Chapter 5 The Data Link Layer Objectives q

Chapter 5: The Data Link Layer Objectives: q understand principles behind data link layer services: v v error detection, correction sharing a broadcast channel: multiple access link layer addressing reliable data transfer, flow control: done! q instantiation and implementation of various link layer technologies 1

Link Layer Services q Framing, link access: v encapsulate datagram into frame, adding header, trailer v channel access if shared medium v ‘physical addresses’ used in frame headers to identify source, dest • different from IP address! q Reliable delivery between adjacent nodes v very similar to the network-layer reliable service v seldom used on low bit error link (fiber, some twisted pair) v wireless links: high error rates • Q: why both link-level and end-end reliability? 2

Link Layer Services (more) q Flow Control: v pacing between adjacent sending and receiving nodes v similar flow control mechanisms as the transport layer q Error Detection: v errors caused by signal attenuation, noise. v receiver detects presence of errors: • signals sender for retransmission or drops frame q Error Correction: v receiver identifies and corrects bit error(s) without resorting to retransmission q Half-duplex and full-duplex v with half duplex, nodes at both ends of link can transmit, but not at the same time 3

Adaptors Communicating datagram sending node frame adapter q link layer implemented in “adaptor” (i. e. NIC) v receiving node link layer protocol Ethernet card, PCMCIA card, 802. 11 card q sending side: v encapsulates datagram in a frame v adds error checking bits, rdt, flow control, etc. frame adapter q receiving side v looks for errors, rdt, flow control, etc v extracts datagram, passes to receiving node q adapter is semi- autonomous q link & physical layers 4

Error Detection EDC= Error Detection and Correction bits (redundancy) D = Data protected by error checking, may include header fields • Error detection not 100% reliable! • protocol may miss some errors, but rarely • larger EDC field yields better detection and correction 5

Parity Checking Single Bit Parity: Detect single bit errors Two Dimensional Bit Parity: Detect and correct single bit errors 0 0 6

Internet checksum Goal: detect “errors” (e. g. , flipped bits) in transmitted segment (note: used at transport layer only) Sender: q treat segment contents as sequence of 16 -bit integers q checksum: addition (1’s complement sum) of segment contents q sender puts checksum value into UDP checksum field Receiver: q compute checksum of received segment q check if computed checksum equals checksum field value: v NO - error detected v YES - no error detected. But maybe errors nonetheless? More later when we present CRC …. 7

Checksumming: Cyclic Redundancy Check q view data bits, D, as a binary number q choose r+1 bit pattern (generator), G q goal: choose r CRC bits, R, such that v v v <D, R> exactly is divisible by G (modulo 2) receiver knows G, divides <D, R> by G. If non-zero remainder: error detected! can detect all burst errors less than r+1 bits q widely used in practice (ATM, HDLC) 8

CRC Example Want: D. 2 r XOR R = n. G equivalently: D. 2 r = n. G XOR R equivalently: if we divide D. 2 r by G, want remainder R R = remainder[ D. 2 r G ] 9

Multiple Access Links and Protocols Two types of “links”: q point-to-point v PPP for dial-up access v point-to-point link between Ethernet switch and host q broadcast (shared wire or medium) v traditional Ethernet v upstream HFC v 802. 11 wireless LAN 10

Multiple Access protocols q single shared broadcast channel q two or more simultaneous transmissions by nodes: interference v only one node can send successfully at a time multiple access protocol q distributed algorithm that determines how nodes share channel, i. e. , determine when node can transmit q communication about channel sharing must use channel itself! q claim: humans use multiple access protocols all the time 11

Ideal Mulitple Access Protocol Broadcast channel of rate R bps 1. When one node wants to transmit, it can send at rate R. 2. When M nodes want to transmit, each can send at average rate R/M 3. Fully decentralized: v v no special node to coordinate transmissions no synchronization of clocks, slots 4. Simple 12

Multiple Access (MA) Protocols: a taxonomy Three broad classes: q Channel Partitioning v v divide channel into smaller “pieces” (time slots, frequency, code) allocate piece to node for exclusive use q Random Access v channel not divided, allow collisions v “recover” from collisions q “Taking turns” v tightly coordinate shared access to avoid collisions 13

Channel Partitioning MA protocols: TDMA: time division multiple access q access to channel in "rounds" q each station gets fixed length slot (length = pkt trans time) in each round q unused slots go idle q example: 6 -station LAN, 1, 3, 4 have pkt, slots 2, 5, 6 idle 14

Channel Partitioning MA protocols: FDMA: frequency division multiple access q channel spectrum divided into frequency bands q each station assigned fixed frequency band q unused transmission time in frequency bands go idle q example: 6 -station LAN, 1, 3, 4 have pkt, frequency bands 2, 5, 6 idle frequency bands time 15

Random Access Protocols q When node has packet to send v transmit at full channel data rate R. v no a priori coordination among nodes q two or more transmitting nodes -> “collision”, q random access MAC protocol specifies: v how to detect collisions v how to recover from collisions (e. g. , via delayed retransmissions) q Examples of random access MAC protocols: v slotted ALOHA v CSMA, CSMA/CD, CSMA/CA 16

Slotted Aloha q All frames have exactly L bits q time is divided into equal size slots (= pkt trans. time) q node with new arriving pkt: transmit at beginning of next slot q if collision: retransmit pkt in future slots with probability p, until successful Success (S), Collision (C), Empty (E) slots 17

$Slotted Aloha efficiency Q: what is max fraction slots successful? A: Suppose N stations$

Slotted Aloha efficiency Q: what is max fraction slots successful? A: Suppose N stations have packets to send v each transmits in slot with probability p v prob. successful transmission S is: by single node: S= p (1 -p)(N-1) by any of N nodes S = Prob (only one transmits) = N p (1 -p)(N-1) … choosing optimum p as n -> infty. . . = 1/e =. 37 as N -> infty At best: channel use for useful transmissions 37% of time! 18

Pure (unslotted) ALOHA q unslotted Aloha: simpler, no synchronization q pkt needs transmission: v send without awaiting for beginning of slot q collision probability increases: v pkt sent at t 0 collide with other pkts sent in [t 0 -1, t 0+1] 19

Pure Aloha (cont. ) P(success by given node) = P(node transmits). P(no other node transmits in [t 0 -1, t 0]. P(no other node transmits in [t 0, t 0+1 ] = p. (1 -p) P(success by any of N nodes) = N p. (1 -p) S = throughput = “goodput” (success rate) … choosing optimum p as n -> infty. . . = 1/(2 e) =. 18 0. 4 0. 3 Slotted Aloha 0. 2 0. 1 protocol constrains effective channel throughput! Pure Aloha 0. 5 1. 0 1. 5 2. 0 G = offered load = Np 20

CSMA: Carrier Sense Multiple Access CSMA: listen before transmit: q If channel sensed idle: transmit entire pkt q If channel sensed busy, defer transmission v Persistent CSMA: retry immediately with probability p when channel becomes idle (may cause instability) v Non-persistent CSMA: retry after random interval q human analogy: don’t interrupt others! 21

CSMA collisions spatial layout of nodes along Ethernet Role of distance and propagation delay is crucial in determining collision prob. Propagation delay means two nodes may not hear each other’s transmission Collision: entire packet transmission time wasted 22

CSMA/CD (Collision Detection) CSMA/CD: carrier sensing, but collisions detected within short time v colliding transmissions aborted, reducing channel wastage v persistent or non-persistent retransmission v q collision detection: v easy in wired LANs: measure signal strengths, compare transmitted, received signals v difficult in wireless LANs: receiver shut off while transmitting q human analogy: the polite conversationalist 23

CSMA/CD collision detection 24

“Taking Turns” Multiple Access protocols channel partitioning MA protocols: v share channel efficiently and fairly at high load v inefficient at low load: delay in channel access, 1/N bandwidth allocated even if only 1 active node! Random access MA protocols v efficient at low load: single node can fully utilize channel v high load: collision overhead “taking turns” protocols look for best of both worlds! 25

“Taking Turns” MAC protocols Token passing: Polling: q control token passed from q master node one node to next “invites” slave nodes sequentially. to transmit in turn q token message q concerns: v polling overhead v v latency single point of failure (master) v v v token overhead latency single point of failure (token) 26

Reservation-based protocols Distributed Polling: q time divided into slots q begins with N short reservation slots reservation slot time equal to channel end-end propagation delay v station with message to send posts reservation v reservation seen by all stations q after reservation slots, message transmissions ordered by v known priority 27

Summary of MAC protocols q What do you do with a shared media? v Channel Partitioning, by time, frequency or code • Time Division, Code Division, Frequency Division v Random partitioning (dynamic), • ALOHA, S-ALOHA, CSMA/CD • carrier sensing: easy in some technologies (wire), hard in others (wireless) • CSMA/CD used in Ethernet v Taking Turns • polling from a central site, token passing 28

Chapter 5: The Data Link Layer Objectives: q LAN technologies q link layer addressing, ARP q specific link layer technologies: v v addressing Gigabit Ethernet ATM Frame Relay 29

LAN technologies Data link layer so far: We talked about v services, error detection/correction, multiple access Next: LAN technologies addressing v Ethernet v ATM v Frame Relay v 30

LAN Addresses 32 -bit IP address: q network-layer address q used to get datagram to destination IP network (recall IP network definition) LAN (or MAC or physical or Ethernet) address: q used to get datagram from one interface to another physically-connected interface (same network) q 48 bit MAC address (for most LANs) burned in the adapter ROM 31

LAN Addresses Each adapter on LAN has unique LAN address 32

LAN Address (more) q MAC address allocation administered by IEEE q manufacturer buys portion of MAC address space (to assure uniqueness) q Analogy: (a) MAC address: like Social Security Number (b) IP address: like postal address q MAC flat address => portability v can move LAN card from one LAN to another q IP hierarchical address NOT portable v depends on IP network to which node is attached 33

Recall earlier routing discussion Starting at A, given IP datagram addressed to B: A 223. 1. 1. 1 223. 1. 2. 1 q look up net. address of B, find B on same net. as A q link layer send datagram to B inside link-layer frame source, dest address B’s MAC A’s MAC addr 223. 1. 1. 2 223. 1. 1. 4 223. 1. 2. 9 B 223. 1. 1. 3 datagram source, dest address A’s IP addr B’s IP addr 223. 1. 3. 27 223. 1. 2. 2 E 223. 1. 3. 2 IP payload datagram frame 34

ARP: Address Resolution Protocol Question: how to determine MAC address of B knowing B’s IP address? q Each IP node (Host or Router) on LAN has ARP table q ARP Table: IP/MAC address mappings for some LAN nodes < IP address; MAC address; TTL> v TTL (Time To Live): time after which address mapping will be forgotten (typically 20 min) 35

ARP protocol q A wants to send datagram to B, and A knows B’s IP address. q Suppose B’s MAC address is not in A’s ARP table. q A broadcasts ARP query packet, containing B's IP address v all machines on LAN receive ARP query q B receives ARP packet, replies to A with its (B's) MAC address v frame sent to A’s MAC address (unicast) q A caches (saves) IP-to- MAC address pair in its ARP table until information becomes old (times out) v soft state: information that times out (goes away) unless refreshed q ARP is “plug-and-play”: v nodes create their ARP tables without intervention from net administrator 36

Routing to another LAN walkthrough: send datagram from A to B via R assume A know’s B IP address A R B q Two ARP tables in router R, one for each IP network (LAN) 37

q A creates datagram with source A, destination B q A uses ARP to get R’s MAC address for 111. 110 q A creates link-layer frame with R's MAC address as dest, q q q frame contains A-to-B IP datagram A’s data link layer sends frame R’s data link layer receives frame R removes IP datagram from Ethernet frame, sees its destined to B R uses ARP to get B’s physical layer address R creates frame containing A-to-B IP datagram sends to B A R B 38

Ethernet “dominant” LAN technology: q cheap $20 for 100 Mbs! q first widely used LAN technology q Simpler, cheaper than token LANs and ATM q Kept up with speed race: 10, 1000 Mbps Metcalfe’s Ethernet sketch 39

Ethernet Frame Structure Sending adapter encapsulates IP datagram (or other network layer protocol packet) in Ethernet frame Preamble: q 7 bytes with pattern 1010 followed by one byte with pattern 10101011 q used to synchronize receiver, sender clock rates 40

Ethernet Frame Structure (more) q Addresses: 6 bytes v if adapter receives frame with matching destination address, or with broadcast address (eg ARP packet), it passes data in frame to net-layer protocol v otherwise, adapter discards frame q Type: indicates the higher layer protocol, mostly IP but others may be supported such as Novell IPX and Apple. Talk) q CRC: checked at receiver, if error is detected, the frame is simply dropped 41

Unreliable, connectionless service q Connectionless: No handshaking between sending and receiving adapter. q Unreliable: receiving adapter doesn’t send acks or nacks to sending adapter v v v stream of datagrams passed to network layer can have gaps will be filled if app is using TCP otherwise, app will see the gaps 42

Ethernet uses CSMA/CD q No slots q adapter doesn’t transmit if it senses that some other adapter is transmitting, that is, carrier sense q transmitting adapter aborts when it senses that another adapter is transmitting, that is, collision detection q Before attempting a retransmission, adapter waits a random time, that is, random access 43

Ethernet: uses CSMA/CD A: sense channel, if idle then { transmit and monitor the channel; If detect another transmission then { abort and send jam signal; update # collisions; delay as required by exponential backoff algorithm; goto A } else {done with the frame; set collisions to zero} } else {wait until ongoing transmission is over and goto A} 44

Ethernet’s CSMA/CD (more) Jam Signal: make sure all other transmitters are aware of collision; 48 bits; Exponential Backoff: q Goal: adapt retransmission attempts to estimated current load v heavy load: random wait will be longer q first collision: choose K from {0, 1}; delay is K x 512 bit transmission times q after second collision: choose K from {0, 1, 2, 3}… q after ten or more collisions, choose K from {0, 1, 2, 3, 4, …, 1023} 45

Ethernet Technologies: 10 Base 2 q 10: 10 Mbps; 2: under 200 meters max cable length q thin coaxial cable in a bus topology q repeaters used to connect up to multiple segments q repeater repeats bits it hears on one interface to its other interfaces: physical layer device only! q has become a legacy technology 46

10 Base. T and 100 Base. T q 10/100 Mbps rate; latter called “fast ethernet” q T stands for Twisted Pair q Nodes connect to a hub: “star topology”; 100 m max distance between nodes and hub nodes hub q Hubs are essentially physical-layer repeaters: bits coming in one link go out all other links v no frame buffering v no CSMA/CD at hub: adapters detect collisions v provides net management functionality v 47

Gbit Ethernet q use standard Ethernet frame format q allows for point-to-point links and shared broadcast q q channels in shared mode, CSMA/CD is used; short distances between nodes to be efficient In Gbit Ethernet terminology, hubs are called “Buffered Distributors” Full-Duplex at 1 Gbps for point-to-point links 10 Gbps now ! 48

Asynchronous Transfer Mode: ATM q 1990’s/00 standard for high-speed (155 Mbps to 622 Mbps and higher) Broadband Integrated Service Digital Network architecture q Goal: integrated, end-end transport of carry voice, video, data v meeting timing/Qo. S requirements of voice, video (versus Internet best-effort model) v packet-switching (fixed length packets, called “cells”) using virtual circuits 49

ATM architecture q adaptation layer: only at edge of ATM network data segmentation/reassembly v roughly analagous to Internet transport layer q ATM layer: “network” layer v cell switching, routing q physical layer v 50

ATM: network or link layer? Vision: end-to-end transport: “ATM from desktop to desktop” v ATM is a network technology Reality: used to connect IP backbone routers v “IP over ATM” v ATM as switched link layer, connecting IP routers 51

ATM Adaptation Layer (AAL) q ATM Adaptation Layer (AAL): “adapts” upper layers (IP or native ATM applications) to ATM layer below q AAL present only in end systems, not in ATM switches q AAL layer segment (header/trailer fields, data) fragmented across multiple ATM cells v analogy: TCP segment in many IP packets 52

ATM Adaptation Layer (AAL) [more] Different versions of AAL layers, depending on ATM service class: q AAL 1: for CBR (Constant Bit Rate) services, e. g. circuit emulation q AAL 2: for VBR (Variable Bit Rate) services, e. g. , MPEG video q AAL 5: for data (eg, IP datagrams) User data AAL PDU ATM cell 53

ATM Layer: Virtual Circuits q VC transport: cells carried on VC from source to dest v call setup, teardown for each call before data can flow v each packet carries VC identifier (not destination ID) v every switch on source-dest path maintain “state” for each passing connection v link, switch resources (bandwidth, buffers) may be allocated to VC: to get circuit-like performance q Permanent VCs (PVCs) long lasting connections v typically: “permanent” route to IP routers q Switched VCs (SVC): v dynamically set up on per-call basis v 54

ATM VCs q Advantages of ATM VC approach: Qo. S performance guarantee for connection mapped to VC (bandwidth, delay jitter) q Drawbacks of ATM VC approach: v Inefficient support of datagram traffic v one PVC between each source/dest pair) does not scale (N*2 connections needed) v SVC introduces call setup latency, processing overhead for short lived connections v 55

ATM Layer: ATM cell q 5 -byte ATM cell header q 48 -byte payload Why? : small payload -> short cell-creation delay for digitized voice v halfway between 32 and 64 (compromise!) v Cell header Cell format 56

ATM Physical Layer Two pieces (sublayers) of physical layer: q Transmission Convergence Sublayer (TCS): adapts ATM layer above to PMD sublayer below q Physical Medium Dependent (PMD): depends on physical medium being used TCS Functions: v Header checksum generation: 8 bits CRC v Cell delineation v With “unstructured” PMD sublayer, transmission of idle cells when no data cells to send 57

ATM Physical Layer (more) Physical Medium Dependent (PMD) sublayer functions q SONET/SDH: transmission frame structure (like a container carrying bits); v bit synchronization; v bandwidth partitions (TDM); v several standardized speeds q TI/T 3: transmission frame structure (old telephone hierarchy): 1. 5 Mbps/ 45 Mbps q unstructured: just cells (busy/idle) 58

X. 25 and Frame Relay q Wide Area Network technologies (like ATM); also, both Virtual Circuit oriented , like ATM q X. 25 was born in mid ‘ 70 s q Frame relay emerged in late ‘ 80 s q Both X. 25 and Frame Relay can be used to carry IP datagrams; q Thus, they are viewed as Link Layers by the IP protocol layer 59

X. 25 q X. 25 builds a VC between source and destination for each user connection q Along the path, error control (with retransmissions) on each hop q Also, on each VC, hop by hop flow control v congestion arising at an intermediate node propagates to source via backpressure 60

X. 25 q As a result, packets are delivered reliably and in sequence to destination q Putting “intelligence into the network” made sense in mid 70 s (dumb terminals without TCP) q Today, TCP and practically error free fibers favor pushing the “intelligence to the edges q As a result, X. 25 is rapidly becoming extinct 61

Frame Relay q Designed in late ‘ 80 s, widely deployed in the ‘ 90 s q Frame relay service: no error control v end-to-end congestion control v 62

Frame Relay (more) q Designed to interconnect corporate customer LANs typically permanent VC’s: “pipe” carrying aggregate traffic between two routers v switched VC’s: as in ATM q corporate customer leases FR service from public Frame Relay network (eg, Sprint, ATT) v 63

Frame Relay -VC Rate Control q Committed Information Rate (CIR) defined, “guaranteed” for each VC v negotiated at VC set up time v customer pays based on CIR v q DE bit: Discard Eligibility bit Edge FR switch measures traffic rate for each VC; marks DE bit v DE = 0: high priority, rate compliant frame; deliver at “all costs” v DE = 1: low priority, eligible for congestion discard v 64

Link Layer: Summary q principles behind data link layer services: v error detection, correction v sharing a broadcast channel: multiple access v link layer addressing v link layer technologies: Ethernet, ATM, Frame Relay q We have finished journey down the protocol stack 65