CSE 524: Lecture 11

1
CSE 524 Lecture 11

• Transport layer (Part 1)

2
Administrative

• Homework 4 due, solutions posted tomorrow
• Midterm Wednesday (11/5/03)
• Office hours Tuesday

3
Roadmap

• Moving up the stack
• Last class
• Finished network layer
• This class and beyond
• Transport and application layers

4
Transport layer outline

• Transport layer functions
• Specific Internet transport layers

5
Transport Layer

• provide logical communication between app
  processes running on different hosts
• transport protocols run in end systems
• transport vs network layer services
• network layer data transfer between end systems
• transport layer data transfer between processes
• relies on, enhances, network layer services

6
Transport Layer Functions

• Demux to upper layer
• Quality of service
• Security
• Delivery semantics
• Flow control
• Congestion control
• Reliable data transfer

7
TL Demux to upper layer (application)

• Recall segment - unit of data exchanged between
  transport layer entities
• aka TPDU transport protocol data unit

Demultiplexing delivering received segments to
correct app layer processes
receiver
P3
P4
application-layer data
segment header
P1
P2
segment
H
t
M
segment
8
TL Quality of service

• Provide predictability and guarantees in
  transport layer
• Operating system issues
• Protocol handler scheduling
• Buffer resource allocation
• Process/application scheduling
• Support for signaling (setup, management,
  teardown)
• L4 (transport) switches, L5 (application)
  switches, and NAT devices
• Issues in supporting QoS at the end systems and
  end clusters

9
TL Security

• Provide at the transport level
• Secrecy
• No eavesdropping
• Integrity
• No man-in-the-middle attacks
• Authenticity
• Ensure identity of source
• What is the difference between transport layer
  security and network layer security?
• Does the end-to-end principle apply?

10
TL Delivery semantics

• Reliable vs. unreliable
• Unicast vs. multicast
• Ordered vs. unordered
• Any others?

11
TL Flow control

• Do not allow sender to overrun receivers buffer
  resources
• Similar to data-link layer flow control, but done
  on an end-to-end basis

12
TL Congestion control

• Congestion
• informally too many sources sending too much
  data too fast for network to handle
• sources compete for resources inside network
• different from flow control!
• manifestations
• lost packets (buffer overflow at routers)
• long delays (queueing in router buffers)

13
TL Congestion

• Why is it a problem?
• Sources are unaware of current state of resource
• Sources are unaware of each other
• In many situations will result in lt 1.5 Mbps of
  goodput (more later)

14
TL Causes/costs of congestion scenario 1

• two senders, two receivers
• one router, infinite buffers
• no retransmission

• large delays when congested
• maximum achievable throughput

15
TL Causes/costs of congestion scenario 2

• one router, finite buffers
• sender retransmission of lost packet

Host A
lout
lin original data
l'in original data, plus retransmitted data
Host B
finite shared output link buffers
16
TL Causes/costs of congestion scenario 2

• no loss (goodput)
• perfect retransmission only when loss
• retransmission of delayed (not lost) packet makes
  larger (than perfect case) for same

• costs of congestion
• more work (retrans) for given goodput
• unneeded retransmissions link carries multiple
  copies of pkt

17
TL Causes/costs of congestion scenario 3

• four senders
• multihop paths
• timeout/retransmit

Q what happens as and increase ?
lout
lin original data
l'in original data, plus retransmitted data
finite shared output link buffers
18
TL Causes/costs of congestion scenario 3
lout

• Another cost of congestion
• when packet dropped, any upstream transmission
  capacity used for that packet was wasted!

19
TL Congestion Collapse

• Increase in network load results in decrease of
  useful work done
• Spurious retransmissions of packets still in
  flight
• Classical congestion collapse
• Solution better timers and congestion control
• Undelivered packets
• Packets consume resources and are dropped
  elsewhere in network
• Solution congestion control for ALL traffic
• Fragments
• Mismatch of transmission and retransmission units
• Solutions
• Make network drop all fragments of a packet
  (early packet discard in ATM)
• Do path MTU discovery
• Control traffic
• Large percentage of traffic is for control
• Headers, routing messages, DNS, etc.
• Stale or unwanted packets
• Packets that are delayed on long queues
• Push data that is never used

20
TL Preventing Congestion Collapse

• End-host vs. network controlled
• Trust hosts to do the right thing
• Hosts adjust rate based on detected congestion
  (TCP)
• Dont trust hosts and enforce within network
• Network adjusts rates at congestion points
• Scheduling
• Queue management
• Hard to prevent global collapse conditions
  locally
• Implicit vs. explicit rate control
• Infer congestion from packet loss or delay
• Increase rate in absence of loss, decrease on
  loss (TCP Tahoe/Reno)
• Increase rate based on delay behavior (TCP Vegas,
  Packet pair)
• Explicit signaling from network
• Congestion notification (DECbit, ECN)
• Rate signaling (ATM ABR)

21
TL Goals for congestion control mechanisms

• Use network resources efficiently
• 100 link utilization, 0 packet loss, Low delay
• Maximize network power (throughputa/delay)
• Efficiency/goodput Xknee Sxi(t)
• Preserve fair network resource allocation
• Fairness (Sxi)2/n(Sxi2)
• Max-min fair sharing
• Small flows get all of the bandwidth they require
• Large flows evenly share leftover
• Example
• 100Mbs link
• S1 and S2 are 1Mbs streams, S3 and S4 are
  infinite greedy streams
• S1 and S2 each get 1Mbs, S3 and S4 each get 49Mbs
• Convergence and stability
• Distributed operation
• Simple router and end-host behavior

22
TL Congestion Control vs. Avoidance

• Avoidance keeps the system performing at the
  knee/cliff
• Control kicks in once the system has reached a
  congested state

23
TL Basic Control Model

• Of all ways to do congestion, the Internet
  chooses.
• Mainly end-host, window-based congestion control
• Only place to really prevent collapse is at
  end-host
• Reduce sender window when congestion is perceived
• Increase sender window otherwise (probe for
  bandwidth)
• Congestion signaling and detection
• Mark/drop packets when queues fill, overflow
• Will cover this separately in later lecture
• Given this, how does one design a windowing
  algorithm which best meets the goals of
  congestion control?

24
TL Linear Control

• Many different possibilities for reaction to
  congestion and probing
• Examine simple linear controls
• Window(t 1) a b Window(t)
• Different ai/bi for increase and ad/bd for
  decrease
• Supports various reaction to signals
• Increase/decrease additively
• Increase/decrease multiplicatively
• Which of the four combinations is optimal?

25
TL Phase plots

• Simple way to visualize behavior of competing
  connections over time

Fairness Line
User 2s Allocation x2
Efficiency Line
User 1s Allocation x1
26
TL Phase plots

• What are desirable properties?
• What if flows are not equal?

Fairness Line
Overload
User 2s Allocation x2
Optimal point
Underutilization
Efficiency Line
User 1s Allocation x1
27
TL Additive Increase/Decrease

• Both X1 and X2 increase/decrease by the same
  amount over time
• Additive increase improves fairness and additive
  decrease reduces fairness

Fairness Line
T1
User 2s Allocation x2
T0
Efficiency Line
User 1s Allocation x1
28
TL Muliplicative Increase/Decrease

• Both X1 and X2 increase by the same factor over
  time
• Extension from origin constant fairness

Fairness Line
T1
User 2s Allocation x2
T0
Efficiency Line
User 1s Allocation x1
29
TL Convergence to Efficiency Fairness

• From any point, want to converge quickly to
  intersection of fairness and efficiency lines

Fairness Line
xH
User 2s Allocation x2
Efficiency Line
User 1s Allocation x1
30
TL What is the Right Choice?

• Constraints limit us to AIMD
• AIMD moves towards optimal point

Fairness Line
x1
x0
User 2s Allocation x2
x2
Efficiency Line
User 1s Allocation x1
31
TL Reliable data transfer

• Error detection, correction
• Retransmission
• Duplicate detection
• Connection integrity

32
TL Principles of Reliable data transfer

• important in app., transport, link layers

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

33
TL Reliable data transfer getting started
send side
receive side
34
TL Reliable data transfer getting started

• Well
• 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 uniquely
determined by next event
35
TL 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 read data from underlying channel

36
TL Rdt2.0 channel with bit errors

• underlying channel may flip bits in packet
• 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
• receiver feedback control msgs (ACK,NAK)
  rcvr-gtsender

37
TL rdt2.0 FSM specification
sender FSM
receiver FSM
38
TL rdt2.0 operation with no errors
rdt_send(data)
snkpkt make_pkt(data, checksum) udt_send(sndpkt)
rdt_rcv(rcvpkt) isNAK(rcvpkt)
Wait for call from above
udt_send(sndpkt)
rdt_rcv(rcvpkt) isACK(rcvpkt)
Wait for call from below
L
rdt_rcv(rcvpkt) notcorrupt(rcvpkt)
extract(rcvpkt,data) deliver_data(data) udt_send(A
CK)
39
TL rdt2.0 error scenario
rdt_send(data)
snkpkt make_pkt(data, checksum) udt_send(sndpkt)
rdt_rcv(rcvpkt) isNAK(rcvpkt)
Wait for call from above
udt_send(sndpkt)
rdt_rcv(rcvpkt) isACK(rcvpkt)
Wait for call from below
L
rdt_rcv(rcvpkt) notcorrupt(rcvpkt)
extract(rcvpkt,data) deliver_data(data) udt_send(A
CK)
40
TL rdt2.0 has a fatal flaw!

• What happens if ACK/NAK corrupted?
• sender doesnt know what happened at receiver!
• cant just retransmit possible duplicate
• What to do?
• sender ACKs/NAKs receivers ACK/NAK? What if
  sender ACK/NAK lost?
• retransmit, but this might cause retransmission
  of correctly received pkt!

• Handling duplicates
• sender adds sequence number to each pkt
• sender retransmits current pkt if ACK/NAK garbled
• receiver discards (doesnt deliver up) duplicate
  pkt

Sender sends one packet, then waits for receiver
response
41
TL rdt2.1 sender, handles garbled ACK/NAKs
42
TL rdt2.1 receiver, handles garbled ACK/NAKs
43
TL rdt2.1 discussion

• Sender
• seq added to pkt
• two seq. s (0,1) will suffice. Why?
• must check if received ACK/NAK corrupted
• twice as many states
• state must remember whether current pkt has 0
  or 1 seq.

• Receiver
• must check if received packet is duplicate
• state indicates whether 0 or 1 is expected pkt
  seq
• note receiver can not know if its last ACK/NAK
  received OK at sender

44
TL rdt2.2 a NAK-free protocol

• same functionality as rdt2.1, using NAKs 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

45
rdt2.2 sender, receiver fragments
rdt_send(data)
sndpkt make_pkt(0, data, checksum) udt_send(sndp
kt)
rdt_rcv(rcvpkt) ( corrupt(rcvpkt)
isACK(rcvpkt,1) )
udt_send(sndpkt)
sender FSM fragment
rdt_rcv(rcvpkt) notcorrupt(rcvpkt)
isACK(rcvpkt,0)
rdt_rcv(rcvpkt) (corrupt(rcvpkt)
has_seq1(rcvpkt))
L
receiver FSM fragment
udt_send(sndpkt)
rdt_rcv(rcvpkt) notcorrupt(rcvpkt)
has_seq1(rcvpkt)
extract(rcvpkt,data) deliver_data(data) sndpkt
make_pkt(ACK1, chksum) udt_send(sndpkt)
46
TL rdt3.0 channels with errors and loss

• New assumption underlying channel can also lose
  packets (data or ACKs)
• checksum, seq. , ACKs, retransmissions will be
  of help, but not enough
• Q how to deal with loss?
• sender waits until certain data or ACK lost, then
  retransmits
• yuck drawbacks?

• Approach sender waits reasonable amount of
  time for ACK
• retransmits if no ACK received in this time
• if pkt (or ACK) just delayed (not lost)
• retransmission will be duplicate, but use of
  seq. s already handles this
• receiver must specify seq of pkt being ACKed
• requires countdown timer

47
TL rdt3.0 sender
rdt_send(data)
rdt_rcv(rcvpkt) ( corrupt(rcvpkt)
isACK(rcvpkt,1) )
sndpkt make_pkt(0, data, checksum) udt_send(sndp
kt) start_timer
L
rdt_rcv(rcvpkt)
L
timeout
udt_send(sndpkt) start_timer
rdt_rcv(rcvpkt) notcorrupt(rcvpkt)
isACK(rcvpkt,1)
rdt_rcv(rcvpkt) notcorrupt(rcvpkt)
isACK(rcvpkt,0)
stop_timer
stop_timer
timeout
udt_send(sndpkt) start_timer
rdt_rcv(rcvpkt)
L
rdt_send(data)
rdt_rcv(rcvpkt) ( corrupt(rcvpkt)
isACK(rcvpkt,0) )
sndpkt make_pkt(1, data, checksum) udt_send(sndp
kt) start_timer
L
48
TL rdt3.0 in action
49
TL rdt3.0 in action
50
TL Performance of rdt3.0

• rdt3.0 works, but performance stinks
• example 1 Gbps link, 15 ms e-e prop. delay, 1KB
  packet

L (packet length in bits)
8kb/pkt
T


8 microsec
transmit
R (transmission rate, bps)
109 b/sec

• U sender utilization fraction of time sender
  busy sending
• 1KB pkt every 30 msec -gt 33kB/sec thruput over 1
  Gbps link
• network protocol limits use of physical resources!

51
TL 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
52
TL 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

53
TL 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
Increase utilization by a factor of 3!
54
TL Go-Back-N

• Sender
• k-bit seq in pkt header
• window of up to N, consecutive unacked pkts
  allowed

• ACK(n) ACKs all pkts up to, including seq n -
  cumulative ACK
• may deceive duplicate ACKs (see receiver)
• timer for each in-flight pkt
• timeout(n) retransmit pkt n and all higher seq
  pkts in window

55
TL GBN sender extended FSM
56
TL GBN receiver extended FSM

• receiver simple
• 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 (dont buffer) -gt no receiver buffering!
• ACK pkt with highest in-order seq

57
TL GBN in action
58
TL 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
• again limits seq s of sent, unACKed pkts

59
TL Selective repeat sender, receiver windows
60
TL Selective repeat

• data from above
• if next available seq in window, send pkt
• timeout(n)
• resend pkt n, restart timer
• ACK(n) in sendbase,sendbaseN
• mark pkt n as received
• if n smallest unACKed pkt, advance window base to
  next unACKed seq

• pkt n in rcvbase, rcvbaseN-1
• send ACK(n)
• out-of-order buffer
• in-order deliver (also deliver buffered,
  in-order pkts), advance window to next
  not-yet-received pkt
• pkt n in rcvbase-N,rcvbase-1
• ACK(n)
• otherwise
• ignore

61
TL Selective repeat in action
62
TL Selective repeat dilemma

• Example
• seq s 0, 1, 2, 3
• window size3
• receiver sees no difference in two scenarios!
• incorrectly passes duplicate data as new in (a)
• Q what relationship between seq size and
  window size?