COMPUTER NETWORKING NETWORK LAYER

Network Layer4-1Chapter 4Network LayerA note on the use of these ppt slides:We’re making these slides freely available to all (faculty, students, readers). They’re in PowerPoint form so you can add, modify, and delete slides (including this one) and slide content to suit your needs. They obviously represent a lotof work on our part. In return for use, we only ask the following:If you use these slides (e.g., in a class) in substantially unaltered form, that you mention their source (after all, we’d like people to use our book!)If you post any slides in substantially unaltered form on a www site, that you note that they are adapted from (or perhaps identical to) our slides, and note our copyright of this material.Thanks and enjoy! JFK/KWRAll material copyright 1996-2007J.F Kurose and K.W. Ross, All Rights ReservedComputer Networking: A Top Down Approach 4thedition. Jim Kurose, Keith RossAddison-Wesley, July 2007. Network Layer4-2Chapter 4: Network LayerChapter goals:understand principles behind network layer services:network layer service modelsforwarding versus routinghow a router worksrouting (path selection)dealing with scaleadvanced topics: IPv6, mobilityinstantiation, implementation in the InternetNetwork Layer4-3Chapter 4: Network Layer4. 1 Introduction4.2 Virtual circuit and datagram networks4.3 What‟s inside a router4.4 IP: Internet ProtocolDatagram formatIPv4 addressingICMPIPv64.5 Routing algorithmsLink stateDistance VectorHierarchical routing4.6 Routing in the InternetRIPOSPFBGP4.7 Broadcast and multicast routingNetwork Layer4-4Network layertransport segment from sending to receiving host on sending side encapsulates segments into datagramson rcving side, delivers segments to transport layernetwork layer protocols in everyhost, routerrouter examines header fields in all IP datagrams passing through itapplicationtransportnetworkdata linkphysicalapplicationtransportnetworkdata linkphysicalnetworkdata linkphysicalnetworkdata linkphysicalnetworkdata linkphysicalnetworkdata linkphysicalnetworkdata linkphysicalnetworkdata linkphysicalnetworkdata linkphysicalnetworkdata linkphysicalnetworkdata linkphysicalnetworkdata linkphysicalnetworkdata linkphysicalNetwork Layer4-5Two Key Network-Layer Functionsforwarding:move packets from router‟s input to appropriate router outputrouting:determine route taken by packets from source to dest. routing algorithmsanalogy:routing:process of planning trip from source to destforwarding:process of getting through single interchangeNetwork Layer4-61230111value in arrivingpacket’s headerrouting algorithmlocal forwarding tableheader valueoutput link01000101011110013221Interplay between routing and forwardingNetwork Layer4-7Connection setup3rdimportant function in somenetwork architectures:ATM, frame relay, X.25before datagrams flow, two end hosts andintervening routers establish virtual connectionrouters get involvednetwork vs transport layer connection service:network:between two hosts (may also involve inervening routers in case of VCs)transport:between two processesNetwork Layer4-8Network service modelQ:What service modelfor “channel” transporting datagrams from sender to receiver?Example services for individual datagrams:guaranteed deliveryguaranteed delivery with less than 40 msec delayExample services for a flow of datagrams:in-order datagram deliveryguaranteed minimum bandwidth to flowrestrictions on changes in inter-packet spacingNetwork Layer4-9Network layer service models:NetworkArchitectureInternetATMATMATMATMServiceModelbest effortCBRVBRABRUBRBandwidthnoneconstantrateguaranteedrateguaranteed minimumnoneLossnoyesyesnonoOrdernoyesyesyesyesTimingnoyesyesnonoCongestionfeedbackno (inferredvia loss)nocongestionnocongestionyesnoGuarantees ?Network Layer4-10Chapter 4: Network Layer4. 1 Introduction4.2 Virtual circuit and datagram networks4.3 What‟s inside a router4.4 IP: Internet ProtocolDatagram formatIPv4 addressingICMPIPv64.5 Routing algorithmsLink stateDistance VectorHierarchical routing4.6 Routing in the InternetRIPOSPFBGP4.7 Broadcast and multicast routingNetwork Layer4-11Network layer connection and connection-less servicedatagram network provides network-layer connectionless serviceVC network provides network-layer connection serviceanalogous to the transport-layer services, but:service: host-to-hostno choice: network provides one or the otherimplementation: in network coreNetwork Layer4-12Virtual circuitscall setup, teardown for each call beforedata can floweach packet carries VC identifier (not destination host address)everyrouter on source-dest path maintains “state” for each passing connectionlink, router resources (bandwidth, buffers) may be allocated to VC (dedicated resources = predictable service)“source-to-dest path behaves much like telephone circuit”performance-wisenetwork actions along source-to-dest pathNetwork Layer4-13VC implementationa VC consists of:1.path from source to destination2.VC numbers, one number for each link along path3.entries in forwarding tables in routers along pathpacket belonging to VC carries VC number (rather than dest address)VC number can be changed on each link.New VC number comes from forwarding tableNetwork Layer4-14Forwarding table122232123VC numberinterfacenumberIncoming interface Incoming VC # Outgoing interface Outgoing VC #1 12 3 222 63 1 18 3 7 2 171 97 3 87… … … …Forwarding table innorthwest router:Routers maintain connection state information!Network Layer4-15Virtual circuits: signaling protocolsused to setup, maintain teardown VCused in ATM, frame-relay, X.25not used in today‟s Internetapplicationtransportnetworkdata linkphysicalapplicationtransportnetworkdata linkphysical1. Initiate call2. incoming call3. Accept call4. Call connected5. Data flow begins6. Receive dataNetwork Layer4-16Datagram networksno call setup at network layerrouters: no state about end-to-end connectionsno network-level concept of “connection”packets forwarded using destination host addresspackets between same source-dest pair may take different pathsapplicationtransportnetworkdata linkphysicalapplicationtransportnetworkdata linkphysical1. Send data2. Receive dataNetwork Layer4-17Forwarding tableDestinationAddressRangeLinkInterface11001000000101110001000000000000through0110010000001011100010111111111111100100000010110001100000000000through11100100000010111000110001111111111001000000101110001100100000000through21100100000010111000111111111111otherwise34 billion possible entriesNetwork Layer4-18Longest prefix matchingPrefixMatchLinkInterface110010000001011100010011001000000101110001100011100100000010111000112otherwise3DA: 11001000 00010111 00011000 10101010 ExamplesDA: 11001000 00010111 00010110 10100001 Which interface?Which interface?Network Layer4-19Datagram or VC network: why?Internet (datagram)data exchange among computers“elastic” service, no strict timing req. “smart” end systems (computers)can adapt, perform control, error recoverysimple inside network, complexity at “edge”many link types different characteristicsuniform service difficultATM (VC)evolved from telephonyhuman conversation: strict timing, reliability requirementsneed for guaranteed service“dumb” end systemstelephonescomplexity inside networkNetwork Layer4-20Chapter 4: Network Layer4. 1 Introduction4.2 Virtual circuit and datagram networks4.3 What‟s inside a router4.4 IP: Internet ProtocolDatagram formatIPv4 addressingICMPIPv64.5 Routing algorithmsLink stateDistance VectorHierarchical routing4.6 Routing in the InternetRIPOSPFBGP4.7 Broadcast and multicast routingNetwork Layer4-21Router Architecture OverviewTwo key router functions:run routing algorithms/protocol (RIP, OSPF, BGP)forwarding datagrams from incoming to outgoing linkNetwork Layer4-22Input Port FunctionsDecentralized switching:given datagram dest., lookup output port using forwarding table in input port memorygoal: complete input port processing at „line speed‟queuing: if datagrams arrive faster than forwarding rate into switch fabricPhysical layer:bit-level receptionData link layer:e.g., Ethernetsee chapter 5Network Layer4-23Three types of switching fabricsNetwork Layer4-24Switching Via MemoryFirst generation routers:traditional computers with switching under direct control of CPUpacket copied to system‟s memoryspeed limited by memory bandwidth (2 bus crossings per datagram)InputPortOutputPortMemorySystem BusNetwork Layer4-25Switching Via a Busdatagram from input port memoryto output port memory via a shared busbus contention:switching speed limited by bus bandwidth32 Gbps bus, Cisco 5600: sufficient speed for access and enterprise routersNetwork Layer4-26Switching Via An Interconnection Networkovercome bus bandwidth limitationsBanyan networks, other interconnection nets initially developed to connect processors in multiprocessoradvanced design: fragmenting datagram into fixed length cells, switch cells through the fabric. Cisco 12000: switches 60 Gbps through the interconnection networkNetwork Layer4-27Output PortsBufferingrequired when datagrams arrive from fabric faster than the transmission rateScheduling disciplinechooses among queued datagrams for transmissionNetwork Layer4-28Output port queueingbuffering when arrival rate via switch exceeds output line speedqueueing (delay) and loss due to output port buffer overflow!Network Layer4-29How much buffering?RFC 3439 rule of thumb: average buffering equal to “typical” RTT (say 250 msec) times link capacity Ce.g., C = 10 Gps link: 2.5 Gbit bufferRecent recommendation: with Nflows, buffering equal to RTT C.NNetwork Layer4-30Input Port QueuingFabric slower than input ports combined -> queueing may occur at input queues Head-of-the-Line (HOL) blocking:queued datagram at front of queue prevents others in queue from moving forwardqueueing delay and loss due to input buffer overflow!Network Layer4-31Chapter 4: Network Layer4. 1 Introduction4.2 Virtual circuit and datagram networks4.3 What‟s inside a router4.4 IP: Internet ProtocolDatagram formatIPv4 addressingICMPIPv64.5 Routing algorithmsLink stateDistance VectorHierarchical routing4.6 Routing in the InternetRIPOSPFBGP4.7 Broadcast and multicast routingNetwork Layer4-32The Internet Network layerforwardingtableHost, router network layer functions:Routing protocols•path selection•RIP, OSPF, BGPIP protocol•addressing conventions•datagram format•packet handling conventionsICMP protocol•error reporting•router “signaling”Transport layer: TCP, UDPLink layerphysical layerNetworklayerNetwork Layer4-33Chapter 4: Network Layer4. 1 Introduction4.2 Virtual circuit and datagram networks4.3 What‟s inside a router4.4 IP: Internet ProtocolDatagram formatIPv4 addressingICMPIPv64.5 Routing algorithmsLink stateDistance VectorHierarchical routing4.6 Routing in the InternetRIPOSPFBGP4.7 Broadcast and multicast routingNetwork Layer4-34IP datagram formatverlength32 bitsdata (variable length,typically a TCP or UDP segment)16-bit identifierheaderchecksumtime tolive32 bit source IP addressIP protocol versionnumberheader length(bytes)max numberremaining hops(decremented at each router)forfragmentation/reassemblytotal datagramlength (bytes)upper layer protocolto deliver payload tohead.lentype ofservice“type” of data flgsfragmentoffsetupperlayer32 bit destination IP addressOptions (if any)E.g. timestamp,record routetaken, specifylist of routers to visit.how much overhead with TCP?20 bytes of TCP20 bytes of IP= 40 bytes + app layer overheadNetwork Layer4-35IP Fragmentation & Reassemblynetwork links have MTU (max.transfer size) -largest possible link-level frame.different link types, different MTUs large IP datagram divided (“fragmented”) within netone datagram becomes several datagrams“reassembled” only at final destinationIP header bits used to identify, order related fragmentsfragmentation: in:one large datagramout:3 smaller datagramsreassemblyNetwork Layer4-36IP Fragmentation and ReassemblyID=xoffset=0fragflag=0length=4000ID=xoffset=0fragflag=1length=1500ID=xoffset=185fragflag=1length=1500ID=xoffset=370fragflag=0length=1040One large datagram becomesseveral smaller datagramsExample4000 byte datagramMTU = 1500 bytes1480 bytes in data fieldoffset =1480/8 Network Layer4-37Chapter 4: Network Layer4. 1 Introduction4.2 Virtual circuit and datagram networks4.3 What‟s inside a router4.4 IP: Internet ProtocolDatagram formatIPv4 addressingICMPIPv64.5 Routing algorithmsLink stateDistance VectorHierarchical routing4.6 Routing in the InternetRIPOSPFBGP4.7 Broadcast and multicast routingNetwork Layer4-38IP Addressing: introductionIP address:32-bit identifier for host, router interfaceinterface:connection between host/router and physical linkrouter‟s typically have multiple interfaceshost typically has one interfaceIP addresses associated with each interface223.1.1.1223.1.1.2223.1.1.3223.1.1.4223.1.2.9223.1.2.2223.1.2.1223.1.3.2223.1.3.1223.1.3.27223.1.1.1 = 11011111 00000001 00000001 00000001223111Network Layer4-39SubnetsIP address:subnet part (high order bits)host part (low order bits) What‟s a subnet ?device interfaces with same subnet part of IP addresscan physically reach each other without intervening router223.1.1.1223.1.1.2223.1.1.3223.1.1.4223.1.2.9223.1.2.2223.1.2.1223.1.3.2223.1.3.1223.1.3.27network consisting of 3 subnetssubnetNetwork Layer4-40Subnets223.1.1.0/24223.1.2.0/24223.1.3.0/24RecipeTo determine the subnets, detach each interface from its host or router, creating islands of isolated networks. Each isolated network is called a subnet.Subnet mask: /24Network Layer4-41SubnetsHow many?223.1.1.1223.1.1.3223.1.1.4223.1.2.2223.1.2.1223.1.2.6223.1.3.2223.1.3.1223.1.3.27223.1.1.2223.1.7.0223.1.7.1223.1.8.0223.1.8.1223.1.9.1223.1.9.2Network Layer4-42IP addressing: CIDRCIDR:Classless InterDomain Routingsubnet portion of address of arbitrary lengthaddress format: a.b.c.d/x, where x is # bits in subnet portion of address11001000 0001011100010000 00000000subnetparthostpart200.23.16.0/23Network Layer4-43IP addresses: how to get one?Q:How does hostget IP address?hard-coded by system admin in a fileWintel: control-panel->network->configuration->tcp/ip->propertiesUNIX: /etc/rc.configDHCP:Dynamic Host Configuration Protocol: dynamically get address from as server“plug-and-play” Network Layer4-44DHCP: Dynamic Host Configuration ProtocolGoal:allow host to dynamically obtain its IP address from network server when it joins networkCan renew its lease on address in useAllows reuse of addresses (only hold address while connected an “on”Support for mobile users who want to join network (more shortly)DHCP overview:host broadcasts “DHCP discover” msgDHCP server responds with “DHCP offer” msghost requests IP address: “DHCP request” msgDHCP server sends address: “DHCP ack” msg Network Layer4-45DHCP client-server scenario223.1.1.1223.1.1.2223.1.1.3223.1.1.4223.1.2.9223.1.2.2223.1.2.1223.1.3.2223.1.3.1223.1.3.27ABEDHCP serverarriving DHCP clientneedsaddress in thisnetworkNetwork Layer4-46DHCP client-server scenarioDHCP server: 223.1.2.5arrivingclienttimeDHCP discoversrc : 0.0.0.0, 68 dest.: 255.255.255.255,67yiaddr: 0.0.0.0transaction ID: 654DHCP offersrc: 223.1.2.5, 67 dest: 255.255.255.255, 68yiaddrr: 223.1.2.4transaction ID: 654Lifetime: 3600 secsDHCP requestsrc: 0.0.0.0, 68 dest:: 255.255.255.255, 67yiaddrr: 223.1.2.4transaction ID: 655Lifetime: 3600 secsDHCP ACKsrc: 223.1.2.5, 67 dest: 255.255.255.255, 68yiaddrr: 223.1.2.4transaction ID: 655Lifetime: 3600 secsNetwork Layer4-47IP addresses: how to get one?Q:How does networkget subnet part of IP addr?A:gets allocated portion of its provider ISP‟s address spaceISP's block 11001000 00010111 00010000 00000000 200.23.16.0/20 Organization 0 11001000 00010111 00010000 00000000 200.23.16.0/23 Organization 1 11001000 00010111 00010010 00000000 200.23.18.0/23 Organization 2 11001000 00010111 00010100 00000000 200.23.20.0/23 ... ….. …. ….Organization 7 11001000 00010111 00011110 00000000 200.23.30.0/23Network Layer4-48Hierarchical addressing: route aggregation“Send me anythingwith addresses beginning 200.23.16.0/20”200.23.16.0/23200.23.18.0/23200.23.30.0/23Fly-By-Night-ISPOrganization 0Organization 7InternetOrganization 1ISPs-R-Us“Send me anythingwith addresses beginning 199.31.0.0/16”200.23.20.0/23Organization 2......Hierarchical addressing allows efficient advertisement of routing information:Network Layer4-49Hierarchical addressing: more specific routesISPs-R-Us has a more specific route to Organization 1“Send me anythingwith addresses beginning 200.23.16.0/20”200.23.16.0/23200.23.18.0/23200.23.30.0/23Fly-By-Night-ISPOrganization 0Organization 7InternetOrganization 1ISPs-R-Us“Send me anythingwith addresses beginning 199.31.0.0/16or 200.23.18.0/23”200.23.20.0/23Organization 2......Network Layer4-50IP addressing: the last word...Q:How does an ISP get block of addresses?A:ICANN: Internet Corporation for Assigned Names and Numbersallocates addressesmanages DNSassigns domain names, resolves disputesNetwork Layer4-51NAT: Network Address Translation10.0.0.110.0.0.210.0.0.310.0.0.4138.76.29.7local network(e.g., home network)10.0.0/24rest ofInternetDatagrams with source or destination in this networkhave 10.0.0/24 address for source, destination (as usual)Alldatagrams leavinglocalnetwork have samesingle source NAT IP address: 138.76.29.7,different source port numbersNetwork Layer4-52NAT: Network Address TranslationMotivation:local network uses just one IP address as far as outside world is concerned:range of addresses not needed from ISP: just one IP address for all devicescan change addresses of devices in local network without notifying outside worldcan change ISP without changing addresses of devices in local networkdevices inside local net not explicitly addressable, visible by outside world (a security plus).Network Layer4-53NAT: Network Address TranslationImplementation:NAT router must:outgoing datagrams:replace(source IP address, port #) of every outgoing datagram to (NAT IP address, new port #). . . remote clients/servers will respond using (NAT IP address, new port #) as destination addr.remember (in NAT translation table) every (source IP address, port #) to (NAT IP address, new port #) translation pairincoming datagrams:replace(NAT IP address, new port #) in dest fields of every incoming datagram with corresponding (source IP address, port #) stored in NAT tableNetwork Layer4-54NAT: Network Address Translation10.0.0.110.0.0.210.0.0.3S: 10.0.0.1, 3345D: 128.119.40.186, 80110.0.0.4138.76.29.71:host 10.0.0.1 sends datagram to 128.119.40.186, 80NAT translation tableWAN side addr LAN side addr138.76.29.7, 5001 10.0.0.1, 3345…… ……S: 128.119.40.186, 80 D: 10.0.0.1, 33454S: 138.76.29.7, 5001D: 128.119.40.186, 8022:NAT routerchanges datagramsource addr from10.0.0.1, 3345 to138.76.29.7, 5001,updates tableS: 128.119.40.186, 80 D: 138.76.29.7, 500133:Reply arrivesdest. address:138.76.29.7, 50014:NAT routerchanges datagramdest addr from138.76.29.7, 5001 to 10.0.0.1, 3345Network Layer4-55NAT: Network Address Translation16-bit port-number field: 60,000 simultaneous connections with a single LAN-side address!NAT is controversial:routers should only process up to layer 3violates end-to-end argument•NAT possibility must be taken into account by app designers, eg, P2P applicationsaddress shortage should instead be solved by IPv6Network Layer4-56NAT traversal problemclient want to connect to server with address 10.0.0.1server address 10.0.0.1 local to LAN (client can‟t use it as destination addr)only one externally visible NATted address: 138.76.29.7solution 1: statically configure NAT to forward incoming connection requests at given port to servere.g., (123.76.29.7, port 2500) always forwarded to 10.0.0.1 port 2500010.0.0.110.0.0.4NAT router138.76.29.7Client?Network Layer4-57NAT traversal problemsolution 2: Universal Plug and Play (UPnP) Internet Gateway Device (IGD) Protocol. Allows NATted host to:learn public IP address (138.76.29.7)enumerate existing port mappingsadd/remove port mappings (with lease times)i.e., automate static NAT port map configuration10.0.0.110.0.0.4NAT router138.76.29.7IGDNetwork Layer4-58NAT traversal problemsolution 3: relaying (used in Skype)NATed server establishes connection to relayExternal client connects to relayrelay bridges packets between to connections10.0.0.1NAT router138.76.29.7Client1.connection torelay initiatedby NATted host2.connection torelay initiatedby client3.relaying establishedNetwork Layer4-59Chapter 4: Network Layer4. 1 Introduction4.2 Virtual circuit and datagram networks4.3 What‟s inside a router4.4 IP: Internet ProtocolDatagram formatIPv4 addressingICMPIPv64.5 Routing algorithmsLink stateDistance VectorHierarchical routing4.6 Routing in the InternetRIPOSPFBGP4.7 Broadcast and multicast routingNetwork Layer4-60ICMP: Internet Control Message Protocolused by hosts & routers to communicate network-level informationerror reporting: unreachable host, network, port, protocolecho request/reply (used by ping)network-layer “above” IP:ICMP msgs carried in IP datagramsICMP message:type, code plus first 8 bytes of IP datagram causing errorTypeCodedescription0 0 echo reply (ping)3 0 dest. network unreachable3 1 dest host unreachable3 2 dest protocol unreachable3 3 dest port unreachable3 6 dest network unknown3 7 dest host unknown4 0 source quench (congestioncontrol -not used)8 0 echo request (ping)9 0 route advertisement10 0 router discovery11 0 TTL expired12 0 bad IP headerNetwork Layer4-61Traceroute and ICMPSource sends series of UDP segments to destFirst has TTL =1Second has TTL=2, etc.Unlikely port numberWhen nth datagram arrives to nth router:Router discards datagramAnd sends to source an ICMP message (type 11, code 0)Message includes name of router& IP addressWhen ICMP message arrives, source calculates RTTTraceroute does this 3 timesStopping criterionUDP segment eventually arrives at destination hostDestination returns ICMP “host unreachable” packet (type 3, code 3)When source gets this ICMP, stops.Network Layer4-62Chapter 4: Network Layer4. 1 Introduction4.2 Virtual circuit and datagram networks4.3 What‟s inside a router4.4 IP: Internet ProtocolDatagram formatIPv4 addressingICMPIPv64.5 Routing algorithmsLink stateDistance VectorHierarchical routing4.6 Routing in the InternetRIPOSPFBGP4.7 Broadcast and multicast routingNetwork Layer4-63IPv6Initial motivation:32-bit address space soon to be completely allocated. Additional motivation:header format helps speed processing/forwardingheader changes to facilitate QoS IPv6 datagram format:fixed-length 40 byte headerno fragmentation allowedNetwork Layer4-64IPv6 Header (Cont)Priority:identify priority among datagrams in flowFlow Label:identify datagrams in same “flow.” (concept of“flow” not well defined).Next header:identify upper layer protocol for data Network Layer4-65Other Changes from IPv4Checksum:removed entirely to reduce processing time at each hopOptions:allowed, but outside of header, indicated by “Next Header” fieldICMPv6:new version of ICMPadditional message types, e.g. “Packet Too Big”multicast group management functionsNetwork Layer4-66Transition From IPv4 To IPv6Not all routers can be upgraded simultaneousno “flag days”How will the network operate with mixed IPv4 and IPv6 routers? Tunneling:IPv6 carried as payload in IPv4 datagram among IPv4 routersNetwork Layer4-67TunnelingABEFIPv6IPv6IPv6IPv6tunnelLogical view:Physical view:ABEFIPv6IPv6IPv6IPv6IPv4IPv4Network Layer4-68TunnelingABEFIPv6IPv6IPv6IPv6tunnelLogical view:Physical view:ABEFIPv6IPv6IPv6IPv6CDIPv4IPv4Flow: XSrc: ADest: FdataFlow: XSrc: ADest: FdataFlow: XSrc: ADest: FdataSrc:BDest: EFlow: XSrc: ADest: FdataSrc:BDest: EA-to-B:IPv6E-to-F:IPv6B-to-C:IPv6 insideIPv4B-to-C:IPv6 insideIPv4Network Layer4-69Chapter 4: Network Layer4. 1 Introduction4.2 Virtual circuit and datagram networks4.3 What‟s inside a router4.4 IP: Internet ProtocolDatagram formatIPv4 addressingICMPIPv64.5 Routing algorithmsLink stateDistance VectorHierarchical routing4.6 Routing in the InternetRIPOSPFBGP4.7 Broadcast and multicast routingNetwork Layer4-701230111value in arrivingpacket’s headerrouting algorithmlocal forwarding tableheader valueoutput link01000101011110013221Interplay between routing, forwardingNetwork Layer4-71uyxwvz2213112535Graph: G = (N,E)N = set of routers = { u, v, w, x, y, z }E = set of links ={ (u,v), (u,x), (v,x), (v,w), (x,w), (x,y), (w,y), (w,z), (y,z) }Graph abstractionRemark: Graph abstraction is useful in other network contextsExample: P2P, where N is set of peers and E is set of TCP connectionsNetwork Layer4-72Graph abstraction: costsuyxwvz2213112535•c(x,x‟) = cost of link (x,x‟)-e.g., c(w,z) = 5•cost could always be 1, or inversely related to bandwidth,or inversely related to congestionCost of path (x1, x2, x3,…, xp) = c(x1,x2) + c(x2,x3) + … + c(xp-1,xp) Question: What‟s the least-cost path between u and z ?Routing algorithm: algorithm that finds least-cost pathNetwork Layer4-73Routing Algorithm classificationGlobal or decentralized information?Global:all routers have complete topology, link cost info“link state” algorithmsDecentralized:router knows physically-connected neighbors, link costs to neighborsiterative process of computation, exchange of info with neighbors“distance vector” algorithmsStatic or dynamic?Static:routes change slowly over timeDynamic:routes change more quicklyperiodic updatein response to link cost changesNetwork Layer4-74Chapter 4: Network Layer4. 1 Introduction4.2 Virtual circuit and datagram networks4.3 What‟s inside a router4.4 IP: Internet ProtocolDatagram formatIPv4 addressingICMPIPv64.5 Routing algorithmsLink stateDistance VectorHierarchical routing4.6 Routing in the InternetRIPOSPFBGP4.7 Broadcast and multicast routingNetwork Layer4-75A Link-State Routing AlgorithmDijkstra‟s algorithmnet topology, link costs known to all nodesaccomplished via “link state broadcast” all nodes have same infocomputes least cost paths from one node („source”) to all other nodesgives forwarding tablefor that nodeiterative: after k iterations, know least cost path to k dest.‟sNotation:c(x,y):link cost from node x to y; = ∞ if not direct neighborsD(v):current value of cost of path from source to dest. vp(v):predecessor node along path from source to vN':set of nodes whose least cost path definitively knownNetwork Layer4-76Dijsktra‟s Algorithm1 Initialization:2 N'= {u} 3 for all nodes v 4 if v adjacent to u 5 then D(v) = c(u,v) 6 else D(v) = ∞7 8 Loop9 find w not in N'such that D(w) is a minimum 10 add w to N'11 update D(v) for all v adjacent to w and not in N': 12 D(v) = min( D(v), D(w) + c(w,v) ) 13 /* new cost to v is either old cost to v or known 14 shortest path cost to w plus cost from w to v */15 until all nodes in N'Network Layer4-77Dijkstra‟s algorithm: exampleStep012345N'uuxuxyuxyvuxyvwuxyvwzD(v),p(v)2,u2,u2,uD(w),p(w)5,u4,x3,y3,yD(x),p(x)1,uD(y),p(y)∞2,xD(z),p(z)∞ ∞ 4,y4,y4,yuyxwvz2213112535Network Layer4-78Dijkstra‟s algorithm: example (2) uyxwvzResulting shortest-path tree from u:vxywz(u,v)(u,x)(u,x)(u,x)(u,x)destinationlinkResulting forwarding table in u:Network Layer4-79Dijkstra‟s algorithm, discussionAlgorithm complexity: n nodeseach iteration: need to check all nodes, w, not in Nn(n+1)/2 comparisons: O(n2)more efficient implementations possible: O(nlogn)Oscillations possible:e.g., link cost = amount of carried trafficADCB11+ee0e1100ADCB2+e0001+e1ADCB02+e1+e100ADCB2+e0e01+e1initially… recomputerouting… recompute… recomputeNetwork Layer4-80Chapter 4: Network Layer4. 1 Introduction4.2 Virtual circuit and datagram networks4.3 What‟s inside a router4.4 IP: Internet ProtocolDatagram formatIPv4 addressingICMPIPv64.5 Routing algorithmsLink stateDistance VectorHierarchical routing4.6 Routing in the InternetRIPOSPFBGP4.7 Broadcast and multicast routingNetwork Layer4-81Distance Vector Algorithm Bellman-Ford Equation (dynamic programming)Definedx(y) := cost of least-cost path from x to yThendx(y) = min {c(x,v) + dv(y) }where min is taken over all neighbors v of xvNetwork Layer4-82Bellman-Ford example uyxwvz2213112535Clearly, dv(z) = 5, dx(z) = 3, dw(z) = 3du(z) = min { c(u,v) + dv(z),c(u,x) + dx(z),c(u,w) + dw(z) }= min {2 + 5,1 + 3,5 + 3} = 4Node that achieves minimum is nexthop in shortest path ➜forwarding tableB-F equation says:Network Layer4-83Distance Vector Algorithm Dx(y)= estimate of least cost from x to yNode x knows cost to each neighbor v: c(x,v)Node x maintains distance vector Dx= [Dx(y): y єN ]Node x also maintains its neighbors‟ distance vectorsFor each neighbor v, x maintains Dv= [Dv(y): y єN ]Network Layer4-84Distance vector algorithm (4)Basic idea:Each node periodically sends its own distance vector estimate to neighborsWhen a node x receives new DV estimate from neighbor, it updates its own DV using B-F equation:Dx(y) ←minv{c(x,v) + Dv(y)} for each node y ∊NUnder minor, natural conditions, the estimate Dx(y) converge to the actual least costdx(y) Network Layer4-85Distance Vector Algorithm (5)Iterative, asynchronous: each local iteration caused by: local link cost change DV update message from neighborDistributed:each node notifies neighbors onlywhen its DV changesneighbors then notify their neighbors if necessarywaitfor (change in local link cost or msg from neighbor)recomputeestimatesif DV to any dest has changed, notifyneighbors Each node:Network Layer4-86x y zxyz0 2 7∞∞∞∞∞∞fromcost tofromfromx y zxyz0fromcost tox y zxyz∞∞∞∞∞cost tox y zxyz∞∞∞710cost to∞2 0 1∞ ∞ ∞2 0 17 1 0timexz127ynode x tablenode y tablenode z tableDx(y)=min{c(x,y)+Dy(y),c(x,z)+Dz(y)}=min{2+0,7+1}=2Dx(z)=min{c(x,y)+Dy(z),c(x,z)+Dz(z)}=min{2+1,7+0}=332 Network Layer4-87x y zxyz0 2 7∞∞∞∞∞∞fromcost tofromfromx y zxyz0 2 3fromcost tox y zxyz0 2 3fromcost tox y zxyz∞∞∞∞∞cost tox y zxyz0 2 7fromcost tox y zxyz0 2 3fromcost tox y zxyz0 2 3fromcost tox y zxyz0 2 7fromcost tox y zxyz∞∞∞710cost to∞2 0 1∞ ∞ ∞2 0 17 1 02 0 17 1 02 0 13 1 02 0 13 1 02 0 13 1 02 0 13 1 0timexz127ynode x tablenode y tablenode z tableDx(y)=min{c(x,y)+Dy(y),c(x,z)+Dz(y)}=min{2+0,7+1}=2Dx(z)=min{c(x,y)+Dy(z),c(x,z)+Dz(z)}=min{2+1,7+0}=3Network Layer4-88Distance Vector: link cost changesLink cost changes:node detects local link cost change updates routing info, recalculates distance vectorif DV changes, notify neighbors “goodnews travelsfast”xz1450y1At time t0, ydetects the link-cost change, updates its DV, and informs its neighbors.At time t1, zreceives the update from yand updates its table. It computes a new least cost to xand sends its neighbors its DV.At time t2, yreceives z‟s update and updates its distance table. y‟s least costs do not change and hence ydoes notsend any message to z. Network Layer4-89Distance Vector: link cost changesLink cost changes:good news travels fast bad news travels slow -“count to infinity” problem!44 iterations before algorithm stabilizes: see textPoisoned reverse:If Z routes through Y to get to X :Z tells Y its (Z‟s) distance to X is infinite (so Y won‟t route to X via Z)will this completely solve count to infinity problem?xz1450y60Network Layer4-90Comparison of LS and DV algorithmsMessage complexityLS:with n nodes, E links, O(nE) msgs sent DV: exchange between neighbors onlyconvergence time variesSpeed of ConvergenceLS:O(n2) algorithm requires O(nE) msgsmay have oscillationsDV: convergence time variesmay be routing loopscount-to-infinity problemRobustness:what happens if router malfunctions?LS:node can advertise incorrect linkcosteach node computes only its owntableDV:DV node can advertise incorrect pathcosteach node‟s table used by others •error propagate thru networkNetwork Layer4-91Chapter 4: Network Layer4. 1 Introduction4.2 Virtual circuit and datagram networks4.3 What‟s inside a router4.4 IP: Internet ProtocolDatagram formatIPv4 addressingICMPIPv64.5 Routing algorithmsLink stateDistance VectorHierarchical routing4.6 Routing in the InternetRIPOSPFBGP4.7 Broadcast and multicast routingNetwork Layer4-92Hierarchical Routingscale:with 200 million destinations:can‟t store all dest‟s in routing tables!routing table exchange would swamp links!administrative autonomyinternet = network of networkseach network admin may want to control routing in its own networkOur routing study thus far -idealization all routers identicalnetwork “flat”… nottrue in practiceNetwork Layer4-93Hierarchical Routingaggregate routers into regions,“autonomous systems” (AS)routers in same AS run same routing protocol“intra-AS” routingprotocolrouters in different AS can run different intra-AS routing protocolGateway routerDirect link to router in another ASNetwork Layer4-943b1d3a1c2aAS3AS1AS21a2c2b1bIntra-ASRouting algorithmInter-ASRouting algorithmForwardingtable3cInterconnected ASesforwarding table configured by both intra-and inter-AS routing algorithmintra-AS sets entries for internal destsinter-AS & Intra-As sets entries for external dests Network Layer4-953b1d3a1c2aAS3AS1AS21a2c2b1b3cInter-AS taskssuppose router in AS1 receives datagram dest outside of AS1router should forward packet to gateway router, but which one?AS1 must:1.learn which dests reachable through AS2, which through AS32.propagate this reachability info to all routers in AS1Job of inter-AS routing!Network Layer4-96Example: Setting forwarding table in router 1dsuppose AS1 learns (via inter-AS protocol) that subnet xreachable via AS3 (gateway 1c) but not via AS2.inter-AS protocol propagates reachability info to all internal routers.router 1d determines from intra-AS routing info that its interface Iis on the least cost path to 1c.installs forwarding table entry (x,I)3b1d3a1c2aAS3AS1AS21a2c2b1b3cx…Network Layer4-97Example: Choosing among multiple ASesnow suppose AS1 learns from inter-AS protocol that subnet xis reachable from AS3 andfrom AS2.to configure forwarding table, router 1d must determine towards which gateway it should forward packets for dest x. this is also job of inter-AS routing protocol!3b1d3a1c2aAS3AS1AS21a2c2b1b3cx……Network Layer4-98Learn from inter-AS protocol that subnet x is reachable via multiple gatewaysUse routing infofrom intra-AS protocol to determinecosts of least-cost paths to eachof the gatewaysHot potato routing:Choose the gatewaythat has the smallest least costDetermine fromforwarding table the interface I that leads to least-cost gateway. Enter (x,I) in forwarding tableExample: Choosing among multiple ASesnow suppose AS1 learns from inter-AS protocol that subnet xis reachable from AS3 andfrom AS2.to configure forwarding table, router 1d must determine towards which gateway it should forward packets for dest x. this is also job of inter-AS routing protocol!hot potato routing:send packet towards closest of two routers.Network Layer4-99Chapter 4: Network Layer4. 1 Introduction4.2 Virtual circuit and datagram networks4.3 What‟s inside a router4.4 IP: Internet ProtocolDatagram formatIPv4 addressingICMPIPv64.5 Routing algorithmsLink stateDistance VectorHierarchical routing4.6 Routing in the InternetRIPOSPFBGP4.7 Broadcast and multicast routingNetwork Layer4-100Intra-AS Routingalso known as Interior Gateway Protocols (IGP)most common Intra-AS routing protocols:RIP: Routing Information ProtocolOSPF: Open Shortest Path FirstIGRP: Interior Gateway Routing Protocol (Cisco proprietary)Network Layer4-101Chapter 4: Network Layer4. 1 Introduction4.2 Virtual circuit and datagram networks4.3 What‟s inside a router4.4 IP: Internet ProtocolDatagram formatIPv4 addressingICMPIPv64.5 Routing algorithmsLink stateDistance VectorHierarchical routing4.6 Routing in the InternetRIPOSPFBGP4.7 Broadcast and multicast routingNetwork Layer4-102RIP ( Routing Information Protocol)distance vector algorithmincluded in BSD-UNIX Distribution in 1982distance metric: # of hops (max = 15 hops)DCBAuvwxyzdestinationhopsu 1v 2w 2x 3y 3z 2From router A to subsets:Network Layer4-103RIP advertisementsdistance vectors:exchanged among neighbors every 30 sec via Response Message (also called advertisement)ech advertisement: list of up to 25 destination nets within ASNetwork Layer4-104RIP: ExampleDestination NetworkNext Router Num. of hops to dest.wA2yB2zB7x--1….….....wxyzACDBRouting table in DNetwork Layer4-105RIP: ExampleDestination NetworkNext Router Num. of hops to dest.wA2yB2zB A7 5x--1….….....Routing table in DwxyzACDBDest Next hopsw-1x-1zC 4….… ...Advertisementfrom A to DNetwork Layer4-106RIP: Link Failure and RecoveryIf no advertisement heard after 180 sec --> neighbor/link declared deadroutes via neighbor invalidatednew advertisements sent to neighborsneighbors in turn send out new advertisements (if tables changed)link failure info quickly (?) propagates to entire netpoison reverseused to prevent ping-pong loops (infinite distance = 16 hops)Network Layer4-107RIP Table processingRIP routing tables managed by application-levelprocess called route-d (daemon)advertisements sent in UDP packets, periodically repeatedphysicallinknetwork forwarding(IP) tableTransprt(UDP)routedphysicallinknetwork(IP)Transprt(UDP)routedforwardingtableNetwork Layer4-108Chapter 4: Network Layer4. 1 Introduction4.2 Virtual circuit and datagram networks4.3 What‟s inside a router4.4 IP: Internet ProtocolDatagram formatIPv4 addressingICMPIPv64.5 Routing algorithmsLink stateDistance VectorHierarchical routing4.6 Routing in the InternetRIPOSPFBGP4.7 Broadcast and multicast routingNetwork Layer4-109OSPF (Open Shortest Path First)“open”: publicly availableuses Link State algorithm LS packet disseminationtopology map at each noderoute computation using Dijkstra‟s algorithmOSPF advertisement carries one entry per neighbor routeradvertisements disseminated to entireAS (via flooding)carried in OSPF messages directly over IP (rather than TCP or UDPNetwork Layer4-110OSPF “advanced” features (not in RIP)security:all OSPF messages authenticated (to prevent malicious intrusion) multiple same-cost paths allowed (only one path in RIP)For each link, multiple cost metrics for different TOS (e.g., satellite link cost set “low” for best effort; high for real time)integrated uni-and multicastsupport: Multicast OSPF (MOSPF) uses same topology data base as OSPFhierarchicalOSPF in large domains.Network Layer4-111Hierarchical OSPFNetwork Layer4-112Hierarchical OSPFtwo-level hierarchy:local area, backbone.Link-state advertisements only in area each nodes has detailed area topology; only know direction (shortest path) to nets in other areas.area border routers:“summarize” distances to nets in own area, advertise to other Area Border routers.backbone routers:run OSPF routing limited to backbone.boundary routers:connect to other AS‟s.Network Layer4-113Chapter 4: Network Layer4. 1 Introduction4.2 Virtual circuit and datagram networks4.3 What‟s inside a router4.4 IP: Internet ProtocolDatagram formatIPv4 addressingICMPIPv64.5 Routing algorithmsLink stateDistance VectorHierarchical routing4.6 Routing in the InternetRIPOSPFBGP4.7 Broadcast and multicast routingNetwork Layer4-114Internet inter-AS routing: BGPBGP (Border Gateway Protocol):thede facto standardBGP provides each AS a means to:1.Obtain subnet reachability information from neighboring ASs.2.Propagate reachability information to all AS-internal routers.3.Determine “good” routes to subnets based on reachability information and policy.allows subnet to advertise its existence to rest of Internet: “I am here”Network Layer4-115BGP basicspairs of routers (BGP peers) exchange routing info over semi-permanent TCP connections: BGP sessionsBGP sessions need not correspond to physical links.when AS2 advertises prefix to AS1:AS2 promisesit will forward any addresses datagrams towards that prefix.AS2 can aggregate prefixes in its advertisement3b1d3a1c2aAS3AS1AS21a2c2b1b3ceBGP sessioniBGP sessionNetwork Layer4-116Distributing reachability infousing eBGP session between 3a and 1c, AS3 sends prefix reachability info to AS1.1c can then use iBGP do distribute new prefix info to all routers in AS11b can then re-advertise new reachability info to AS2 over 1b-to-2a eBGP sessionwhen router learns of new prefix, creates entry for prefix in its forwarding table.3b1d3a1c2aAS3AS1AS21a2c2b1b3ceBGP sessioniBGP sessionNetwork Layer4-117Path attributes & BGP routesadvertised prefix includes BGP attributes. prefix + attributes = “route”two important attributes:AS-PATH:contains ASs through which prefix advertisement has passed: e.g, AS 67, AS 17 NEXT-HOP:indicates specific internal-AS router to next-hop AS. (may be multiple links from current AS to next-hop-AS)when gateway router receives route advertisement, uses import policyto accept/decline.Network Layer4-118BGP route selectionrouter may learn about more than 1 route to some prefix. Router must select route.elimination rules:1.local preference value attribute: policy decision2.shortest AS-PATH 3.closest NEXT-HOP router: hot potato routing4.additional criteria Network Layer4-119BGP messagesBGP messages exchanged using TCP.BGP messages:OPEN:opens TCP connection to peer and authenticates senderUPDATE:advertises new path (or withdraws old)KEEPALIVEkeeps connection alive in absence of UPDATES; also ACKs OPEN requestNOTIFICATION:reports errors in previous msg; also used to close connectionNetwork Layer4-120BGP routing policyA,B,C are provider networksX,W,Y are customer (of provider networks)X is dual-homed:attached to two networksX does not want to route from B via X to C.. so X will not advertise to B a route to CABCWXYlegend:customer network:providernetworkNetwork Layer4-121BGP routing policy (2)A advertises path AW to BB advertises path BAW to X Should B advertise path BAW to C?No way! B gets no “revenue” for routing CBAW since neither W nor C are B‟s customers B wants to force C to route to w via AB wants to route only to/from its customers!ABCWXYlegend:customer network:providernetworkNetwork Layer4-122Why different Intra-and Inter-AS routing ?Policy:Inter-AS: admin wants control over how its traffic routed, who routes through its net. Intra-AS: single admin, so no policy decisions neededScale:hierarchical routing saves table size, reduced update trafficPerformance:Intra-AS: can focus on performanceInter-AS: policy may dominate over performanceNetwork Layer4-123Chapter 4: Network Layer4. 1 Introduction4.2 Virtual circuit and datagram networks4.3 What‟s inside a router4.4 IP: Internet ProtocolDatagram formatIPv4 addressingICMPIPv64.5 Routing algorithmsLink stateDistance VectorHierarchical routing4.6 Routing in the InternetRIPOSPFBGP4.7 Broadcast and multicast routingNetwork Layer4-124R1R2R3R4sourceduplicationR1R2R3R4in-networkduplicationduplicatecreation/transmissionduplicateduplicateBroadcast Routingdeliver packets from source to all other nodessource duplication is inefficient:source duplication: how does source determine recipient addresses?Network Layer4-125In-network duplicationflooding: when node receives brdcst pckt, sends copy to all neighborsProblems: cycles & broadcast stormcontrolled flooding: node only brdcsts pkt if it hasn‟t brdcst same packet beforeNode keeps track of pckt ids already brdcstedOr reverse path forwarding (RPF): only forward pckt if it arrived on shortest path between node and sourcespanning treeNo redundant packets received by any nodeNetwork Layer4-126ABGDEcFABGDEcF(a) Broadcast initiated at A(b) Broadcast initiated at DSpanning TreeFirst construct a spanning treeNodes forward copies only along spanning treeNetwork Layer4-127ABGDEcF12345(a)Stepwise construction of spanning treeABGDEcF(b) Constructed spanning treeSpanning Tree: CreationCenter nodeEach node sends unicast join message to center nodeMessage forwarded until it arrives at a node already belonging to spanning treeMulticast Routing: Problem StatementGoal:find a tree (or trees) connecting routers having local mcast group members tree:not all paths between routers usedsource-based:different tree from each sender to rcvrsshared-tree:same tree used by all group membersShared treeSource-based treesApproaches for building mcast treesApproaches:source-based tree:one tree per sourceshortest path treesreverse path forwardinggroup-shared tree:group uses one treeminimal spanning (Steiner) center-based trees…we first look at basic approaches, then specific protocols adopting these approachesShortest Path Treemcast forwarding tree: tree of shortest path routes from source to all receiversDijkstra‟s algorithmR1R2R3R4R5R6R7216345irouter with attachedgroup memberrouter with no attachedgroup memberlink used for forwarding,i indicates order linkadded by algorithmLEGENDS: sourceReverse Path Forwardingif (mcast datagram received on incoming link on shortest path back to center)thenflood datagram onto all outgoing linkselseignore datagramrely on router‟s knowledge of unicast shortest path from it to sendereach router has simple forwarding behavior:Reverse Path Forwarding: example•result is a source-specific reverseSPT–may be a bad choice with asymmetric linksR1R2R3R4R5R6R7router with attachedgroup memberrouter with no attachedgroup memberdatagram will be forwardedLEGENDS: sourcedatagram will not be forwardedReverse Path Forwarding: pruningforwarding tree contains subtrees with no mcast group membersno need to forward datagrams down subtree“prune” msgs sent upstream by router with no downstream group membersR1R2R3R4R5R6R7router with attachedgroup memberrouter with no attachedgroup memberprune messageLEGENDS: sourcelinks with multicastforwardingPPPShared-Tree: Steiner TreeSteiner Tree:minimum cost tree connecting all routers with attached group membersproblem is NP-completeexcellent heuristics existsnot used in practice:computational complexityinformation about entire network neededmonolithic: rerun whenever a router needs to join/leaveCenter-based treessingle delivery tree shared by allone router identified as “center”of treeto join:edge router sends unicast join-msgaddressed to center routerjoin-msg “processed” by intermediate routers and forwarded towards centerjoin-msgeither hits existing tree branch for this center, or arrives at centerpath taken by join-msgbecomes new branch of tree for this routerCenter-based trees: an exampleSuppose R6 chosen as center:R1R2R3R4R5R6R7router with attachedgroup memberrouter with no attachedgroup memberpath order in which join messages generatedLEGEND2131Internet Multicasting Routing: DVMRPDVMRP:distance vector multicast routing protocol, RFC1075flood and prune:reverse path forwarding, source-based treeRPF tree based on DVMRP‟s own routing tables constructed by communicating DVMRP routers no assumptions about underlying unicastinitial datagram to mcast group flooded everywhere via RPFrouters not wanting group: send upstream prune msgsDVMRP: continued…soft state:DVMRP router periodically (1 min.) “forgets” branches are pruned: mcast data again flows down unpruned branchdownstream router: reprune or else continue to receive datarouters can quickly regraft to tree following IGMP join at leafodds and endscommonly implemented in commercial routersMbone routing done using DVMRPTunnelingQ:How to connect “islands” of multicast routers in a “sea” of unicast routers? mcast datagram encapsulated inside “normal” (non-multicast-addressed) datagramnormal IP datagram sent thru “tunnel” via regular IP unicast to receiving mcast routerreceiving mcast router unencapsulates to get mcast datagramphysical topologylogical topologyPIM: Protocol Independent Multicastnot dependent on any specific underlying unicast routing algorithm (works with all)two different multicast distribution scenarios :Dense:group members densely packed, in “close” proximity.bandwidth more plentifulSparse:# networks with group members small wrt # interconnected networksgroup members “widely dispersed”bandwidth not plentifulConsequences of Sparse-Dense Dichotomy:Densegroup membership by routers assumed until routers explicitly prunedata-drivenconstruction on mcast tree (e.g., RPF)bandwidth and non-group-router processing profligateSparse:no membership until routers explicitly joinreceiver-drivenconstruction of mcast tree (e.g., center-based)bandwidth and non-group-router processing conservativePIM-Dense Modeflood-and-prune RPF, similar to DVMRP butunderlying unicast protocol provides RPF info for incoming datagramless complicated (less efficient) downstream flood than DVMRP reduces reliance on underlying routing algorithmhas protocol mechanism for router to detect it is a leaf-node routerPIM -Sparse Modecenter-based approachrouter sends joinmsg to rendezvous point (RP)intermediate routers update state and forward joinafter joining via RP, router can switch to source-specific treeincreased performance: less concentration, shorter pathsR1R2R3R4R5R6R7joinjoinjoinall data multicastfrom rendezvouspointrendezvouspointPIM -Sparse Modesender(s):unicast data to RP, which distributes down RP-rooted treeRP can extend mcast tree upstream to sourceRP can send stopmsg if no attached receivers“no one is listening!”R1R2R3R4R5R6R7joinjoinjoinall data multicastfrom rendezvouspointrendezvouspointNetwork Layer4-145Chapter 4: summary4. 1 Introduction4.2 Virtual circuit and datagram networks4.3 What‟s inside a router4.4 IP: Internet ProtocolDatagram formatIPv4 addressingICMPIPv64.5 Routing algorithmsLink stateDistance VectorHierarchical routing4.6 Routing in the InternetRIPOSPFBGP4.7 Broadcast and multicast routing