from a sourceto a destinationacross a carriernetwork

- Statisticalmultiplexing is usedto enable devices toshare these circuits

WAN Virtual CircuitsA l i l i it i


93/126

93

- A logical circuit ensuringreliable communication betweentwo devices

- Switched virtual circuits (SVCs)

- Dynamically established ondemand

- Torn down whentransmission is complete

- Used when data transmissionis sporadic

- Permanent virtual circuits(PVCs)

- Permanently established- Save bandwidth for cases

where certain virtual circuits mustexist all the time

A virtual circuit is alogical circuit createdto ensure reliablecommunicationbetween two networkdevices.

-Two types of virtualcircuits exist:

switched virtualcircuits (SVCs) andpermanent virtualcircuits (PVCs

SVCs are dynamicallyestablished on demandand are torn down whentransmission is complete.

SVCs are used insituations where data

transmission is sporadicor periodic,or random.

PVCs are permanentlyestablished. PVCs savebandwidth associatedwith circuit establishment

and tear down insituations where certainvirtual circuits must existall the time.

WAN Protocols


94/126

94

-The OSI model providesa conceptual frameworkfor communication

between computers, butthe model itself is not amethod ofcommunication.

-Actual communication ismade possible by using

communication protocols.-A protocol implementsthe functions of one ormore of the OSI layers.

Variety of Comm. Protocols

LAN protocols: operate at thephysical and data link layers and

define communication over thevarious LAN media

- WAN protocols: operate at thelowest three layers and definecommunication over the variouswide-area media.

- Network protocols: are thevarious upper-layer protocols ina given protocol suite.

- Routing protocols: network-layer protocols responsible forpath determination and trafficswitching

ContdSDLC:-S h D t Li k C t l

PPP:-Point-to-Point Protocol


95/126

95

Synchronous Data Link Control.SDLC is a bit-oriented, full-duplexserial protocol that has spawnedor generated numerous similarprotocols, including HDLC and

LAPB.

HDLC:-High-Level Data Link Control. Bit-oriented synchronous data linklayer protocol developed by ISO.

-Specifies a data encapsulationmethod on synchronous seriallinks using frame characters andchecksums.

LAPB:-Link Access Procedure, Balanced.Data link layer protocol in theX.25 protocol stack.LAPB is a bit-oriented protocolderived from HDLC

Point to Point Protocol.Provides router-to-routerand host-to-network

X.25 PTP:-

Packet level protocol.Network layer protocol in theX.25 protocol stack. Defineshow connections aremaintained for remoteterminal accessISDN:-Integrated Services DigitalNetwork. Communication

protocol, offered bytelephone companies, thatpermits telephone networksto carry data, voiceFrame Relay:-Industry-standard, switcheddata link layer protocol thathandles multiple virtualcircuits using HDLCencapsulation betweenconnected devices.

Transmission Options orWANServices


96/126

96

-A number oftransmission optionsavailable today

POTS (Plain OldTelephone Service)

Using Modem Dialup

Analog modems using basic telephoneservice are asynchronous transmission-based, and have the following benefits:

- Available everywhere

- Easy to set up- Dial anywhere on demand- The lowest cost alternative of any wide-area service

IntegratedServices Digital Network(ISDN)


97/126

97

(ISDN)

-ISDN ,is a digital service, that cantransmit data, voice, and video overexisting copper phone lines.

-Instead of leasing a dedicated line forhigh-speed digital transmission, ISDNoffers the option of dialup connectivityincurring charges only when the line isactive

-ISDN provides a high-bandwidth, cost-effectivesolution for companies

requiring light or sporadic orperiodic high-speed access toeither a central or branchoffice.

Types of ISDN


98/126

98

Basic Rate Interface (BRI)and Primary Rate Interface(PRI).

BRI provides twoB or bearerchannels of 64 Kbps

each and oneadditional signalingchannel called the

D or delta channel

PRI provides up to23 bearer channelsof 64 Kbps each and

one D channel forsignaling.

Thats 23 channels butwith only one physicalconnection, which

makes it an elegant orneat solution- theresno wiring mess

NB: PRI can be used atyour central site if youplan to have manyISDN dial-in clients.

Leased Line


99/126

99

-Leased lines are most cost-effective if acustomers daily usage exceeds four to sixhours.

-Leased lines offer predictable throughputwith bandwidth typically 56 Kbps to 1.544Mbps.

-They require one connection per physicalinterface (namely, a synchronous serialport).

- One connection per physical interface- Bandwidth: 56 kbps1.544 Mbps- T1/E1 and fractional T1/E1- Cost effective at 46 hours daily usage- Dedicated connections with predictablethroughput- Permanent- Cost varies by distance

rame e ay-Frame Relay frames travel


100/126

100

Frame Relay provides a

standard interface to thewide-area network forbridges, routers ,LAN Devicesetc

-A Frame Relay interface isdesigned to act like a wide-area LAN- it relays ortransmits data frames directlyto their destinations at veryhigh speeds.

-Frame Relay frames travelover predetermined virtualcircuit paths, are self-routing and arrive at their

destination in the correctorder

-Frame Relay is designed tohandle the LAN-type burstytraffic efficiently.

-The guaranteed bandwidth(known as committedinformation rate or CIR) istypically between 56 Kbps

and 1.544 Mbps.

-cost is normally notdistance-sensitive

Connecting Offices with Frame Relay


101/126

101

-Companies, thatrequire office-to-officecommunications,

usually choosebetween a dedicatedleased-line connectionor a packet-basedservice, such as FrameRelay or X.25.

-As a rule, higherconnect times makeleased-line solutionsmore cost-effective.

Like ISDN, FrameRelay requires only onephysical connection tothe Frame Relaynetwork, but cansupport manyPermanent VirtualCircuits, or PVCs.

Frame Relay service is often less expensivethan leased lines, and the cost is based on:

- The committed information rate (CIR), whichcan be exceeded up to the port speed when thecapacity is available on your carriers network.

-The number of permanent virtual circuits(PVCs) you require; a benefit to users whoneed reliable and dedicated connections to

resources simultaneously.

X.25


102/126

102

-X.25 networks implement theinternationally accepted ITU-T standardgoverning the operation of packetswitching networks.

-Transmission links are used only whenneeded

-It performs error checking along eachhop from source node to destination node.The bandwidth is typically between

9.6Kbps and 64Kbps.

Digital Subscriber Line (xDSL)


103/126

103

- DSL is a pair of modems on eachend of a copper wire pair- DSL converts ordinary phone linesinto high-speed data conduits

-End-users dont buy DSL, they buyservices, such as high-speed Internetaccess, intranet, leased line,

voice, VPN, and video on demand-Service is limited to certain

geographical areas

-Digital subscriber line (DSL)technology is a high-speed servicethat, like ISDN, operates over

ordinary twisted-pair copper wiressupplying phone service to businessesand homes in most areas

-Using special modems and dedicatedequipment in the phone company'sswitching office, DSL offers fasterdata transmission than either analogmodems or ISDN service, plus-inmost cases-simultaneous voicecommunications over the same lines

DSL Modem Technology


104/126

104

gy

-DSL has several flavors.ADSL delivers

asymmetrical data rates(for example, data movesfaster on the way to yourPC than it does on theway out to Internet).

-Other DSL technologiesdeliver symmetrical data(same speed traveling inand out of your PC).

The type of service available to you willdepend on the carriers operating in your

area. Because DSL works over theexisting telephone infrastructure, itshould be easy to deploy over a widearea in a relatively short time

How to choose Service?


105/126

105

-Analog services are the least expensivetype of service. ISDN costs somewhat morebut improves performance over even thefastest current analog offerings

-Leased lines are the costliest of thesethree options, but offer dedicated, digitalservice for more demanding situations.Which is right?

Youll need to answer a few questions:

- Will employees use the Internet

frequently?- Will the Internet be used forconducting business (for example,inventory management,online catalog selling or accountinformation or bidding on new jobs)?- Do you anticipate a large volume of

traffic between branch offices of thebusiness?- Is there a plan to usevideoconferencing or video trainingbetween locations?- Who will use the main officesconnection to the Internet - individual

employees at the central office,telecommuting workers dialing in fromhome, mobile workers dialing in fromthe road?

Contd


106/126

106

Cont d

The more times the answer is yes , themore likely that leased line services arerequired. It is also possible to mix andmatch services

-For example, small branch offices orindividual employees dialing in fromhome might connect to the central officeusing ISDN, while the main connectionfrom the central office to the Internet canbe a T1.

-It is important to understand that

as the bandwidth increases, so dothe charges, both from the ISP andthe phone company. Keep in mindthat rates for different kinds ofconnections vary from location tolocation.

Compare the technology options, assuming all services areavailable


107/126

107

available

- A leased-line service provides a dedicated connection

with a fixed bandwidth at a flat rate. You pay the samemonthly fee regardless how much or how little you usethe connection.

- A packet-switched service typically provides apermanent connection with specific, guaranteedbandwidth (Frame Relay). Temporary connections

(such as X.25) may also be available. The cost of the lineis typically a flat rate, plus an additional charge basedon actual usage.

- A circuit-switched service provides a temporaryconnection with variable bandwidth, with costprimarily based on actual usage.

Wide-Area Network Requirements

b d d h


108/126

108

Minimize bandwidth costs- Maximize efficiency- Maximize performance

- Support new/emergingapplications- Maximize availability- Minimize management andmaintenance

Manage Bandwidth to Control Cost

Bandwidth-on-demandgives you the flexibility toadd additional WANbandwidth when its

needed to accommodateheavy network loads suchas file transfers

Dial-on-Demand Routing


109/126

109

-Dial-on-demand routingallows a router toautomatically initiate andclose a circuit-switchedsession

-With dial-on-demand routing,the router dials up the WANlink only when it senses

interesting traffic

-Interesting traffic might be

defined as any trafficdestined for the remotenetwork, or only trafficrelated to a specific hostaddress or service.

-Equally important, dial-on-demand routing enables therouter to take down theconnection when it is no longerneeded, ensuring that the userwill not have unnecessary WANusage charges

Bandwidth-on-Demand


110/126

110

-Bandwidth-on-demand works in asimilar way.

-When the router senses that thetraffic level on the primary link hasreached a certainsay, threshold orentrance when a user starts a largefile transferit automatically dials upadditional bandwidth through thePSTN to accommodate the increasedload.

Snapshot Routing

Snapshot routing

-this feature ensures that theremote router always has themost up-to-date routinginformation but only whenneeded.

-Without Snapshot Routing, your

ISDN connection would be dialedevery 30 seconds

IPX Protocol Spoofing


111/126

111

-Protocol spoofing allows the user toimprove performance while providingthe ability to use lower line speeds over

the WAN.

Compression

-Compression reduces thespace required to store data,thus reducing the bandwidthrequired to transmit

-The benefit of thesecompression algorithms is thatusers can utilize lower line

speeds if needed to save costs

-Compression also provides theability to move more data overa link than it would normally

bear

- Dial Backup


112/126

112

-Dial backup addresses a

customers need for reliability andguaranteed uptime

-Dial backup capability offers usersprotection against WAN downtimeby allowing them to configure abackup serial line via a circuit-

switched connection such as ISDN

-When the software detects theloss of a signal from the primary

line device or finds that the lineprotocol is down, it activates thesecondary line to establish a newsession and continue the job oftransmitting traffic over thebackup line.

Chapter 7: Understanding Routing


113/126

113

What Is Routing?

-Routing is the processof finding a path to adestination host and ofmoving informationacross an internet workfrom a source to adestination

-A router is a devicethat forwards packetsfrom one network toanother and determinesthe optimal path along

which network trafficshould be forwarded

-Routers forwardpackets from one

network to anotherbased on networklayer information

RoutersLayer 3

It determines theappropriate networkpath to send thepacket

LAN-to-LAN Connectivity


114/126

114

This illustrates the flow of packetsthrough a routed network using theexample of an e-mail message beingsent from system X to system Y

-The message exits system X andtravel through an organizationsinternal network until it gets to apoint where it needs an Internetservice provider

Read about the following


115/126

115

Path determination Multi-protocol routing Routing table

Routing Tables

-To aid the process of pathdetermination, routingalgorithms initialize and maintainrouting tables, which containroute information

-Routing Information-Include

-Destination/next hopassociations tell a routerthat a particulardestination is linked to aparticular router

representing the nexthop on the way to thefinal destination

-path desirability

-Routers compare metrics todetermine optimal routes.Metrics differ depending onthe routing algorithm used.

A metric is a standard ofmeasurement, such aspath length, that is usedby routing algorithms todetermine the optimal pathto a destination.

Routing Algorithm Goals

-Routing tables contain information used by softwareb d


116/126

116

g yto select the best route.

But how, specifically, are routing tables built?

What is the specific nature of the information theycontain?

How do routing algorithms determine that one routeis preferable to others?

Routing algorithms arebased on the followingdesign goals:

-Optimality- thecapability of the routing

algorithm to select thebest route, depending on

metrics e.g. hops anddelays

-Simplicity and lowoverhead- efficientrouting algorithmfunctionality with aminimum ofsoftware and

utilization overhead

Robustness andstabilitychecks theunforeseen

circumstances, such ashardware failures, highload conditions

Rapid convergence- Convergence is the

process ofagreement, by allrouters, on optimalroutes. When anetwork event causeschanges in routeravailability,

recalculations areneed to reestablishnetworks

Routing MetricsRouting algorithms have used many different metrics to determine the best route


117/126

117

Path length - The most common metric. The sum of either an assignedcost per network link or hop count, a metric specify the number of

passes through network devices between source and destination.

Reliability- dependability (bit-error rate) of each network link. Somenetwork links might go down more often than others. Also, some linksmay be easier or faster to repair after a failure.

Delay- The length of time required to move a packet from source to

destination through the internet work. Depends on bandwidth ofintermediate links, port queues at each router, network congestion, andphysical distance. A common and useful metric.

Bandwidth - available traffic capacity of a link.

Load- Degree to which a network resource, such as a router, is busy

(uses CPU utilization or packets processed per second).

Communication cost- operating expenses of network links (privateversus public lines).

Now lets talk a little about network addressing.

Network AddressingS b tti ll


118/126

118

Sub network Addressing

-Sub networks or subnets arenetworks arbitrarily segmented bya network administrator in orderto provide a multilevel,

hierarchical routing structurewhile shielding the sub networkfrom the addressing complexity ofattached networks

-Sub netting allowssingle routing entries torefer either to the larger

block or to its individualconstituents

-A subnet mask is a 32-bit number thatdetermines how an IPaddress is split intonetwork and hostportions, on a bitwisebasis

-For example,131.108.0.0 is astandard Class B subnetmask; the first two bytes

identify the network andthe last two bytesidentify the host.

Contd Read about


119/126

119

-A subnet mask is a 32-bitaddress mask used in IP to

indicate the bits of an IPaddress that are being used forthe subnet address.Sometimes referred to simplyas mask. The term maskderives from the fact that thenon-host portions of the IP

address bits are masked by 0sto form the subnet mask.

-Sub netting helps to organizethe network, allows rules to bedeveloped and applied to thenetwork, and provides security

and shielding

Routing Algorithms Types

Static and Dynamic Routing

Routing protocol evolutions

Required to make individualpresentations

Deadline next class !!!

Chap 8:Layer 3 Switching


120/126

120

Chap 8:Layer 3 SwitchingAssignment: Required to read the above

chapter and answer the followingquestions

Qn1. What Is Layer 3 Switching?

Qn2. What is the Difference Between

Layer 2 Switching, and Routing?

Qn3.Discuss the advantages of layer 3switches

Qn4. Why Do We Need Layer 3Switching?

Chapter 9: Understanding


121/126

121

Chapter 9: UnderstandingVirtual LANs

Guidelines for exam preparation


122/126

122

p p Chap1=Networking Basics

Chap2=OSI Reference Model Chap4 and 6

Exam Structure !!!!Sections( A= 10marks and B=70marks)=Exam Total 80%)( SectionA-Multiple choices and B-Structured

questionsAssignments=20%


123/126

123


124/126

124


125/126

125


126/126

16657572 Lecture Slides on Networking (1)

Documents