Chapter 3 Transport Layer

A note on the use of these Powerpoint slides: We’re making these slides freely available to all (faculty, students, readers). They’re in PowerPoint form so you see the animations; and can add, modify, and delete slides (including this one) and slide content to suit your needs. They obviously represent a lot of work on our part. In return for use, we only ask the following: Computer § If you use these slides (e.g., in a class) that you mention their source (after all, we’d like people to use our book!) Networking: A Top § If you post any slides on a www site, that you note that they are adapted from (or perhaps identical to) our slides, and note our copyright of this Down Approach material. 7th edition Thanks and enjoy! JFK/KWR Jim Kurose, Keith Ross All material copyright 1996-2016 Pearson/Addison Wesley J.F Kurose and K.W. Ross, All Rights Reserved April 2016 Transport Layer 2-1 Chapter 3: Transport Layer our goals: § understand principles § learn about Internet behind transport transport layer protocols: layer services: • UDP: connectionless • multiplexing, transport demultiplexing • TCP: connection-oriented • reliable data transfer reliable transport • flow control • TCP congestion control • congestion control

Transport Layer 3-2 Chapter 3 outline

3.1 transport-layer 3.5 connection-oriented services transport: TCP 3.2 multiplexing and • segment structure demultiplexing • reliable data transfer 3.3 connectionless • flow control transport: UDP • connection management 3.4 principles of reliable 3.6 principles of congestion data transfer control 3.7 TCP congestion control

Transport Layer 3-3 Transport services and protocols

application transport § provide logical communication network data link between app processes physical running on different hosts § transport protocols run in end systems • send side: breaks app messages into segments, passes to network layer • rcv side: reassembles application segments into messages, transport network passes to app layer data link physical § more than one transport protocol available to apps • Internet: TCP and UDP

Transport Layer 3-4 Transport vs. network layer

§ network layer: logical communication household analogy: 12 kids in Ann’s house sending between hosts letters to 12 kids in Bill’s § transport layer: house: logical § hosts = houses communication § processes = kids between processes § app messages = letters in envelopes • relies on, enhances, § transport protocol = Ann network layer and Bill who demux to in- services house siblings § network-layer protocol = postal service

Transport Layer 3-5 Internet transport-layer protocols

application transport § reliable, in-order network data link delivery (TCP) physical network network data link • congestion control data link physical physical network • flow control data link • connection setup physical network data link § unreliable, unordered physical network delivery: UDP data link physical network • no-frills extension of data link application physical “best-effort” IP network transport data link network physical data link § services not available: physical • delay guarantees • bandwidth guarantees

Transport Layer 3-6 Chapter 3 outline

Transport Layer 3-7 Multiplexing/demultiplexing

multiplexing at sender: handle data from multiple demultiplexing at receiver: sockets, add transport header use header info to deliver (later used for demultiplexing) received segments to correct socket

application

application P1 P2 application socket P3 P4 transport process transport network transport network link network link physical link physical physical

Transport Layer 3-8 How demultiplexing works

§ host receives IP datagrams 32 bits • each datagram has source IP source port # dest port # address, destination IP address other header fields • each datagram carries one transport-layer segment • each segment has source, application destination port number data § host uses IP addresses & (payload) port numbers to direct segment to appropriate socket TCP/UDP segment format

Transport Layer 3-9 Connectionless demultiplexing

§ recall: created socket has § recall: when creating host-local port #: datagram to send into UDP DatagramSocket mySocket1 socket, must specify = new DatagramSocket(12534); • destination IP address • destination port #

§ when host receives UDP IP datagrams with same segment: dest. port #, but different • checks destination port # source IP addresses in segment and/or source port numbers will be directed • directs UDP segment to to same socket at dest socket with that port #

Transport Layer 3-10 Connectionless demux: example

DatagramSocket DatagramSocket serverSocket = new DatagramSocket DatagramSocket mySocket2 = new mySocket1 = new DatagramSocket (6428); DatagramSocket (9157); application (5775); application P1 application P3 P4 transport transport transport network network link network link physical link physical physical

source port: 6428 source port: ? dest port: 9157 dest port: ?

source port: 9157 source port: ? dest port: 6428 dest port: ?

Transport Layer 3-11 Connection-oriented demux

§ TCP socket identified § server host may support by 4-tuple: many simultaneous TCP • source IP address sockets: • source port number • each socket identified by • dest IP address its own 4-tuple • dest port number § web servers have § demux: receiver uses all different sockets for four values to direct each connecting client segment to appropriate • non-persistent HTTP will socket have different socket for each request

Transport Layer 3-12 Connection-oriented demux: example

application application P4 P5 P6 application P3 P2 P3 transport transport transport network network link network link physical link physical server: IP physical address B

host: IP source IP,port: B,80 host: IP address A dest IP,port: A,9157 source IP,port: C,5775 address C dest IP,port: B,80 source IP,port: A,9157 dest IP, port: B,80 source IP,port: C,9157 dest IP,port: B,80 three segments, all destined to IP address: B, dest port: 80 are demultiplexed to different sockets Transport Layer 3-13 Connection-oriented demux: example

threaded server

application application application P4 P3 P2 P3 transport transport transport network network link network link physical link physical server: IP physical address B

Transport Layer 3-14 Chapter 3 outline

Transport Layer 3-15 UDP: User Datagram Protocol [RFC 768]

§ “no frills,” “bare bones” § UDP use: Internet transport § streaming multimedia protocol apps (loss tolerant, rate § “best effort” service, UDP sensitive) segments may be: § DNS • lost § SNMP • delivered out-of-order § reliable transfer over to app UDP: § connectionless: § add reliability at • no handshaking application layer between UDP sender, receiver § application-specific error recovery! • each UDP segment handled independently of others Transport Layer 3-16 UDP: segment header

length, in bytes of 32 bits UDP segment, source port # dest port # including header length checksum why is there a UDP? § no connection application establishment (which can data add delay) (payload) § simple: no connection state at sender, receiver § small header size UDP segment format § no congestion control: UDP can blast away as fast as desired

Transport Layer 3-17 UDP checksum

Goal: detect “errors” (e.g., flipped bits) in transmitted segment sender: receiver: § treat segment contents, § compute checksum of including header fields, received segment as sequence of 16-bit § integers check if computed checksum equals checksum field value: § checksum: addition (one’s complement sum) • NO - error detected of segment contents • YES - no error detected. § sender puts checksum But maybe errors value into UDP checksum nonetheless? More later field ….

Transport Layer 3-18 Internet checksum: example

example: add two 16-bit integers

1 1 1 1 0 0 1 1 0 0 1 1 0 0 1 1 0 1 1 1 0 1 0 1 0 1 0 1 0 1 0 1 0 1 wraparound 1 1 0 1 1 1 0 1 1 1 0 1 1 1 0 1 1

sum 1 1 0 1 1 1 0 1 1 1 0 1 1 1 1 0 0 checksum 1 0 1 0 0 0 1 0 0 0 1 0 0 0 0 1 1

Note: when adding numbers, a carryout from the most significant bit needs to be added to the result

* Check out the online interactive exercises for more examples: http://gaia.cs.umass.edu/kurose_ross/interactive/ Transport Layer 3-19 Chapter 3 outline

Transport Layer 3-20 Principles of reliable data transfer § important in application, transport, link layers • top-10 list of important networking topics!

§ characteristics of unreliable channel will determine complexity of reliable data transfer protocol (rdt)

Transport Layer 3-21 Principles of reliable data transfer § important in application, transport, link layers • top-10 list of important networking topics!

§ characteristics of unreliable channel will determine complexity of reliable data transfer protocol (rdt)

Transport Layer 3-22 Principles of reliable data transfer § important in application, transport, link layers • top-10 list of important networking topics!

§ characteristics of unreliable channel will determine complexity of reliable data transfer protocol (rdt)

Transport Layer 3-23 Reliable data transfer: getting started rdt_send(): called from above, deliver_data(): called by (e.g., by app.). Passed data to rdt to deliver data to upper deliver to receiver upper layer

send receive side side

udt_send(): called by rdt, rdt_rcv(): called when packet to transfer packet over arrives on rcv-side of channel unreliable channel to receiver

Transport Layer 3-24 Reliable data transfer: getting started we’ll: § incrementally develop sender, receiver sides of reliable data transfer protocol (rdt) § consider only unidirectional data transfer • but control info will flow on both directions! § use finite state machines (FSM) to specify sender, receiver event causing state transition actions taken on state transition state: when in this “state” next state state state uniquely determined event 1 2 by next event actions

Transport Layer 3-25 rdt1.0: reliable transfer over a reliable channel

§ underlying channel perfectly reliable • no bit errors • no loss of packets § separate FSMs for sender, receiver: • sender sends data into underlying channel • receiver reads data from underlying channel

Wait for rdt_send(data) Wait for rdt_rcv(packet) call from call from extract (packet,data) above packet = make_pkt(data) below deliver_data(data) udt_send(packet)

sender receiver

Transport Layer 3-26 rdt2.0: channel with bit errors

§ underlying channel may flip bits in packet • checksum to detect bit errors § the question: how to recover from errors: • acknowledgements (ACKs): receiver explicitly tells sender that pkt received OK • negative acknowledgements (NAKs): receiver explicitly tells sender that pkt had errors • senderHow retransmits do humans pkt on recover receipt fromof NAK “errors” § new mechanisms in rdt2.0 (beyond rdt1.0): • error detectionduring conversation? • receiver feedback: control msgs (ACK,NAK) rcvr- >sender

Transport Layer 3-27 rdt2.0: channel with bit errors

§ underlying channel may flip bits in packet • checksum to detect bit errors § the question: how to recover from errors: • acknowledgements (ACKs): receiver explicitly tells sender that pkt received OK • negative acknowledgements (NAKs): receiver explicitly tells sender that pkt had errors • sender retransmits pkt on receipt of NAK § new mechanisms in rdt2.0 (beyond rdt1.0): • error detection • feedback: control msgs (ACK,NAK) from receiver to sender

Transport Layer 3-28 rdt2.0: FSM specification

rdt_send(data) sndpkt = make_pkt(data, checksum) receiver udt_send(sndpkt) rdt_rcv(rcvpkt) && isNAK(rcvpkt) Wait for Wait for rdt_rcv(rcvpkt) && call from ACK or udt_send(sndpkt) corrupt(rcvpkt) above NAK udt_send(NAK)

rdt_rcv(rcvpkt) && isACK(rcvpkt) Wait for L call from sender below

rdt_rcv(rcvpkt) && notcorrupt(rcvpkt) extract(rcvpkt,data) deliver_data(data) udt_send(ACK)

Transport Layer 3-29 rdt2.0: operation with no errors

rdt_send(data) snkpkt = make_pkt(data, checksum) udt_send(sndpkt) rdt_rcv(rcvpkt) && isNAK(rcvpkt) Wait for Wait for rdt_rcv(rcvpkt) && call from ACK or udt_send(sndpkt) corrupt(rcvpkt) above NAK udt_send(NAK)

rdt_rcv(rcvpkt) && isACK(rcvpkt) Wait for L call from below

rdt_rcv(rcvpkt) && notcorrupt(rcvpkt) extract(rcvpkt,data) deliver_data(data) udt_send(ACK)

Transport Layer 3-30 rdt2.0: error scenario

rdt_rcv(rcvpkt) && isACK(rcvpkt) Wait for L call from below

rdt_rcv(rcvpkt) && notcorrupt(rcvpkt) extract(rcvpkt,data) deliver_data(data) udt_send(ACK)

Transport Layer 3-31 rdt2.0 has a fatal flaw! what happens if handling duplicates: ACK/NAK corrupted? § sender retransmits § sender doesn’t know current pkt if ACK/NAK what happened at corrupted receiver! § sender adds sequence ’ § can t just retransmit: number to each pkt possible duplicate § receiver discards (doesn’t deliver up) duplicate pkt stop and wait sender sends one packet, then waits for receiver response

Transport Layer 3-32 rdt2.1: sender, handles garbled ACK/NAKs

rdt_send(data) sndpkt = make_pkt(0, data, checksum) udt_send(sndpkt) rdt_rcv(rcvpkt) && ( corrupt(rcvpkt) || Wait for Wait for ACK or isNAK(rcvpkt) ) call 0 from udt_send(sndpkt) above NAK 0 rdt_rcv(rcvpkt) && notcorrupt(rcvpkt) rdt_rcv(rcvpkt) && isACK(rcvpkt) && notcorrupt(rcvpkt) && isACK(rcvpkt) L L Wait for Wait for ACK or call 1 from rdt_rcv(rcvpkt) && NAK 1 above ( corrupt(rcvpkt) || isNAK(rcvpkt) ) rdt_send(data) udt_send(sndpkt) sndpkt = make_pkt(1, data, checksum) udt_send(sndpkt)

Transport Layer 3-33 rdt2.1: receiver, handles garbled ACK/NAKs

rdt_rcv(rcvpkt) && notcorrupt(rcvpkt) && has_seq0(rcvpkt) extract(rcvpkt,data) deliver_data(data) sndpkt = make_pkt(ACK, chksum) udt_send(sndpkt) rdt_rcv(rcvpkt) && (corrupt(rcvpkt) rdt_rcv(rcvpkt) && (corrupt(rcvpkt) sndpkt = make_pkt(NAK, chksum) sndpkt = make_pkt(NAK, chksum) udt_send(sndpkt) udt_send(sndpkt) Wait for Wait for rdt_rcv(rcvpkt) && 0 from 1 from rdt_rcv(rcvpkt) && not corrupt(rcvpkt) && below below not corrupt(rcvpkt) && has_seq1(rcvpkt) has_seq0(rcvpkt) sndpkt = make_pkt(ACK, chksum) sndpkt = make_pkt(ACK, chksum) udt_send(sndpkt) udt_send(sndpkt) rdt_rcv(rcvpkt) && notcorrupt(rcvpkt) && has_seq1(rcvpkt)

extract(rcvpkt,data) deliver_data(data) sndpkt = make_pkt(ACK, chksum) udt_send(sndpkt)

Transport Layer 3-34 rdt2.1: discussion sender: receiver: § seq # added to pkt § must check if received § two seq. #’s (0,1) will packet is duplicate suffice. Why? • state indicates whether 0 or 1 is expected pkt § must check if received seq # ACK/NAK corrupted § note: receiver can not § twice as many states know if its last • state must ACK/NAK received “remember” whether OK at sender “expected” pkt should have seq # of 0 or 1

Transport Layer 3-35 rdt2.2: a NAK-free protocol

§ same functionality as rdt2.1, using ACKs only § instead of NAK, receiver sends ACK for last pkt received OK • receiver must explicitly include seq # of pkt being ACKed § duplicate ACK at sender results in same action as NAK: retransmit current pkt

Transport Layer 3-36 rdt2.2: sender, receiver fragments

rdt_send(data) sndpkt = make_pkt(0, data, checksum) udt_send(sndpkt) rdt_rcv(rcvpkt) && ( corrupt(rcvpkt) || Wait for Wait for isACK(rcvpkt,1) ) call 0 from ACK above 0 udt_send(sndpkt) sender FSM fragment rdt_rcv(rcvpkt) && notcorrupt(rcvpkt) rdt_rcv(rcvpkt) && && isACK(rcvpkt,0) (corrupt(rcvpkt) || L has_seq1(rcvpkt)) Wait for receiver FSM 0 from udt_send(sndpkt) below fragment

rdt_rcv(rcvpkt) && notcorrupt(rcvpkt) && has_seq1(rcvpkt) extract(rcvpkt,data) deliver_data(data) sndpkt = make_pkt(ACK1, chksum) udt_send(sndpkt) Transport Layer 3-37 rdt3.0: channels with errors and loss new assumption: approach: sender waits underlying channel can “reasonable” amount of also lose packets (data, time for ACK ACKs) § retransmits if no ACK • checksum, seq. #, received in this time ACKs, retransmissions § if pkt (or ACK) just delayed (not lost): will be of help … but not enough • retransmission will be duplicate, but seq. #’s already handles this • receiver must specify seq # of pkt being ACKed § requires countdown timer

Transport Layer 3-38 rdt3.0 sender

rdt_send(data) rdt_rcv(rcvpkt) && sndpkt = make_pkt(0, data, checksum) ( corrupt(rcvpkt) || udt_send(sndpkt) isACK(rcvpkt,1) ) rdt_rcv(rcvpkt) start_timer L L Wait Wait for timeout call 0from for udt_send(sndpkt) above ACK0 start_timer rdt_rcv(rcvpkt) && notcorrupt(rcvpkt) rdt_rcv(rcvpkt) && isACK(rcvpkt,1) && notcorrupt(rcvpkt) stop_timer && isACK(rcvpkt,0) stop_timer Wait Wait for timeout for call 1 from udt_send(sndpkt) ACK1 above start_timer rdt_rcv(rcvpkt) rdt_send(data) L rdt_rcv(rcvpkt) && ( corrupt(rcvpkt) || sndpkt = make_pkt(1, data, checksum) isACK(rcvpkt,0) ) udt_send(sndpkt) start_timer L

Transport Layer 3-39 rdt3.0 in action

sender receiver sender receiver send pkt0 pkt0 send pkt0 pkt0 rcv pkt0 rcv pkt0 ack0 send ack0 ack0 send ack0 rcv ack0 rcv ack0 send pkt1 pkt1 send pkt1 pkt1 rcv pkt1 X ack1 send ack1 loss rcv ack1 send pkt0 pkt0 rcv pkt0 timeout ack0 send ack0 resend pkt1 pkt1 rcv pkt1 ack1 send ack1 rcv ack1 send pkt0 pkt0 (a) no loss rcv pkt0 ack0 send ack0

(b) packet loss

Transport Layer 3-40 rdt3.0 in action sender receiver sender receiver send pkt0 pkt0 send pkt0 rcv pkt0 pkt0 send ack0 rcv pkt0 ack0 send ack0 rcv ack0 ack0 send pkt1 pkt1 rcv ack0 rcv pkt1 send pkt1 pkt1 rcv pkt1 send ack1 ack1 send ack1 ack1 X loss timeout resend pkt1 pkt1 timeout rcv pkt1 resend pkt1 pkt1 rcv ack1 pkt0 (detect duplicate) rcv pkt1 send pkt0 send ack1 (detect duplicate) ack1 ack1 send ack1 rcv ack1 rcv pkt0 rcv ack1 ack0 send ack0 send pkt0 pkt0 send pkt0 pkt0 rcv pkt0 rcv pkt0 ack0 (detect duplicate) ack0 send ack0 send ack0

Transport Layer 3-41 Performance of rdt3.0

§ rdt3.0 is correct, but performance stinks § e.g.: 1 Gbps link, 15 ms prop. delay, 8000 bit packet: L 8000 bits D = = = 8 microsecs trans R 109 bits/sec

§ U sender: utilization – fraction of time sender busy sending L / R .008 U = = = 0.00027 sender RTT + L / R 30.008 § if RTT=30 msec, 1KB pkt every 30 msec: 33kB/sec thruput over 1 Gbps link § network protocol limits use of physical resources!

Transport Layer 3-42 rdt3.0: stop-and-wait operation

sender receiver first packet bit transmitted, t = 0 last packet bit transmitted, t = L / R

first packet bit arrives RTT last packet bit arrives, send ACK

ACK arrives, send next packet, t = RTT + L / R

L / R .008 U = = = 0.00027 sender RTT + L / R 30.008

Transport Layer 3-43 Pipelined protocols pipelining: sender allows multiple, “in-flight”, yet- to-be-acknowledged pkts • range of sequence numbers must be increased • buffering at sender and/or receiver

§ two generic forms of pipelined protocols: go-Back-N, selective repeat

Transport Layer 3-44 Pipelining: increased utilization

sender receiver first packet bit transmitted, t = 0 last bit transmitted, t = L / R

first packet bit arrives RTT last packet bit arrives, send ACK last bit of 2nd packet arrives, send ACK last bit of 3rd packet arrives, send ACK ACK arrives, send next packet, t = RTT + L / R 3-packet pipelining increases utilization by a factor of 3!

3L / R .0024 U = = = 0.00081 sender RTT + L / R 30.008

Transport Layer 3-45 Pipelined protocols: overview

Go-back-N: Selective Repeat: § sender can have up to § sender can have up to N N unacked packets in unack’ed packets in pipeline pipeline § receiver only sends § rcvr sends individual ack cumulative ack for each packet • doesn’t ack packet if there’s a gap § sender has timer for § sender maintains timer oldest unacked packet for each unacked packet • when timer expires, • when timer expires, retransmit all unacked retransmit only that packets unacked packet

Transport Layer 3-46 Go-Back-N: sender

§ k-bit seq # in pkt header § “window” of up to N, consecutive unack’ed pkts allowed

§ ACK(n): ACKs all pkts up to, including seq # n - “cumulative ACK” • may receive duplicate ACKs (see receiver) § timer for oldest in-flight pkt § timeout(n): retransmit packet n and all higher seq # pkts in window

Transport Layer 3-47 GBN: sender extended FSM

rdt_send(data) if (nextseqnum < base+N) { sndpkt[nextseqnum] = make_pkt(nextseqnum,data,chksum) udt_send(sndpkt[nextseqnum]) if (base == nextseqnum) start_timer nextseqnum++ } L else refuse_data(data) base=1 nextseqnum=1 timeout Wait start_timer udt_send(sndpkt[base]) rdt_rcv(rcvpkt) udt_send(sndpkt[base+1]) && corrupt(rcvpkt) … udt_send(sndpkt[nextseqnum-1]) rdt_rcv(rcvpkt) && notcorrupt(rcvpkt) base = getacknum(rcvpkt)+1 If (base == nextseqnum) stop_timer else start_timer Transport Layer 3-48 GBN: receiver extended FSM

default

udt_send(sndpkt) rdt_rcv(rcvpkt) && notcurrupt(rcvpkt) L && hasseqnum(rcvpkt,expectedseqnum) expectedseqnum=1 Wait extract(rcvpkt,data) sndpkt = deliver_data(data) make_pkt(0,ACK,chksum) sndpkt = make_pkt(expectedseqnum,ACK,chksum) udt_send(sndpkt) expectedseqnum++ ACK-only: always send ACK for correctly-received pkt with highest in-order seq # • may generate duplicate ACKs • need only remember expectedseqnum § out-of-order pkt: • discard (don’t buffer): no receiver buffering! • re-ACK pkt with highest in-order seq #

Transport Layer 3-49 GBN in action sender window (N=4) sender receiver 0 1 2 3 4 5 6 7 8 send pkt0 0 1 2 3 4 5 6 7 8 send pkt1 receive pkt0, send ack0 0 1 2 3 4 5 6 7 8 send pkt2 receive pkt1, send ack1 0 1 2 3 4 5 6 7 8 send pkt3 Xloss (wait) receive pkt3, discard, 0 1 2 3 4 5 6 7 8 rcv ack0, send pkt4 (re)send ack1 0 1 2 3 4 5 6 7 8 rcv ack1, send pkt5 receive pkt4, discard, (re)send ack1 ignore duplicate ACK receive pkt5, discard, (re)send ack1 pkt 2 timeout 0 1 2 3 4 5 6 7 8 send pkt2 0 1 2 3 4 5 6 7 8 send pkt3 0 1 2 3 4 5 6 7 8 send pkt4 rcv pkt2, deliver, send ack2 0 1 2 3 4 5 6 7 8 send pkt5 rcv pkt3, deliver, send ack3 rcv pkt4, deliver, send ack4 rcv pkt5, deliver, send ack5

Transport Layer 3-50 Selective repeat

§ receiver individually acknowledges all correctly received pkts • buffers pkts, as needed, for eventual in-order delivery to upper layer § sender only resends pkts for which ACK not received • sender timer for each unACKed pkt § sender window • N consecutive seq #’s • limits seq #s of sent, unACKed pkts

Transport Layer 3-51 Selective repeat: sender, receiver windows

Transport Layer 3-52 Selective repeat sender receiver data from above: pkt n in [rcvbase, rcvbase+N-1] § if next available seq # in § send ACK(n) window, send pkt § out-of-order: buffer timeout(n): § in-order: deliver (also § resend pkt n, restart timer deliver buffered, in-order pkts), advance window to ACK(n) in [sendbase,sendbase+N]: next not-yet-received pkt § mark pkt n as received pkt n in [rcvbase-N,rcvbase-1] § if n smallest unACKed pkt, § ACK(n) advance window base to next unACKed seq # otherwise: § ignore

Transport Layer 3-53 Selective repeat in action sender window (N=4) sender receiver 0 1 2 3 4 5 6 7 8 send pkt0 0 1 2 3 4 5 6 7 8 send pkt1 receive pkt0, send ack0 0 1 2 3 4 5 6 7 8 send pkt2 receive pkt1, send ack1 0 1 2 3 4 5 6 7 8 send pkt3 Xloss (wait) receive pkt3, buffer, 0 1 2 3 4 5 6 7 8 rcv ack0, send pkt4 send ack3 0 1 2 3 4 5 6 7 8 rcv ack1, send pkt5 receive pkt4, buffer, send ack4 record ack3 arrived receive pkt5, buffer, send ack5 pkt 2 timeout 0 1 2 3 4 5 6 7 8 send pkt2 0 1 2 3 4 5 6 7 8 record ack4 arrived rcv pkt2; deliver pkt2, 0 1 2 3 4 5 6 7 8 record ack5 arrived 0 1 2 3 4 5 6 7 8 pkt3, pkt4, pkt5; send ack2

Q: what happens when ack2 arrives?

Transport Layer 3-54 sender window receiver window Selective repeat: (after receipt) (after receipt) dilemma 0 1 2 3 0 1 2 pkt0 0 1 2 3 0 1 2 pkt1 0 1 2 3 0 1 2 0 1 2 3 0 1 2 pkt2 0 1 2 3 0 1 2 example: 0 1 2 3 0 1 2 0 1 2 3 0 1 2 pkt3 § seq #’s: 0, 1, 2, 3 X 0 1 2 3 0 1 2 § window size=3 pkt0 will accept packet with seq number 0 § receiver sees no (a) no problem difference in two receiver can’t see sender side. scenarios! receiver behavior identical in both cases! ’ § duplicate data something s (very) wrong!

accepted as new in (b) 0 1 2 3 0 1 2 pkt0 0 1 2 3 0 1 2 pkt1 0 1 2 3 0 1 2 Q: what relationship 0 1 2 3 0 1 2 pkt2 0 1 2 3 0 1 2 X 0 1 2 3 0 1 2 between seq # size X timeout and window size to retransmit pkt0 X avoid problem in (b)? 0 1 2 3 0 1 2 pkt0 will accept packet with seq number 0 (b) oops!

Transport Layer 3-55 Chapter 3 outline

Transport Layer 3-56 TCP: Overview RFCs: 793,1122,1323, 2018, 2581

§ point-to-point: § full duplex data: • one sender, one receiver • bi-directional data flow § reliable, in-order byte in same connection stream: • MSS: maximum segment size • no “message boundaries” § connection-oriented: § pipelined: • handshaking (exchange of control msgs) inits • TCP congestion and sender, receiver state flow control set window before data exchange size § flow controlled: • sender will not overwhelm receiver

Transport Layer 3-57 TCP segment structure

32 bits URG: urgent data source port # dest port # counting (generally not used) by bytes sequence number ACK: ACK # of data valid acknowledgement number (not segments!) head not PSH: push data now len used UAPRS F receive window # bytes (generally not used) checksum Urg data pointer rcvr willing RST, SYN, FIN: options (variable length) to accept connection estab (setup, teardown commands) application Internet data checksum (variable length) (as in UDP)

Transport Layer 3-58 TCP seq. numbers, ACKs

outgoing segment from sender sequence numbers: source port # dest port # sequence number • byte stream “number” of acknowledgement number first byte in segment’s rwnd data checksum urg pointer window size acknowledgements: N • seq # of next byte

expected from other side sender sequence number space • cumulative ACK sent sent, not- usable not Q: how receiver handles ACKed yet ACKed but not usable out-of-order segments (“in- yet sent flight”) • A: TCP spec doesn’t say, incoming segment to sender - up to implementor source port # dest port # sequence number acknowledgement number A rwnd checksum urg pointer

Transport Layer 3-59 TCP seq. numbers, ACKs

Host A Host B

User types ‘C’ Seq=42, ACK=79, data = ‘C’ host ACKs receipt of ‘C’, echoes Seq=79, ACK=43, data = ‘C’ back ‘C’ host ACKs receipt of echoed ‘C’ Seq=43, ACK=80

simple telnet scenario

Transport Layer 3-60 TCP round trip time, timeout

Q: how to set TCP Q: how to estimate RTT? timeout value? § SampleRTT: measured time from segment § longer than RTT transmission until ACK • but RTT varies receipt § too short: premature • ignore retransmissions timeout, unnecessary § SampleRTT will vary, want retransmissions estimated RTT “smoother” • average several recent § too long: slow reaction measurements, not just to segment loss current SampleRTT

Transport Layer 3-61 TCP round trip time, timeout

EstimatedRTT = (1- a)*EstimatedRTT + a*SampleRTT § exponential weighted moving average § influence of past sample decreases exponentially fast

§ typical value: a = 0.125RTT: gaia.cs.umass.edu to fantasia.eurecom.fr

350 RTT: gaia.cs.umass.edu to fantasia.eurecom.fr

300

250

RTT (milliseconds) 200 RTT (milliseconds) sampleRTT 150 EstimatedRTT

100 1 8 15 22 29 36 43 50 57 64 71 78 85 92 99 106 time (seconnds) time (seconds) Transport Layer 3-62 SampleRTT Estimated RTT TCP round trip time, timeout

§ timeout interval: EstimatedRTT plus “safety margin” • large variation in EstimatedRTT -> larger safety margin § estimate SampleRTT deviation from EstimatedRTT: DevRTT = (1-b)*DevRTT + b*|SampleRTT-EstimatedRTT| (typically, b = 0.25)

TimeoutInterval = EstimatedRTT + 4*DevRTT

estimated RTT “safety margin”

* Check out the online interactive exercises for more examples: http://gaia.cs.umass.edu/kurose_ross/interactive/ Transport Layer 3-63 Chapter 3 outline

Transport Layer 3-64 TCP reliable data transfer

§ TCP creates rdt service on top of IP’s unreliable service • pipelined segments • cumulative acks let’s initially consider • single retransmission simplified TCP sender: timer • ignore duplicate acks § retransmissions • ignore flow control, triggered by: congestion control • timeout events • duplicate acks

Transport Layer 3-65 TCP sender events: data rcvd from app: timeout: § create segment with § retransmit segment seq # that caused timeout § seq # is byte-stream § restart timer number of first data ack rcvd: byte in segment § if ack acknowledges § start timer if not previously unacked already running segments • think of timer as for • update what is known oldest unacked to be ACKed segment • start timer if there are • expiration interval: still unacked segments TimeOutInterval

Transport Layer 3-66 TCP sender (simplified)

data received from application above create segment, seq. #: NextSeqNum pass segment to IP (i.e., “send”) NextSeqNum = NextSeqNum + length(data) if (timer currently not running) L start timer NextSeqNum = InitialSeqNum wait SendBase = InitialSeqNum for event timeout retransmit not-yet-acked segment with smallest seq. # start timer ACK received, with ACK field value y if (y > SendBase) { SendBase = y /* SendBase–1: last cumulatively ACKed byte */ if (there are currently not-yet-acked segments) start timer else stop timer } Transport Layer 3-67 TCP: retransmission scenarios

Host A Host B Host A Host B

SendBase=92 Seq=92, 8 bytes of data Seq=92, 8 bytes of data

Seq=100, 20 bytes of data ACK=100 timeout X timeout ACK=100 ACK=120

Seq=92, 8 bytes of data Seq=92, 8 SendBase=100 bytes of data SendBase=120 ACK=100 ACK=120

SendBase=120

lost ACK scenario premature timeout

Transport Layer 3-68 TCP: retransmission scenarios

Host A Host B

Seq=92, 8 bytes of data

Seq=100, 20 bytes of data ACK=100 X timeout ACK=120

Seq=120, 15 bytes of data

cumulative ACK

Transport Layer 3-69 TCP ACK generation [RFC 1122, RFC 2581]

event at receiver TCP receiver action

arrival of in-order segment with delayed ACK. Wait up to 500ms expected seq #. All data up to for next segment. If no next segment, expected seq # already ACKed send ACK

arrival of in-order segment with immediately send single cumulative expected seq #. One other ACK, ACKing both in-order segments segment has ACK pending

arrival of out-of-order segment immediately send duplicate ACK, higher-than-expect seq. # . indicating seq. # of next expected byte Gap detected

arrival of segment that immediate send ACK, provided that partially or completely fills gap segment starts at lower end of gap

Transport Layer 3-70 TCP fast retransmit

§ time-out period often relatively long: TCP fast retransmit • long delay before if sender receives 3 resending lost packet ACKs for same data § detect lost segments ((““tripletriple duplicateduplicate ACKsACKs””),), via duplicate ACKs. resend unacked • sender often sends segment with smallest many segments back- seq # to-back § likely that unacked • if segment is lost, there segment lost, so don’t will likely be many wait for timeout duplicate ACKs.

Transport Layer 3-71 TCP fast retransmit

Host A Host B

Seq=92, 8 bytes of data Seq=100, 20 bytes of data X

ACK=100 ACK=100

timeout ACK=100 ACK=100 Seq=100, 20 bytes of data

fast retransmit after sender receipt of triple duplicate ACK Transport Layer 3-72 Chapter 3 outline

Transport Layer 3-73 TCP flow control application application may process remove data from application TCP socket buffers …. TCP socket OS receiver buffers … slower than TCP receiver is delivering (sender is sending) TCP code

IP flow control code receiver controls sender, so sender won’t overflow receiver’s buffer by transmitting from sender too much, too fast receiver protocol stack

Transport Layer 3-74 TCP flow control

§ receiver “advertises” free buffer space by including to application process rwnd value in TCP header of receiver-to-sender segments RcvBuffer buffered data • RcvBuffer size set via socket options (typical default rwnd free buffer space is 4096 bytes) • many operating systems autoadjust RcvBuffer TCP segment payloads § sender limits amount of “ ” unacked ( in-flight ) data to receiver-side buffering receiver’s rwnd value § guarantees receive buffer will not overflow

Transport Layer 3-75 Chapter 3 outline

Transport Layer 3-76 Connection Management before exchanging data, sender/receiver “handshake”: § agree to establish connection (each knowing the other willing to establish connection) § agree on connection parameters

application application

connection state: ESTAB connection state: ESTAB connection variables: connection Variables: seq # client-to-server seq # client-to-server server-to-client server-to-client rcvBuffer size rcvBuffer size at server,client at server,client

network network

Socket clientSocket = Socket connectionSocket = newSocket("hostname","port welcomeSocket.accept(); number"); Transport Layer 3-77 Agreeing to establish a connection

2-way handshake: Q: will 2-way handshake always work in network? Let’s talk ESTAB § variable delays OK ESTAB § retransmitted messages (e.g. req_conn(x)) due to message loss § message reordering choose x ’ “ ” req_conn(x) § can t see other side ESTAB acc_conn(x) ESTAB

Transport Layer 3-78 Agreeing to establish a connection

2-way handshake failure scenarios:

choose x choose x req_conn(x) req_conn(x) ESTAB ESTAB retransmit acc_conn(x) retransmit acc_conn(x) req_conn(x) req_conn(x)

ESTAB ESTAB data(x+1) req_conn(x) accept retransmit data(x+1) data(x+1) connection connection client x completes server client x completes server terminates forgets x forgets x terminates req_conn(x)

ESTAB ESTAB data(x+1) accept half open connection! data(x+1) (no client!) Transport Layer 3-79 TCP 3-way handshake

client state server state LISTEN LISTEN choose init seq num, x send TCP SYN msg SYNSENT SYNbit=1, Seq=x choose init seq num, y send TCP SYNACK msg, acking SYN SYN RCVD SYNbit=1, Seq=y ACKbit=1; ACKnum=x+1 received SYNACK(x) ESTAB indicates server is live; send ACK for SYNACK; this segment may contain ACKbit=1, ACKnum=y+1 client-to-server data received ACK(y) indicates client is live ESTAB

Transport Layer 3-80 TCP 3-way handshake: FSM

closed

Socket connectionSocket = welcomeSocket.accept(); L Socket clientSocket = SYN(x) newSocket("hostname","port number"); SYNACK(seq=y,ACKnum=x+1) create new socket for SYN(seq=x) communication back to client listen

SYN SYN rcvd sent

SYNACK(seq=y,ACKnum=x+1) ESTAB ACK(ACKnum=y+1) ACK(ACKnum=y+1) L

Transport Layer 3-81 TCP: closing a connection

§ client, server each close their side of connection • send TCP segment with FIN bit = 1 § respond to received FIN with ACK • on receiving FIN, ACK can be combined with own FIN § simultaneous FIN exchanges can be handled

Transport Layer 3-82 TCP: closing a connection

client state server state ESTAB ESTAB clientSocket.close() FIN_WAIT_1 can no longer FINbit=1, seq=x send but can receive data CLOSE_WAIT ACKbit=1; ACKnum=x+1 can still FIN_WAIT_2 wait for server send data close

LAST_ACK FINbit=1, seq=y TIMED_WAIT can no longer send data ACKbit=1; ACKnum=y+1 timed wait for 2*max CLOSED segment lifetime

CLOSED

Transport Layer 3-83 Chapter 3 outline

Transport Layer 3-84 Principles of congestion control congestion: § informally: “too many sources sending too much data too fast for network to handle” § different from flow control! § manifestations: • lost packets (buffer overflow at routers) • long delays (queueing in router buffers) § a top-10 problem!

Transport Layer 3-85 Causes/costs of congestion: scenario 1

original data: lin throughput: l § two senders, two out receivers Host A

§ one router, infinite buffers unlimited shared § output link capacity: R output link buffers § no retransmission

Host B

R/2 out l delay

lin R/2 lin R/2 § maximum per-connection v large delays as arrival rate, lin, throughput: R/2 approaches capacity

Transport Layer 3-86 Causes/costs of congestion: scenario 2 § one router, finite buffers § sender retransmission of timed-out packet

• application-layer input = application-layer output: lin = lout ‘ • transport-layer input includes retransmissions : lin lin

lin : original data lout l'in: original data, plus retransmitted data

Host A

finite shared output Host B link buffers Transport Layer 3-87 Causes/costs of congestion: scenario 2

R/2 idealization: perfect knowledge out § sender sends only when l router buffers available lin R/2

lin : original data copy lout l'in: original data, plus retransmitted data

A free buffer space!

finite shared output Host B link buffers Transport Layer 3-88 Causes/costs of congestion: scenario 2 Idealization: known loss packets can be lost, dropped at router due to full buffers § sender only resends if packet known to be lost

lin : original data copy lout l'in: original data, plus retransmitted data

A no buffer space!

Host B

Transport Layer 3-89 Causes/costs of congestion: scenario 2

Idealization: known loss R/2 packets can be lost, when sending at R/2, dropped at router due some packets are out

to full buffers l retransmissions but asymptotic goodput § sender only resends if is still R/2 (why?) packet known to be lost R/2 lin

lin : original data lout l'in: original data, plus retransmitted data

A free buffer space!

Host B

Transport Layer 3-90 Causes/costs of congestion: scenario 2 Realistic: duplicates R/2 § packets can be lost, dropped at when sending at R/2, router due to full buffers some packets are out

§ sender times out prematurely, l retransmissions including duplicated sending two copies, both of that are delivered! which are delivered R/2 lin

lin timeoutcopy lout l'in

A free buffer space!

Host B

Transport Layer 3-91 Causes/costs of congestion: scenario 2 Realistic: duplicates R/2 § packets can be lost, dropped at when sending at R/2, router due to full buffers some packets are out

§ sender times out prematurely, l retransmissions including duplicated sending two copies, both of that are delivered! which are delivered R/2 lin

“costs” of congestion: § more work (retrans) for given “goodput” § unneeded retransmissions: link carries multiple copies of pkt • decreasing goodput

Transport Layer 3-92 Causes/costs of congestion: scenario 3

’ § four senders Q: what happens as lin and lin increase ? § multihop paths ’ A: as red lin increases, all arriving § timeout/retransmit blue pkts at upper queue are dropped, blue throughput g 0 Host A l lin : original data out Host B l'in: original data, plus retransmitted data finite shared output link buffers

Host D Host C

Transport Layer 3-93 Causes/costs of congestion: scenario 3

C/2 out l

’ lin C/2

another “cost” of congestion: § when packet dropped, any “upstream transmission capacity used for that packet was wasted!

Transport Layer 3-94 Chapter 3 outline

Transport Layer 3-95 TCP congestion control: additive increase multiplicative decrease § approach: sender increases transmission rate (window size), probing for usable bandwidth, until loss occurs • additive increase: increase cwnd by 1 MSS every RTT until loss detected • multiplicative decrease: cut cwnd in half after loss

additively increase window size … …. until loss occurs (then cut window in half)

AIMD saw tooth behavior: probing

for bandwidth sender TCP cwnd: congestion window size window congestion

time Transport Layer 3-96 TCP Congestion Control: details

sender sequence number space cwnd TCP sending rate: § roughly: send cwnd bytes, wait RTT for

last byte last byte ACKS, then send ACKed sent, not- sent yet ACKed more bytes (“in- flight”) cwnd rate ~ bytes/sec § sender limits transmission: RTT LastByteSent- < cwnd LastByteAcked § cwnd is dynamic, function of perceived network congestion

Transport Layer 3-97 TCP Slow Start

Host A Host B § when connection begins, increase rate exponentially until first

loss event: RTT • initially cwnd = 1 MSS • double cwnd every RTT • done by incrementing cwnd for every ACK received § summary: initial rate is slow but ramps up exponentially fast time

Transport Layer 3-98 TCP: detecting, reacting to loss

§ loss indicated by timeout: • cwnd set to 1 MSS; • window then grows exponentially (as in slow start) to threshold, then grows linearly § loss indicated by 3 duplicate ACKs: TCP RENO • dup ACKs indicate network capable of delivering some segments • cwnd is cut in half window then grows linearly § TCP Tahoe always sets cwnd to 1 (timeout or 3 duplicate acks)

Transport Layer 3-99 TCP: switching from slow start to CA Q: when should the exponential increase switch to linear? A: when cwnd gets to 1/2 of its value before timeout.

Implementation: § variable ssthresh § on loss event, ssthresh is set to 1/2 of cwnd just before loss event

* Check out the online interactive exercises for more examples: http://gaia.cs.umass.edu/kurose_ross/interactive/ Transport Layer 3-100 Summary: TCP Congestion Control

New New new ACK ACK! duplicate ACK ACK! cwnd = cwnd + MSS . (MSS/cwnd) dupACKcount++ new ACK dupACKcount = 0 cwnd = cwnd+MSS transmit new segment(s), as allowed dupACKcount = 0 L transmit new segment(s), as allowed cwnd = 1 MSS ssthresh = 64 KB cwnd > ssthresh dupACKcount = 0 slow L congestion

start timeout avoidance ssthresh = cwnd/2 cwnd = 1 MSS duplicate ACK timeout dupACKcount = 0 dupACKcount++ ssthresh = cwnd/2 retransmit missing segment cwnd = 1 MSS dupACKcount = 0 retransmit missing segment timeout New ACK! ssthresh = cwnd/2 cwnd = 1 New ACK dupACKcount = 0 cwnd = ssthresh dupACKcount == 3 dupACKcount == 3 retransmit missing segment dupACKcount = 0 ssthresh= cwnd/2 ssthresh= cwnd/2 cwnd = ssthresh + 3 cwnd = ssthresh + 3 retransmit missing segment retransmit missing segment fast recovery

duplicate ACK cwnd = cwnd + MSS transmit new segment(s), as allowed

Transport Layer 3-101 TCP throughput

§ avg. TCP thruput as function of window size, RTT? • ignore slow start, assume always data to send § W: window size (measured in bytes) where loss occurs • avg. window size (# in-flight bytes) is ¾ W • avg. thruput is 3/4W per RTT 3 W avg TCP thruput = bytes/sec 4 RTT

W/2

Transport Layer 3-102 TCP Futures: TCP over “long, fat pipes”

§ example: 1500 byte segments, 100ms RTT, want 10 Gbps throughput § requires W = 83,333 in-flight segments § throughput in terms of segment loss probability, L [Mathis 1997]: 1.22 . TCP throughput = MSS RTT L

➜ to achieve 10 Gbps throughput, need a loss rate of L = 2·10-10 – a very small loss rate! § new versions of TCP for high-speed

Transport Layer 3-103 TCP Fairness fairness goal: if K TCP sessions share same bottleneck link of bandwidth R, each should have average rate of R/K

TCP connection 1

bottleneck router capacity R TCP connection 2

Transport Layer 3-104 Why is TCP fair? two competing sessions: § additive increase gives slope of 1, as throughout increases § multiplicative decrease decreases throughput proportionally

R equal bandwidth share

loss: decrease window by factor of 2 congestion avoidance: additive increase loss: decrease window by factor of 2 congestion avoidance: additive increase

Connection 1 throughput R

Transport Layer 3-105 Fairness (more)

Fairness and UDP Fairness, parallel TCP § multimedia apps often connections do not use TCP § application can open • do not want rate multiple parallel throttled by congestion connections between control two hosts § instead use UDP: § • send audio/video at web browsers do this constant rate, tolerate § e.g., link of rate R with 9 packet loss existing connections: • new app asks for 1 TCP, gets rate R/10 • new app asks for 11 TCPs, gets R/2

Transport Layer 3-106 Explicit Congestion Notification (ECN) network-assisted congestion control: § two bits in IP header (ToS field) marked by network router to indicate congestion § congestion indication carried to receiving host § receiver (seeing congestion indication in IP datagram) ) sets ECE bit on receiver-to-sender ACK segment to notify sender of congestion

TCP ACK segment source destination application application ECE=1 transport transport network network link link physical physical

ECN=00 ECN=11

IP datagram Transport Layer 3-107 Chapter 3: summary

§ principles behind transport layer services: next: • multiplexing, § leaving the network demultiplexing “edge” (application, • reliable data transfer transport layers) • flow control § into the network • congestion control “core” § instantiation, § two network layer implementation in the chapters: Internet • data plane • UDP • control plane • TCP

Transport Layer 3-108