Routing Protocol

1. INTRODUCTION 1. 1 Whatis Computer lucre? The groupofcomputers and devices involvemented by communication channels all(prenominal)owing mathematical controlrs to sh atomic number 18 tuition, entropy, softw are and hardware with further users is dream upt to be computer network. Network protocols bound hardware as well as software comp geniusnts of network. Two or more(prenominal)computers are saidto beina network if and solely if they are attachedmutuallyandareableto commune. Computers are committed to a network by the use of allthe ports i. e. , parallel ports, modem ports, Ethernet ports, serial ports, USB ports, fire wire ports and many more in iodin or more way.But Ethernet port is the more or less broadly employ portsfor net work. Hosts, end stations or workstations are referred tour talkingabout networks. Anythingattachedtothe networkincluding hubs, bridges, switches, routers,access points, firewalls, workstations, servers, mainframes, printers, scanners, copi ers, fax machinesand more are included low Host or end stations . Computers are connected in a network for overlap of software and hardware resources, info and entropy as well as smooth the progress of communication. 1. 2 TCP/IP flooredarchitecture figure TCP/IP Layeredarchitecture The pursuitarethe stratumsofthe TCP/IParchitecture Application Layer In theapplication social class bare(a) Mail Transfer Protocol (SMTP) and File Transfer Protocol (FTP) uses protocolfor network communication. Application layer protocolsare roughlyfrequentlylinked with client-serverapplications. Transport Layer End-to-end message transfer capability, diminish control, misconduct control and fragmentation etc are providedby the post layer. The transport layer ensures source to goal delivery of big moneys safely and reliably.The service finished which applications are connectedtogether viathe useof ports is provided by transport layer. Network Layer Packets are crystallinely transmitted o verthe total network in the OSIs Network layer. Hosts addressing by assigningtheman IPaddressand piece of land routing among multiple networks are handled in this layer. This layer is concerned with routing selective tuition end to end message delivery etc. Inter guinea pig Layer The data exchange betweenthe serverandthe network are monitored by theinterface layer. The protocols forphysical transmissionof data is de exquisited by Interface Layer . 1. 3Autonomous SystemIP networksand routers collection belowthe controlof one entity representing a common routing insurance is called anAutonomous System. EachAS pretend a uniqueAS summatefor usein routing. Each network is uniquely identified onthe earnings by ASN. IANA (Internetassigned Numbersauthority) assign AS numbersand supplyto Regionalinternet Registries (RIRs)in blocks. Autonomous System can be dividedinto three categories MultihomedAutonomous System Connectionsto more than oneAS is maintained by a MultihomedAS. Stubautonom ous System Connectionto tho one otherAS is Stubautonomous System.Transitautonomous SystemConnections through itselfto conk out networks are provided by Transitautonomous System. 1. 4 Routing The methodof selecting caterpillar tracksina network via whichto channelise data is meant to be routing. The turnof findinga racecourseway froma sendertoa in demand(p) end is also said to be routing. The telephone network,theinternetand transport networks, etc perform routing. Network Layerof either TCP/IP layered model orthe OSI (Open Systeminterconnect) Reference model primarily carry out routing. The logicallyaddressed packets are passed fromtheir sourceto destination viaintermediary nodes i. e. orwarding is directed by routing. Routing tasks are performed by routers. Routing and packet forwarding is performed by ordinarycomputers available with multiple network card in a limited manner. Forwarding is directed by the routing process onthe basisof routing hold overs where routing remember to different network destinations are maintained. In revision to have efficient routing, construction of routing table heldinthe routers repositing is most necessary thing. Only one network thoroughfare are frequently used by routing algorithmic programs ata time, butthe useof multiplealternative paths is made possible by multi-path routing techniques.Following are the facesof routing delivery semantics Unicast A message is delivered toa wizard specified node by router. Fig Unicasting Broadcast A message is deliveredtoall nodesinthe network by router. Fig beam Multicast A message is deliveredto assemblyof nodes that have expressedinte detainin gettingthe message by router. Fig Multicasting Anycast A message is deliveredtoany one outofa setof nodes, typicallythe one nexttothe source. Figanycasting 2. TYPESOF ROUTING Following are the typesof Routing mechanisms. Theyare Static Routing Dynamic Routing 2. Static Routing The processby which routes can be manually entered into the routing table with the inspection and repair of a configuration file which institutionalizes automatically as soon as router starts is called smooth routing. Networkadministrator, who configures the routes, can enter these routes as an option. Thus static routes mean the routes that can non be changed (excepta person changesthem) after their configuration. The simplesttypeof routing is static routing. In end of change of routing tuition often or configuration on a huge number of routing devices (router) it doesnt work fine as it is a manual process.The outages or down connections are not handled properly by static routing becausemanually configured route must be reconfigured physically in orderto fix or renovateany lost connectivity. 2. 2 Dynamic Routing Network destinations are discovered dynamicallyby promoter of softwareapplications called Dynamic routing protocols. A routing table is created and managed by routerin Dynamic Routing. Firstly, a router get out lea rn routestothe directly connected entire networks. It willthen learn routes from other routers using the same routing protocol. ane or more ruff routes are selected from the list of routes for each(prenominal) and all(prenominal) network destination by router. Best routein stageion are distributedto other routers runningthe same routing protocol by Dynamic protocols, distributingthe teaching on what networks it subsistand can be reached. This provide dynamic routing protocolsthecapabilitytoget used to logical networktopology changes, equipment failures or network outages onthe fly. 2. 3 partsof Dynamic Routing withdrawnness- sender Routing Paths are calculated using Bellman Ford Algorithm bya distance-transmitter routing protocol.RIPv1and 2and IGRP (Interior Gateway Routing Protocol) are examplesof distance-vector routing protocols. Earlier, distance vector protocols such as RIPv1 show classful behavior but newer distance vector protocols suchas RIPv2and enhanceinterior Gatewa y Routing Protocol (EIGRP) show signs of classless behavior. Distance-vector routing protocols Easyand competentin small networks divest convergence properties Facilitate inthe growthof more complex but more scalable link-state routing protocolsfor usein large networks. Periodic copiesofa routing table are passed from routerto router by distance vector routingalgorithms. reproducible broadcast is the most commonly usedaddressing scheme. Periodic updates are sent by routers runninga distance vector routing protocol even ifthereare no changesinthe network. Complete routing table is included underthe periodic routing update in a pure distance vector environment. all in all know routes can be verified and changes can be madeby gettinga neighbors complete routing table based on simplifiedinformation also called as routing by rumor. Fig Distance Vector RoutingPeriodic routing updates are received from router A to router B inthe figure. Distance vector metric (suchas hop count) ar e added by Router B to each route learned from router A,risingthe distance vector. Its own routing tablesare passed to its neighbor, router C. This process occursbetween directly connected neighbor routers inall directions. The chief purposeisto decidethe top routetocontaininthe table when the routing table is updated bya routing protocolalgorithm. Different routing metric is used to determinethe beat route by each distance vector routing protocol. calculated valueis generated for each path through network by thealgorithm. Usually, the path is better if metric is smaller. Single characteristicofa path helps in calculation of metrics and combination of several path characteristics helps in calculation of more complex metrics. The most commonly usedmetrics used by distance vector routing protocols are Hop press Packets numberof passages throughoutthe output portof one router Bandwidth Links data capability Delay Time necessaryto shifta packet from starting placeto destination. stre tch along work load onrouter or link. Reliability each network linkbit error rate Maximum Transmission Unit (MTU)the utmost message extentin octets satisfactorytoall links onthe path. Link-State Routing Packet-switched networks use link-state routing protocolfor computer communications. OSPFandIS-IS are its examples. Atopological database is built by the help of link-state routing that describes extrapreciseinter-network routes. Large networks use link state routing protocols and now used by most of the organization and ISP.Router performs the link-state protocol inthe network. A mapofthe connectivityofthe network is constructed by every node in the form of graph showing node connection to other node is the basic conceptof link-state routing. The best next hop is calculated by each nodeindependently for every possible destinationinthe network. The routing table for the node is formed bythe collectionof best next hops. Fig Link-State Routing To find outthe shortest path from itselfto every other nodeinthe network analgorithm is run by each nodeindependently overthe map.OSPF, EIGRP and Novells NLSP (NetWare Link State Protocol) are the examples of link state routing protocol. IPX is only supported by Novells NLSP. A partial mapofthe network is maintained by each router in this typeof routing protocol. Link stateadvertisement (LSA)is flooded throughoutthe network whena network link changes state (upto down, or vice versa). The changes are noted and routes are re-computed by allthe routersaccordingly. Greater supplenessand sophistication are provided by Link State Routing protocols thanthe Distance Vector routing protocols.Overall broadcast calling is reducedand better finales are madeabout routing by taking characteristics suchas bandwidth, delay, reliability,and loadinto consideration,insteadof takingtheir decisions only on hop count. 3. ROUTINGALGORITHMS 3. 1 Bellman-FordAlgorithm Also called as chase after Correctingalgorithm Used for negative edge weigh t Same as Dijkstrasalgorithm In order to maintain distance tables, this algorithm is used by router Exchanginginformation withthe neighboring nodes help to update information in the distance table All nodesin the network is represented by the numberof datainthe table The directlyattached neighbors are represented by the columnsof table and all destinationsinthe network are represented by the row. The numberof hops, latency,the numberof outgoing packets, etc. are measurements in this algorithm. 3. 2 DijkstrasAlgorithm Edsger Dijkstraconceived Dijkstrasalgorithm virtuallyly used for routing Is a graph search algorithm The single-source shortest path problemfora graph is solved by this algorithm with non negative edge path monetary values The shortest path tree is produced as a output Helps in finding shortest route from one router to other A shortest-path spanning tree having route to all possible destinationis built by this algorithm for router The router usingthealgorit hmisthe sourceof its shortest-path spanning tree 4. ROUTING PROTOCOLS Routing protocol describe the way of communication between routers which helps in the selection of routes between any two nodes on a network. Usually, knowledge of immediate neighbors is known by each router. Thisinformation is shared bya routing protocol to have routers the knowledgeofthe networktopology. Most commonly used Routing protocols are as follows 4. RIP (Routinginformation Protocol) dynamicinter-network routing protocol used in private network routes are automatically discovered routing tables are built a Distance-Vector routing protocol uses Bellman-Fordalgorithm 15 hops areallowed with RIP 180 sec is the hold down time broad updates are transmitted every 30 sec by each RIP router Works at network layer Prevent routing loops Hop limit incorrect routinginformation are bared from cosmos propagated palmy configuration no parameter required Two versionsof RIP are as follows RIPv1 classf ul routing is used subnet information is not carried by periodic routing updates no support for VLSM (variable length subnet masks) Same network class have different sized subnet by the use of RIPv1 No router credentials Broadcast based and 15 is the supreme hop count A RIPv1 packetformatis shown below picFig RIP packetformat wantdetermine whetherthe packetisa call for ora response. A router sendall or partof its routing table is asked bythe request. Replytoa request or regular routing update means the response. Routing table entries are contained in responses. Version number RIP version used is specified.Potentially repugnant versions can be signaled by this field. Zero RFC 1058 RIP doesnt use this field it wasadded to have retrograde compatibility provided to pre-standard varietiesof RIP. Address family identifier (AFI) Theaddress family used is specified. Address-family identifier is contained ineach gatetospecifythe categoryofaddress being particularized. TheAFIfor IPi s 2. Address The IPaddress is particularizedforthe entry. MetricThe number of inter-network hops traversedinthe trigger offtothe destination is indicated. 1and 15foran applicable route, or 16foran unapproachable route. RIPv2 Developedin 1994 classlessinter-Domain Routing (CIDR) is supported Subnetinformation can be carried Addition of MD5 authentication and Rudimentary plain textauthentication for the security of routing updates. Routing updates are multicast to 224. 0. 0. 9 15 is the maximum hop count A RIPv2 packetformat is shown below pic Fig RIPv2 packetformat Commanddetermine whetherthe packetisa request ora response. A router sendall or partof its routing table is asked bythe request. Replytoa request or regular routing update means the response. Routing table entries are contained in responses.Version number RIP version used is specified. Unused Zero is the value set. Address-family identifier (AFI)Theaddress family used is specified. Authenticationinformation is conta ined in the remainder of the entry iftheAFIforthe initial entryis 0xFFFF inthe message. At present,simple password is the onlyauthentication type. Route tag The methodology is providedfor distinguishing betweeninternal routes (learned by RIP)and external routes (learned from other protocols). IPaddress IPaddress is particularizedforthe entry. Subnet maskThe subnet mask is containedforthe entry.No subnet mask has been particularizedforthe entry if this fieldis zero. Next hop The IPaddressofthe next hop is indicatedto which packetsforthe entry should beforwarded. MetricThe number of inter-network hops traversedinthe triptothe destination is indicated. 1and 15foran applicable route, or 16foran unapproachable route. 4. 2 OSPF (Open Shortest Path First) A Link-State protocol usedfor routing between routers belongingtoa singleautonomous system link-state technology is used informationaboutthe direct connectionsand links is communicated between the routers kindred database is maintain ed by each OSPF router for the description of theautonomous Systemstopology Calculation of a routing table by the construction of a shortest- path tree from this database. Routes are promptly recalculated in the face of topological changes equal-cost multi-path are supported Authentication of all OSPF routing protocol exchanges Designed for TCP/IP environment routing updates authentication IP multicast are utilized in sending/receivingthe updates routes IP packets based exclusively onthe commit IPaddress originateinthe IP packet header Grouping of sets of networks IP subnets are flexibly configured Destinationand mask is available to the route distributed by OSPF The hobby figure showsthe packetformat used by OSPF picFig OSPF packetformat Version numberthe OSPF version used is specified. Typethe OSPF packet type is identifiedas oneofthe following Hello neighbor relationships are set up and maintained. Database descriptionthe contentsofthetopological database are descri bed. Link-state request piecesofthetopological database are requested from neighbor routers.Link-state updatea link-state request packet is responded. Link-stateacknowledgment link-state update packets are acknowledged. Packet lengththe packet length,the OSPF header is specified. Router ID the sourceofthe packet is identified. Area ID Thearea of packet is identified. All OSPF packetsarelinked witha singlearea. Checksumthe complete packet contents are checkedforany harm sufferedin travel. Authentication typetheauthentication type is contained. Authentication ofall OSPF protocol exchanges. Configuration of theauthentication type on per-area basis.Authentication authenticationinformation is contained. Data encapsulated upper-layerinformation is contained. 5. WORKING 5. 1 Distance Vector Routing The following methods showthe boilers suit workingofthe Distance-Vector Routing . There is no pre define route i. e. entire route for a particular destination is not known to any router. The po rt to send out a unicast packet is known by each router on the basis of destination address. Progressively the route is made and there is the formation of the route by the theatrical role of each router when it receives the packet.The optimum tree is not predefined in DVRP actually. No routers have knowledge for making an optimal tree. Slowly and gradually the tree is made. The tree is formed as soon as a router receives a packet it is forwarded by router through some of the ports, on the basis of source address. Other down-stream routers make the rest of the tree. The formation of the loops must be prevented by this protocol. Duplications are also prevented in order to make the entire network receive only one copy. In addition to this, the shortest path from source to the destination is the path travelled by a copy.Inconsistencies occurring with Distance-Vector Routing Incorrect routing entries are caused by slowinter-network convergence which may meet inconsistencies maintaining routing information. .The following example describes howinconsistencies occurin Distance-Vector routing The entire figure describes the inconsistencies occurring with Distance-Vector Routing. Defininga maximumto prevent countto timeless existence . With thisapproach,the routing table update loop is permitted by routing protocol untilthe metric exceeds its maximumallowed value. Fig Defininga maximumto prevent counttoinfinity 6 hops are defined as the maximumallowed value. Whenthe metric value exceeds 16 hops, we cannot reach network 10. 4. 0. 0 Routing Loopsin Distance-Vector Routing A routing loop is said to be occurred if two or more routers havefalse routinginformationrepresenting thata applicable pathtoan unapproachable destination exists via other routers. Fig Routing Loop Solutionsto eliminate routing loops Split vistaThe information is not sent in the direction from where original information comes. The give out horizon function is illustrated by the following figureFig Sp lit Horizon Route PoisoningRouting loops are eliminated. The following figure providesan exampleof Route Poisoning Fig Route Poisoning Inadditionto split horizon, route poisoningand holddown timers, poison reverse, holddown timersand triggered updatesare other methodsto eliminate routing loops. 5. 2 Link-State Routing The following methods showthe overall workingof Link-State Routing. Gathering of the neighborinformation continuously. Router answering to this protocol are broadcasted the list of neighborinformation, process knownas flooding.Soon, thisinformation is distributed to all routers onthe network. Flooding of the neighborinformation in caseofa (routing-significant) changeinthe network. The best path can be calculated to any host on any destination network as everythingaboutthe network is known by every router. 6. ADVANTAGESAND DISADVANTAGES Distance-Vector Routing Advantagesof Distance-Vector Routing simpleand flat network No special hierarchical design is required. Impl ementation of hub-and-spoke networks No concern for worst-case convergence timesina network less memoryand processing authority functionDisadvantagesof Distance-Vector Routing Incorrect routing entries create inconsistencies in maintainingthe routinginformation Rise of a condition counttoinfinity Occurrence of a routing loop Variable Length Subnet Masking (VLSM) or super netting is not supported multi-vendor routing environment is not supported Link-State Routing Advantagesof Link-State Routing Paths are chosen via network by the use of cost metrics changesinthe networktopology are reported toall routersinthe network quickly fast convergence times No occurrence of routing loops routing decisions are based on the most new-fashioned setofinformation Link-State protocols use cost metricsto choose paths thoughthe network. The cost metric reflectsthe capacityofthe links on those paths. Disadvantagesof Link-State Routing Topology database,anadjacency database,andaforwardin g database is required. a significantamountof memoryis required in large or complex networks significantamountof mainframe power usage need of a strict hierarchical network design to reduce significantamountof CPU power usage network capability or performance is low to transport data . occupationAREAS Distance-Vector Routing used in mobile, wireless and hoc networks (MANETs) used for mobilead hoc routing (Ad hoc On-Demand Distance Vector Routing) . Link-State Routing usedin larger, more complicated networks Optimized Link State Routing Protocol (OLSR) designed for mobile, wireless and hoc networks 8. COMPARING DISTANCE-VECTORAND LINK-STATE ROUTING STRATEGIES Mostly, best path is determined by Distance Vector protocols, while bandwidth, delay, reliabilityand load are considered to make routing decision by Link-State protocols Distance Vector protocols are simple and efficient where as Link-State protocols are flexible and advanced Routinginformation Protocol (RIP v1and v2 )andinterior Gateway Routing Protocol (IGRP developed by Cisco) are Distance Vector protocols where as OSPF, EIGRP, Novells NLSP (NetWare Link State Protocol) are Link-State protocols Notion of a distance is not required in Distance Vector routing where as Link-State routing is based on minimizing some notion of distance Uniform policies are not required at all routers in Distance Vector routing but uniform policy is required in Link-State routing Router have little knowledge about network topology in Distance Vector routing where as routing domain has excessive knowledge about topology information in Link-State routing 9. coating Introduction, working, use, advantages and disadvantages of Distance-Vectorand Link-State routingare explainedin this project. Bellmanfordand Dijkstrasalgorithm are also discussed. This project describes the popularity of Distance-Vectorand Link-State routingbecause of their complex, sophisticated, flexible features in recent computer networking field..