L2: Introduction to Communication Networks: Internet Outline

EECS 3213 Fall 2014

L2: Introduction to Communication Networks: Internet

Sebastian Magierowski York University

EECS 3213, F14 L2: Intro Internet 1

Outline

• ARPANET – A connectionless datagram network • Internet – A connectionless/connection-oriented datagram network – best-effort service • Local Area Networks – Ethernet: connectionless protocol, medium access control

EECS 3213, F14 L2: Intro Internet 2

1 Internet

• An internetwork • Multi-tiered, decentralized organization • A network of computers – Powerful processing at network edge – Move communication complexity towards the edge – Develop sophisticated protocols

SEC. 1.5 EXAMPLE NETWORKS 55 EECS 3213, F14 SEC. 1.5L2: Intro EXAMPLEInternet NETWORKS 55 3 Of course, countless technical books have been written about the Internet and its protocols as well. For more information,Of course, countless see, for example,technical Maufer books have (1999). been written about the Internet and its protocols as well. For more information, see, for example, Maufer (1999). The ARPANET The ARPANET The story begins in the late 1950s. At the height of the Cold War, the U.S. The story begins in the late 1950s. At the height of the Cold War, the U.S. DoD wanted a command-and-control network that could survive a nuclear war. DoD wanted a command-and-control network that could survive a nuclear war. At that time, all military communicationsAt that time, all usedmilitary the publiccommunications telephone used network, the public telephone network, which was considered vulnerable.which wasThe considered reason for vulnerable.this belief can The be reason gleaned for from this belief can be gleaned from Fig. 1-25(a). Here the blackFig. dots 1-25(a).represent Here telephone the blackswitching dots represent offices, telephoneeach of switching offices, each of which was connected to thousandswhichTelephone was of telephones.connected toThese thousands vs. switching ofInternet telephones. offices were,These switching offices were, in turn, connected to higher-levelin turn, switching connected officesto higher-level (toll offices), switching to form offices a (toll offices), to form a national hierarchy with onlynational a small hierarchy amount of withredundancy. only a small The amount vulnerability of redundancy. of The vulnerability of the system was that the destructionthe system of a was few that key the toll destruction offices could of a fragment few key toll it into offices could fragment it into Publicmany isolatedSwitched islands. Telephonemany isolated Network islands. Distributed Switching System (PSTN) Switching Switching office office

Toll Toll office office

• U.S. military(a) dependent (a)on PSTN(b) in 50’s (b) Figure 1-25. (a) Structure of the Figuretelephone 1-25. system.(a) Structure (b) Baran’s of the proposedtelephone dis- system. (b) Baran’s proposed dis- • Easytributed to switching cripple system. bytributed taking switching system.out switching centers Around 1960, the DoD awardedAround a contract 1960, the to theDoD RAND awarded Corporation a contract to to find the RAND a Corporation to find a • solution.RAND One ofCorp. its employees, (Paulsolution. Paul One Baran,Baran of cameits employees,) up proposes with the Paul highly Baran, distributed camea distributed up with the highly distributed network and fault-tolerant design of Fig. 1-25(b). Since the paths between any two switch- and fault-tolerant design of Fig. 1-25(b). Since the paths between any two switching offices were now much longer than analog signals could travel without distor- ing offices were now much longer than analog signals could travel without distor- • AT&T rejectedtion, the Baran idea proposed when using digital askedpacket-switching to build technology. prototypeBaran wrote sev- tion, Baran proposed using digital packet-switching technology. Baran wrote several reports for the DoD describing his ideas in detail (Baran, 1964). Officials at eral reports for the DoD describing his ideas in detail (Baran, 1964). Officials at the Pentagon liked the concept and asked AT&T, then the U.S.’ national tele- EECS 3213,the Pentagon F14 liked the conceptphone and monopoly, asked AT&T,L2: to build Intro then a Internet prototype. the U.S.’ national AT&T dismissed tele- Baran’s ideas out of 4 phone monopoly, to build ahand. prototype. The biggest AT&T and dismissed richest corporation Baran’s ideas in the out world of was not about to allow hand. The biggest and richest corporation in the world was not about to allow

2 5858 INTRODUCTIONINTRODUCTION CHAP. 1 CHAP. 1 58 INTRODUCTIONARPANET CHAP. 1 nodenode by by a ateam team58 led led by Lenby58 Len Kleinrock Kleinrock (a pioneer (aINTRODUCTION pioneer of the theoryINTRODUCTION of the of theory packet of switching) packet switching)CHAP.CHAP. 1 1 node by atoto team the the SRI led SRI by node. node. Len The Kleinrock The network network (a grew pioneer grew quickly of quicklythe as theory more as of IMPs more packet were IMPs switching) delivered were delivered and and to the SRIinstalled; node. it The soonnode network spanned by a team grewnode the led Unitedby quickly by a team Len States. asled Kleinrock moreby Len Figure IMPs Kleinrock (a pioneer 1-27 were shows (a of pioneer delivered the how theory of rapidlythe and of theory packet the of packetswitching) switching) installed; itto soon the spannedSRI node.to the the SRIThe United node. network The States. grew network quickly Figure grew as quickly 1-27 more shows as IMPs more were how IMPs delivered rapidly were delivered andthe and • installed;RANDARPANET it soon idea spanned grew implemented in the the United firstinstalled; 3 years. States. it soon Figurein spanned late 1-27 the 60’s shows United States.howas rapidlynetwork Figure 1-27the shows of how rapidly the ARPANETinstalled; grew inthe it soon first spanned 3 years. the United States. Figure 1-27 shows how rapidly the ARPANET grew in the first 3 years.ARPANET grew in the first 3 years. computersSRI UTAH betweenARPANETSRI grew UTAH research inMIT the first 3 years.SRI centers UTAH ILLINOIS MIT LINCOLN CASE SRI UTAH MIT SRI UTAH ILLINOIS MIT LINCOLN CASE SRI UTAH SRI UTAH MIT SRI UTAH ILLINOIS MIT LINCOLN CASE SRI UTAH SRI UTAH MIT SRI UTAHSRI UTAH ILLINOIS MIT LINCOLN CASE SRI UTAH SRI UTAH MIT SRI UTAH ILLINOIS MIT LINCOLN CASE UCSB UCSB SDC UCSB SDC CARN UCSBSTAN SDC UCSB SDC CARN UCSBUCSB SDC UCSB SDC CARN UCSB UCSB SDC UCSB SDC STANCARN UCSB UCSB UCSB SDC UCSB SDC CARN STAN STAN UCLA UCLA RAND BBN UCLA RAND STANBBN HARVARD BURROUGHS UCLA UCLA RAND BBN UCLA RAND BBN HARVARD BURROUGHS (a) (b) (c) UCLA UCLAUCLA RANDUCLAUCLABBN RAND(a) UCLAUCLABBNRANDRANDBBN(b)UCLABBNUCLARANDHARVARDRAND BURROUGHSBBN BBNHARVARDHARVARD(c) BURROUGHS BURROUGHS Dec. (a)’69 (a) (b) (a) July(b) ’70(b) (c) Mar.(c) (c) ’71 SRI LBL MCCLELLAN SRIUTAH ILLINOISLBL MCCLELLAN MIT UTAH ILLINOIS MIT

CCA CCA MCCLELLAN MCCLELLAN SRI LBL MCCLELLANAMES TIP UTAH ILLINOIS MIT AMES TIP BBN SRI LBL MCCLELLANSRI LBLUTAHMCCLELLANILLINOISUTAHBBN MIT ILLINOIS MIT SRI UTAH NCAR GWC LINCOLNSRI CASEUTAH NCAR GWC LINCOLN CASE HARVARD AMES IMP HARVARDAMES IMP LINC CCA LINC MCCLELLAN AMES TIPX-PARC ABERDEEN X-PARC RADC ABERDEENCCA BBN MCCLELLAN RADC AMES TIPILLINOIS STANFORD CCA ILLINOIS CARN NBS Apr. ’72 MCCLELLAN SRI UTAH AMES NCAR CARN GWC USC LINCOLN Sept. CASESTANFORDAMES ’72 TIP BBNNBS HARVARD AMES USC LINC AMES IMP BBNETAC LINC SRI UTAH NCAR GWC LINCOLNUCSB CASELINC X-PARCMITRE HARVARDFNWCETAC RAND SRI UTAH NCAR GWC LINCOLN CASEAMESRADC IMP MIT LINC ABERDEEN UCSB MITREILLINOIS FNWC RAND HARVARDTINKER ARPA MIT STANX-PARCSDC CARN AMESABERDEENTINKER IMP STANFORD NBS LINC AMESRADC USC ETAC ARPA STAN ILLINOISSDC LINC X-PARC ABERDEEN MITRE CARN ETACRADC STANFORD NBS ETAC RADC AMES USC UCSB ILLINOIS MITRE FNWC RANDUCSBMITRE UCSD SAAC LINC MIT ETACSTANFORD TINKER RADCNBS UCLA RANDCARNTINKERUCSBBBN UCSDHARVARD NBS ARPA UCSB AMES USC STAN MITRESDC LINCFNWC RAND SAAC BELVOIR MIT ETAC TINKER ETAC UCSBUCLA RAND TINKER BBN HARVARD NBS FNWC RAND BELVOIR MITRE RADC CMU STAN SDC MIT MITRE ARPA ETAC UCSB UCSD TINKER CMU SAAC STAN SDC MITRE RADC ARPA UCLA RAND TINKER BBNETACHARVARD NBS UCLA SDC USC BELVOIRNOAA GWC CASE UCSB UCSD SAAC MITRE UCLA (d) SDC USC NOAA GWC CASE(e) CMURADC • RetiredUCLA RAND inTINKER ‘90BBN atHARVARD >100 NBS hosts UCSB UCSD BELVOIR SAAC UCLA RAND TINKER(d) BBN HARVARD NBS UCLA (e)SDC USCCMU NOAA BELVOIRGWC CASE – California - Norway Figure(d) 1-27. Growth of the ARPANET. (a) December(e)1969. (b) JulyCMU 1970. UCLA(c) March 1971.SDC (d)USC April 1972.NOAA (e) SeptemberGWC 1972. CASE EECS 3213, F14 Figure(d) 1-27. Growth of the ARPANET.L2: Intro(a) Internet DecemberUCLA (e) 1969.SDC (b) JulyUSC 1970. NOAA GWC CASE 5 (c) March 1971. (d) April 1972. (e) September 1972. Figure(d) 1-27. Growth of the ARPANET. (a) December 1969.(e) (b) July 1970. (c) MarchIn 1971. addition (d) April to helping1972. (e) the September fledgling 1972. ARPANET grow, ARPA also funded re- Figure 1-27. Growth of the ARPANET.search on(a) the December use of satellite1969.networks (b) July 1970. and mobile packet radio networks. In one In addition to helping the fledgling ARPANET grow, ARPA also funded re- (c) March 1971.Figure (d) April 1-27. 1972.Growthnow (e) September offamous the ARPANET. demonstration, 1972. (a) December a truck driving1969. around (b) July in 1970. California used the packet search on the useIn of additionsatellite tonetworks helping and the fledglingmobile packet ARPANET radio networks. grow, ARPA In onealso funded re- (c) March 1971. (d)radio April network 1972.(e) to send September messages 1972. to SRI, which were then forwarded over the now famous demonstration,search on theARPANET use a truck of satellite drivingto thenetworks East around Coast, and in where California mobile they packet were used shipped radio the packet networks. to University In College one in In additionradio network to helpingnow to sendthe famous fledgling messagesLondon demonstration, ARPANET over to SRI,the s aatellite which truck grow,network. driving were ARPA then around This also forwardedallowed in funded California a researcher over re- used the in the truck packet to use a computer in London while driving around in California. search onARPANET theIn use addition of tosatelliteradio the to East network helpingnetworks Coast, to thewhere sendand fledgling mobile messages they were packet ARPANET to shipped SRI, radio which to networks. grow,University were ARPA then In College one forwarded also in funded over re-the now famous demonstration,ARPANET a truck to the drivingThis East experimentCoast, around where alsoin California demonstrated they were used shipped that the the to packet existing University ARPANET College protocols in Londonsearchover on the the use satellite of satellitenetwork.networks This allowed and a mobile researcher packet in the radio truck networks. to use a In one radio network to sendLondon messages overwere theto SRI, s notatellite suitable whichnetwork. for were running This then allowed over forwarded different a researcher networks. over the in the This truck observation to use a led to SEC. 1.5computer in London EXAMPLE whilemore drivingNETWORKS research around on protocols,in California. culminating with57 the invention of the TCP/IP model ARPANETnow toThis famousthe experiment Eastcomputer demonstration,Coast, also where in London demonstrated they a while truck were driving thatshipped driving the around toexistingaround University in California. inARPANET California College protocols in used the packet radio networkThis to send experimentand messages protocols also (Cerf to demonstrated SRI, and Kahn, which 1974). that were the TCP/IP existing then was forwarded specifically ARPANET designed over protocols the to handle toLondon build the overwere subnet the not andsatellite suitable writeARPANET:network. the for subnet runningcommunication This software. over allowed different BBNBasic over a researcher chose internetworks, networks. to Structure use in speciallyThis thesomething truck observationto becoming use aled increasingly to important were not suitable for running over different networks. This observation led to modifiedcomputer HoneywellmoreARPANET in London research DDP-316 while to on the protocols, driving minicomputers Eastas around Coast, more culminating and inwith where more California. 12K with networks they 16-bit the wereinvention were words hookedshipped of of core up the to TCP/IPthe University ARPANET. model College in memory asand theLondon IMPs. protocols over Themore (Cerf IMPs theresearch ands didatellite notKahn, have onnetwork. 1974). protocols, disks, TCP/IPsince This culminating moving wasallowed specifically parts with a were researcher the con- invention designed in ofto the handlethe truck TCP/IP tomodel use a This experiment alsoand protocols demonstrated (Cerf and that Kahn, the 1974). existing TCP/IP ARPANET was specifically protocols designed to handle sideredwerenot unreliable.communicationcomputer suitable The for in IMPs runningLondon over were internetworks, overinterconnected while different driving something bynetworks. around 56-kbps inbecoming lines California. This leased observation increasingly from led important to •telephone Nodescompanies.as more consisted and Although morecommunication networks 56 kbpsofwere isminicomputers over now hooked theinternetworks, choice up to of the teenagers something ARPANET. connected who becoming can- increasingly to hosts important notmore afford research DSL orThis on cable, protocols, experiment itas was more then culminating and the also more best moneydemonstrated networks with could the were buy.invention hooked that the up of to the existing the TCP/IP ARPANET. ARPANET model protocols and–The protocolsInterface softwarewere (Cerf was not Message split and suitable into Kahn, two for 1974). parts:Processors running subnet TCP/IP over and was host. different(IMPs specifically The subnet) networks. designed software This to handle observation led to consistedcommunication ofmore the IMP research over endinternetworks, of the onhost-IMP protocols, connection, something culminatingthe becoming IMP-IMP with increasinglythe protocol, invention and important of the TCP/IP model •a as sourceLinked more IMP andand to more protocolsby destination networks 56-kbps (Cerf IMP wereprotocol and hooked lines Kahn, designed up 1974). leased to totheimprove ARPANET. TCP/IP reliability. was specificallyThe designed to handle originalfrom ARPANETcommunication telephone design is shown over company in internetworks, Fig. 1-26. something becoming increasingly important as more and more networks were hooked up to the ARPANET. Host-host protocol Host Host-IMP protocol

Source IMP to destination IMP protocol

IMP-IMP protocol IMP-IMP protocol Subnet

IMP

• Protocols Figuredeveloped 1-26. The original ARPANETfor communication design. – agreement/rules on how communications are to proceed Outside the subnet, software was also needed, namely, the host end of the host-IMP– IMP-IMP, connection, theS/IMP-D/IMP, host-host protocol, andHost-IMP, the application Host-Host software. It soon became clear that BBN was of the opinion that when it had accepted a message on a host-IMP wire and placed it on the host-IMP wire at the destination, its job was EECS 3213, F14 L2: Intro Internet 6 done. Roberts had a problem, though: the hosts needed software too. To deal with it, he convened a meeting of network researchers, mostly graduate students, at Snowbird, Utah, in the summer of 1969. The graduate students expected some network expert to explain the grand design of the network and its software to them and then assign each of them the job of writing part of it. They were astounded when there was no network expert and no grand design. They had to figure out what to do on their own. Nevertheless, somehow an experimental network went online in December 1969 with four nodes: at UCLA, UCSB, SRI, and the University of Utah. These four were chosen because all had a large number of ARPA contracts, and all had 3 different and completely incompatible host computers (just to make it more fun). The first host-to-host message had been sent two months earlier from the UCLA Key ARPANET Characteristics

• Datagram service (just like telegraph) – connectionless (contrast with connection-oriented) – unreliable (unacknowledged) • Packet switched – messages up to 8063 bits could be sent – BUT…IMPs broke it up into 1008 bit (max) packets • Automated routing – no connection setup prior to packet transmission – distributed routing algorithm to update routing tables • Error control • Congestion control • Flow control

58 EECS 3213, F14 INTRODUCTION L2: Intro Internet CHAP. 1 7 node by a team led by Len Kleinrock (a pioneer of the theory of packet switching) to the SRI node. The network grew quickly as more IMPs were delivered and installed; it soon spanned the United States. Figure 1-27 shows how rapidly the ARPANET grew in the first 3 years.

SRI UTAH MIT SRI UTAH ILLINOIS MIT LINCOLN CASE SRI UTAH ARPANET Applications

UCSB UCSB SDC UCSB SDC CARN • ARPANET introducedSTAN many new applications

UCLA UCLA – RANDEmailBBN UCLA RAND BBN HARVARD BURROUGHS (a) –(b) remote login (c) – file transfer…

SRI LBL MCCLELLAN UTAH ILLINOIS MIT

CCA MCCLELLAN AMES TIP BBN SRI UTAH NCAR GWC LINCOLN CASE AMES IMP HARVARD LINC X-PARC RADC ABERDEEN ILLINOIS CARN STANFORD NBS AMES USC LINC ETAC UCSB MITRE FNWC RAND MIT TINKER STAN SDC ARPA ETAC MITRE RADC UCSB UCSD SAAC UCLA RAND TINKER BBN HARVARD NBS BELVOIR CMU

UCLA SDC USC NOAA GWC CASE (d) (e) EECS 3213, F14 L2: Intro Internet 8

Figure 1-27. Growth of the ARPANET. (a) December 1969. (b) July 1970. (c) March 1971. (d) April 1972. (e) September 1972.

In addition to helping the fledgling ARPANET grow, ARPA also funded research on the use of satellite networks and mobile packet radio networks. In one now famous demonstration, a truck driving around in California used the packet 4 radio network to send messages to SRI, which were then forwarded over the ARPANET to the East Coast, where they were shipped to University College in London over the satellite network. This allowed a researcher in the truck to use a computer in London while driving around in California. This experiment also demonstrated that the existing ARPANET protocols were not suitable for running over different networks. This observation led to more research on protocols, culminating with the invention of the TCP/IP model and protocols (Cerf and Kahn, 1974). TCP/IP was specifically designed to handle communication over internetworks, something becoming increasingly important as more and more networks were hooked up to the ARPANET. Internetworking

• ARPANET was a great WAN demonstration – A robust network – Capable of supporting a variety of applications • But… – Its protocol structure did not support the merging of various networks well – Not an internet – E.g. ARPANET + packet radio + satellite performed poorly • A reorganized design was proposed…

EECS 3213, F14 L2: Intro Internet 9

TCP/IP The TCP/IP Protocol Suite

• New set of rules proposed HTTP SMTP DNS RTP to enable internetworking • Kahn & Cerf argued for TCP UDP common rule layer – Hide differences between different networks IP instead of translation • The layer was eventually Network Network Network separated into 2 protocols Interface 1 Interface 2 Interface 3 – IP (Internet Protocol) • A means of getting messages The hourglass shape of the TCP/IP protocol suite underscores the features that make moving over multiple links: connectionlessTCP/IP so powerful. The operation of the single IP protocol over various networks provides independence from the underlying network technologies. The – TCP (Transmission Control Protocol)communication services of TCP and UDP provide a network-independent platform on which applications can be developed. By allowing multiple network technologies • A means of strengthening delivery guaranteesto coexist, the Internet is able to provide ubiquitous connectivity and to achieve between end-points: connection-oriented enormous economies of scale.

EECS 3213, F14 L2: Intro Internet 10

provides independence from

5 Layers & Structural Ideas

• With universally understood communication rules hosts in different types of network can talk to each other

H Net 1 Net 2 H router • Routers talk IP, hosts talk TCP & IP – Encapsulation

EECS 3213, F14 L2: Intro Internet 11

IP Addressing and Routing

• Location based addressing • Hierarchical address: Net ID + Host ID • IP packets routed according to Net ID • Routers compute routing tables using distributed algorithm H

H Net 3 G Net 1 G G G Net 5 Net 2 Net 4 H G G H

EECS 3213, F14 L2: Intro Internet 12

6 62 INTRODUCTION CHAP. 1

The big picture is shown in Fig. 1-29. Let us examine this figure piece by piece, starting with a computer at home (at the edges of the figure). To join the Internet, the computer is connected to an Internet Service Provider, or simply ISP, from who the user purchases Internet access or connectivity. This lets the computer exchange packets with all of the other accessible hosts on the Internet. The user might send packets to surf the Web or for any of a thousand other uses, it does not matter. There are many kinds of Internet access, and they are usually distinguished by how much bandwidthInternetthey Today provide and how much they cost, but the most important attribute is connectivity.

Data Tier 1 ISP International ISPs (Tier 1) center Backbone Router

Peering 3G mobile at IXP TorIX phone Fiber (Front & (FTTH) Uni.)

Dialup Cable DSL Other ISPs Cable DSLAM POP modem Data CMTS path DSL modem

Figure 1-29.RegionalOverview or of National the Internet ISPs architecture. (e.g. Telus, SaskTel, Bell) EECS 3213, F14 L2: Intro Internet 13 A common way to connect to an ISP is to use the phone line to your house, in which case your phone company is your ISP. DSL, short for Digital Subscriber Line,reusesthetelephonelinethatconnectsto your house for digital data transmission. The computer is connected to a device called a DSL modem that converts between digital packets and analog signals that can pass unhindered over the telephone line. At the other end, a device called a DSLAM (Digital Sub- scriber Line Access Multiplexer)convertsbetweensignals and packets. Several other popular ways toTier-1connect toISPs an ISP are shown in Fig. 1-29. DSL is a higher-bandwidth way to use the local telephone line than to send bits over a traditional telephone call instead of avoiceconversation.1.That Level is called 3 dial-up and done with a different kind of modem at both ends. The2. word Globalmodem Crossingis short for ‘‘modulator demodulator’’ and refers to any device that3. converts NTT between digital bits and analog signals. 4. Sprint Another method is to send signals over the cable TV system. Like DSL, this 5. TeliaSonera is a way to reuse existing infrastructure, in this case otherwise unused cable TV 6. Tinet 7. Tata 8. Cogent 9. Verizon 10. AT&T

• Google, Facebook, etc. also setting up private pathways to move data between centers

EECS 3213, F14 L2: Intro Internet 14

7 Local Area Networks

334 THE MEDIUM ACCESS CONTROL SUBLAYER CHAP. 4 • A major component of the internet 334are concentratedTHE MEDIUM ACCESS CONTROL SUBLAYER CHAP. 4 4.8.2 Learning Bridges networks of computers 4.8.2 Learning Bridges – university, business The topology of two LANs bridged together is shown in Fig. 4-41 for two The topology of two LANs bridged together is shown in Fig. 4-41 for two cases. On the left-hand side, two• multidropThese LANs,simpler such networks as classic Ethernets, interface are tocases. the On internet the left-hand via side, two multidrop LANs, such as classic Ethernets, are joined by a special station—the bridge—thatrouters sits but on bothwhat LANs. happens On the right-hand inside? joined by a special station—the bridge—that sits on both LANs. On the right-hand side, LANs with point-to-point cables, including one hub, are joined together. The side, LANs with point-to-point cables, including one hub, are joined together. The bridges are the devices to which• theBasic stations components and hub are attached. If the LAN bridges are the devices to which the stations and hub are attached. If the LAN technology is Ethernet, the bridges are better known as Ethernet switches. technology is Ethernet, the bridges are– hubsbetter known as Ethernet switches. A D A D – bridges/switches Hub Port A Port D Hub E 1 1 Port E Port A D 2 2 E 1 1 B B1 B B1 B2 H1 E 1 2 4 4 2 2 3 3 F B B1 B B1 B2 H1 F 1 2 3 4 4 3 C G F Bridge Bridge F C G C G Bridge Bridge (a) (b) C G EECS 3213, F14 L2: Intro Internet Figure 4-41. (a) Bridge connecting15 two multidrop LANs. (b) Bridges (and a (a) (b) hub) connecting seven point-to-point stations.

Figure 4-41. (a) Bridge connecting two multidrop LANs. (b) Bridges (and a Bridges were developed when classic Ethernets were in use, so they are often hub) connecting seven point-to-point stations. shown in topologies with multidrop cables, as in Fig. 4-41(a). However, all the topologies that are encountered today are comprised of point-to-point cables and switches. The bridges work the same way in both settings. All of the stations at- Bridges were developed when classic Ethernets were in use, so they are often tached to the same port on a bridge belong to the same collision domain, and this shown in topologies with multidrop cables, as in Fig. 4-41(a). However, all the is different than the collision domain for other ports. If there is more than one sta- topologies that are encountered today are comprised of point-to-pointPopular cables LANs and tion, as in a classic Ethernet, a hub, or a half-duplex link, the CSMA/CD protocol switches. The bridges work the same way in both settings. All of the stations at- is used to send frames. tached to the same port on a bridge belong to the same collision domain, and this There is a difference, however, in how the bridged LANs are built. To bridge is different than the collision domain for other ports. If there is more than one sta- multidrop LANs, a bridge is added as a new station on each of the multidrop • IEEE 802.11 (WiFi) & IEEE 802.3 (LANs,Ethernet as in Fig.) 4-41(a). To bridge point-to-point LANs, the hubs are either con- tion, as in a classic Ethernet, a hub, or a half-duplex link, the CSMA/CD protocol nected to a bridge or, preferably, replaced with a bridge to increase performance. is used to send frames. – Best-effort connectionless service In Fig. 4-41(b), bridges have replaced all but one hub. INTRODUCTION CHAP. 1 There is a difference, however,20 in how the bridged LANs are built. To bridge Different kinds of cables can also be attached to one bridge. For example, the multidrop LANs, a bridge is added as a new station on each of the multidrop cable connecting bridge B1 to bridge B2 in Fig. 4-41(b) might be a long-distance LANs, as in Fig. 4-41(a). To bridge point-to-point LANs, the hubs are either con- fiber optic link, while the cable connecting the bridges to stations might be a Access To wired network short-haulEthernet twisted-pair line. This arrangement is useful for bridging LANs in dif- nected to a bridge or, preferably, replaced with a pointbridge to increase performance.Portsferent buildings.switch To rest of In Fig. 4-41(b), bridges have replaced all but one hub. Now let us considernetwork what happens inside the bridges. Each bridge operates in Different kinds of cables can also be attached to one bridge. For example, the promiscuous mode, that is, it accepts every frame transmitted by the stations cable connecting bridge B1 to bridge B2 in Fig. 4-41(b) might be a long-distance fiber optic link, while the cable connecting the bridges to stations might be a short-haul twisted-pair line. This arrangement is useful for bridging LANs in different buildings. Now let us consider what happens inside the bridges. Each bridge operates in promiscuous mode, that is, it accepts every frame transmitted by the stations

Figure 1-8. Wireless and wired LANs. (a) 802.11. (b) Switched Ethernet.

EECS 3213, F14 L2: Intro Internet 16 to hundreds of Mbps. (In this book we will adhere to tradition and measure line speeds in megabits/sec, where 1 Mbps is 1,000,000 bits/sec, and gigabits/sec, where 1 Gbps is 1,000,000,000 bits/sec.) We will discuss 802.11 in Chap. 4. Wired LANs use a range of different transmission technologies. Most of them use copper wires, but some use optical fiber. LANs are restricted in size, which means that the worst-case transmission time is bounded and known in ad- vance. Knowing these bounds helps with the task of designing network protocols. Typically, wired LANs run at speeds of 100 Mbps to 1 Gbps, have low delay 8 (microseconds or nanoseconds), and make very few errors. Newer LANs can op- erate at up to 10 Gbps. Compared to wireless networks, wired LANs exceed them in all dimensions of performance. It is just easier to send signals over a wire or through a fiber than through the air. The topology of many wired LANs is built from point-to-point links. IEEE 802.3, popularly called Ethernet, is, by far, the most common type of wired LAN. Fig. 1-8(b) shows a sample topology of switched Ethernet. Each computer speaks the Ethernet protocol and connects to a box called a switch with a point-to-point link. Hence the name. Aswitchhasmultipleports, each of which can connect to one computer. The job of the switch is to relay packets between computers that are attached to it, using the address in each packet to determine which computer to send it to. To build larger LANs, switches can be plugged into each other using their ports. What happens if you plug them together in a loop? Will the network still work? Luckily, the designers thought of this case. It is the job of the protocol to sort out what paths packets should travel to safely reach the intended computer. We will see how this works in Chap. 4. It is also possible to divide one large physical LAN into two smaller logical LANs. You might wonder why this would be useful. Sometimes, the layout of the network equipment does not match the organization’s structure. For example, the Medium Access Control (MAC)

• A common challenge: communicating with multiple nodes over a shared medium • Medium Access Controls for sharing were developed • 20Example: Polling protocolINTRODUCTION on a multidrop line CHAP. 1

Access To wired network Ethernet point Ports switch To rest of network

EECS 3213, F14 Figure 1-8. Wireless and wiredL2: Intro LANs. Internet (a) 802.11. (b) Switched Ethernet. 17

to hundreds of Mbps. (In this book we will adhere to tradition and measure line speeds in megabits/sec, where 1 Mbps is 1,000,000 bits/sec, and gigabits/sec, where 1 Gbps is 1,000,000,000 bits/sec.) We will discuss 802.11 in Chap. 4. Wired LANs use a range of different transmission technologies. Most of them use copper wires, but some use optical fiber. LANs are restricted in size, which means that the worst-caseLAN transmission Addressing time is bounded and known in ad- vance. Knowing these bounds helps with the task of designing network protocols. Typically, wired LANs run at speeds of 100 Mbps to 1 Gbps, have low delay • (microsecondsHow do LANs or nanoseconds), identify themselves? and make very few errors. Newer LANs can op- erate at up to 10 Gbps. Compared to wireless networks, wired LANs exceed them – If they share a medium some means of identification is necessary in all dimensions of performance. It is just easier to send signals over a wire or • throughGlobally a fiber unique than through address the air. The topology of many wired LANs is built from point-to-point links. IEEE – MAC address, MAC-48, physical address 802.3, popularly called Ethernet, is, by far, the most common type of wired LAN.– consists Fig. 1-8(b) of 48-bits shows a sample topology of switched Ethernet. Each computer– burned speaks inside the Ethernet network protocol interface and connectscard (NIC to) a box called a switch with a point-to-point link. Hence the name. Aswitchhasmultipleports, each of which • can How connect does to one thiscomputer. work with The jobIP? of the switch is to relay packets between computers– The layering that are concept attached to it, using the address in each packet to determine which computer to send it to. To build larger LANs, switches can be plugged into each other using their ports. What happens if you plug them together in a loop? Will the network still work? Luckily, the designers thought of this case. It is the job of the protocol to sort out what paths packets should travel to safely reach the intended computer. We will see how this works in Chap. 4. It is also possible to divide one large physical LAN into two smaller logical LANs. You might wonder why this would be useful. Sometimes, the layout of the EECS 3213,network F14 equipment does not matchL2: theIntro organization’s Internet structure. For example,18 the

9 Summary of Some Network Terms

• connectionless – Send to source before you know that source is accepting • connection-oriented – Send to source only after you hear that it is willing to accept • packet-switching – Non-dedicated link to source made on fly for each chunk of message • circuit-switching – Dedicated link created to source for duration of message • best-effort service – not guaranteed • datagram service – unacknowledged connectionless service

EECS 3213, F14 L2: Intro Internet 19