Transcript 3rd Edition: Chapter 3 - Georgia Institute of Technology

Chapter 3
Transport Layer
Modified by John Copeland
Georgia Tech
for use in ECE3600
A note on the use of these ppt slides:
We’re making these slides freely available to all (faculty, students, readers).
They’re in PowerPoint form so you can add, modify, and delete slides
(including this one) and slide content to suit your needs. They obviously
represent a lot of work on our part. In return for use, we only ask the
following:
 If you use these slides (e.g., in a class) in substantially unaltered form,
that you mention their source (after all, we’d like people to use our book!)
 If you post any slides in substantially unaltered form on a www site, that
you note that they are adapted from (or perhaps identical to) our slides, and
note our copyright of this material.
Computer Networking:
A Top Down Approach
Featuring the Internet,
5th edition.
Jim Kurose, Keith Ross
Addison-Wesley, July
2009.
Thanks and enjoy! JFK/KWR
All material copyright 1996-2009
J.F Kurose and K.W. Ross, All Rights Reserved
Transport Layer
10-4-2014
3-1
Chapter 3: Transport Layer
Our goals:
 understand principles
behind transport
layer services:




multiplexing/demultiplexing
reliable data transfer
flow control
congestion control
 learn about transport
layer protocols in the
Internet:



UDP: connectionless
transport
TCP: connection-oriented
transport
TCP congestion control
Transport Layer
3-2
Chapter 3 outline
 3.1 Transport-layer
services
 3.2 Multiplexing and
demultiplexing
 3.3 Connectionless
transport: UDP
 3.4 Principles of
reliable data transfer
 3.5 Connection-oriented
transport: TCP




segment structure
reliable data transfer
flow control
connection management
 3.6 Principles of
congestion control
 3.7 TCP congestion
control
Transport Layer
3-3
Transport services and protocols
 provide
logical communication
between app processes
running on different hosts
 transport protocols run in
end systems
 send side: breaks app
messages into segments,
passes to network layer
 receive side: reassembles
segments into messages,
passes to app layer
 more than one transport
protocol available to apps
 Internet: TCP and UDP
application
transport
network
data link
physical
network
data link
physical
network
data link
physical
network
data link
physical
network
data link
physical
network
data link
physical
application
transport
network
data link
physical
Transport Layer
3-4
Transport vs. network layer

network layer: logical
Household analogy:

transport layer: logical
 processes = kids
communication
between hosts
communication
between processes*

relies on, enhances,
network layer services
* running on different hosts
12 kids sending letters to
12 kids
 app messages = letters
in envelopes
 hosts = houses
 transport protocol =
Ann and Bill
 network-layer protocol
= postal service
Transport Layer
3-5
Internet transport-layer protocols
 reliable, in-order
delivery (TCP)



congestion control
flow control
connection setup
 unreliable, unordered
delivery: UDP

no-frills extension of
“best-effort” IP
 services not available:
 delay guarantees
 bandwidth guarantees
application
transport
network
data link
physical
network
data link
physical
network
data link
physical
network
data link
physical
network
data link
physical
network
data link
physical
application
transport
network
data link
physical
Transport Layer
3-6
Chapter 3 outline
 3.1 Transport-layer
services
 3.2 Multiplexing and
demultiplexing
 3.3 Connectionless
transport: UDP
 3.4 Principles of
reliable data transfer
 3.5 Connection-oriented
transport: TCP




segment structure
reliable data transfer
flow control
connection management
 3.6 Principles of
congestion control
 3.7 TCP congestion
control
Transport Layer
3-7
Multiplexing/demultiplexing
Multiplexing at send host:
gathering data from multiple
sockets, enveloping data with
header (later used for
demultiplexing)
Demultiplexing at rcv host:
delivering received segments
to correct socket
= socket
application
transport
network
link
= process
P3
P1
P1
application
transport
network
P2
P4
application
transport
network
link
link
physical
host 1
physical
host 2
physical
host 3
Transport Layer
3-8
How demultiplexing works
 host receives IP datagrams
each datagram has source
IP address, destination IP
address
 each datagram carries 1
transport-layer segment
 each segment has source,
destination port number
 host uses IP addresses & port
numbers to direct segment to
appropriate socket

32 bits
source port #
dest port #
other header fields
application
data
(message)
TCP and UDP segment format
Transport Layer
3-9
Connectionless demultiplexing
 Create sockets with port
numbers:
DatagramSocket mySocket1 = new
DatagramSocket(12534);
DatagramSocket mySocket2 = new
DatagramSocket(12535);
 UDP socket identified by
two-tuple
 (plus "UDP", it is a 3-tuple):
(For incoming segment:
destination IP address,
destination port number)
 When host receives UDP
segment:
 checks destination port
number in segment
 directs UDP segment to
socket with that port number
 IP datagrams with different
source IP addresses and/or
source port numbers directed to
same UDP socket
This means a single process thread
can handle all UDP packets coming to
a host that have the same local
(destination) UDP port number.
This is not the case for TCP (an
addition thread is needed for each
remote IP address and TCP port pair).
Transport Layer 3-10
Connectionless (UDP) demux (cont)
DatagramSocket serverSocket = new DatagramSocket(6428);
P2
SP: 6428
SP: 6428
DP: 9157
DP: 5775
SP: 9157
client
IP: A
P1
P1
P3
DP: 6428
SP provides “return address”
SP: 5775
server
IP: C
DP: 6428
Client
IP:B
Same incoming UDP destination port,
delivered to same process, the one
that opened that socket.
Transport Layer
3-11
Connection-oriented demux
 TCP socket identified
by a 4-tuple
 (+ "TCP" -> a 5-tuple):




source IP address
source port number
dest IP address
dest port number
 recv host TCP layer uses
all four values to direct
segment to appropriate
socket
 Server host may support
many simultaneous TCP
sockets:

each TCP socket identified
by its own 4-tuple (5-tuple)
 Web servers have
different sockets for
each connecting client

non-persistent HTTP will
have different socket
(different client port
number) for each request
Transport Layer 3-12
Connection-oriented (TCP)
demux (cont)
Different TCP source port
from P5 and P6 on client C
-> different process (thread)
(P2, P3 on server B)
P1
P4
P5
P2
P6
P1P3
SP: 5775
DP: 80
S-IP: C
D-IP:B
server
IP: A
SP:5776
DP: 80
S-IP: C
D-IP:A
SP: 5777
client
IP: C
DP: 80
S-IP: C
D-IP:B
server
IP:B
Transport Layer 3-13
Connection-oriented (TCP) demux:
Threaded Web Server
P1
Listening
Socket
SP: *
DP: 80
Different TCP source port
-> different sub-process (thread)
*
DP: 80
D-IP:C
S-IP: B
D-IP:C
* = "any"
SP: 9157
DP: 80
S-IP: A
D-IP:C
P1P3
SP: 5775
S-IP: *
client
IP: A
P2
P4
SP: 9157
server
IP: C
DP: 80
S-IP: B
D-IP:C
Client
IP:B
thread, spawned or "forked" from main process,
manages each connected socket.
Transport Layer 3-14
Active Socket states: ESTABLISHED or LISTEN (netstat)
mc:/Users/copeland root# netstat -f inet -al
Active Internet connections
Proto Recv-Q Send-Q Local Address (IP.port) Foreign Address
tcp4
0
0 mc.55185
www.suntrust.com.https
tcp4
0
48 mc.ssh
home.49901
tcp4
0
0 *.9503
*.*
tcp4
0
0 localhost.ipp
*.*
tcp4
0
0 *.afpovertcp
*.*
tcp4
0
0 *.timbuktu
*.*
tcp4
0
0 *.ssh
*.*
udp4
0
0 *.mdns
*.*
udp4
0
0 *.52521
*.*
udp4
0
0 *.9912
*.*
udp4
0
0 *.ipp
*.*
udp4
0
0 localhost.49156
localhost.1022
udp4
0
0 localhost.49155
localhost.1022
udp4
0
0 localhost.1022
*.*
udp4
0
0 localhost.1023
*.*
udp4
0
0 *.timbuktu-srv3
*.*
udp4
0
0 *.timbuktu
*.*
udp4
0
0 *.*
*.*
udp4
0
0 mc.ntp
*.*
udp4
0
0 localhost.ntp
*.*
udp4
0
0 *.ntp
*.*
udp4
0
0 *.tftp
*.*
(state)
ESTABLISHED
ESTABLISHED
LISTEN
LISTEN
LISTEN
LISTEN
LISTEN
Well Known Ports (from /etc/services): ssh = 22, mdns = 5353, timbuktu = 407,
timbuktu-srv3 = 1419, ntp = 123, tftp = 69, ipp = 631
Host names (from /etc/hosts): mc, home (from DNS) www.suntrust.com
Transport Layer 3-15
Chapter 3 outline
 3.1 Transport-layer
services
 3.2 Multiplexing and
demultiplexing
 3.3 Connectionless
transport: UDP
 3.4 Principles of
reliable data transfer
 3.5 Connection-oriented
transport: TCP




segment structure
reliable data transfer
flow control
connection management
 3.6 Principles of
congestion control
 3.7 TCP congestion
control
Transport Layer 3-16
UDP: User Datagram Protocol [RFC 768]
 “no frills,” “bare bones”
Internet transport
protocol
 “best effort” service, UDP
segments may be:
 lost
 delivered out of order
to app

connectionless:


no handshaking between
UDP sender, receiver
each UDP segment
handled independently
of others
Why is there a UDP?
 no connection
establishment (which can
add delay)
 simple: no connection state
at sender, receiver
 small segment header
 no congestion control: UDP
can blast away as fast as
desired
Transport Layer 3-17
UDP: more
 often used for streaming
multimedia apps
 loss tolerant
 rate sensitive
Length, in
bytes of UDP
segment,
including
 other UDP uses
header
 DNS (name lookup)
SNMP (network management)
 NTP (network time protocol)
 to get reliable transfer over
UDP: add reliability at application
layer
 application-specific error
recovery!

32 bits
source port #
dest port #
length
checksum
Application
data
(message)
UDP segment format
Transport Layer 3-18
UDP checksum
Goal: detect “errors” (e.g., flipped bits) in
transmitted segment
Sender:
Receiver:
 treat segment contents
 compute checksum of
as sequence of 16-bit
integers
 checksum: addition (1’s
complement sum) of
segment contents
 sender puts checksum
value into UDP checksum
field
received segment
 check if computed checksum
equals checksum field value:
 NO - error detected
 YES - no error detected.
But maybe errors
nonetheless? More later
….
Transport Layer 3-19
Internet Checksum Example
 Note

When adding numbers, a carryout from the
most significant bit needs to be added to the
result
 Example: add two 16-bit integers (1's compliment)
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
Transport Layer 3-20
Chapter 3 outline
 3.1 Transport-layer
services
 3.2 Multiplexing and
demultiplexing
 3.3 Connectionless
transport: UDP
 3.4 Principles of
reliable data transfer
 3.5 Connection-oriented
transport: TCP




segment structure
reliable data transfer
flow control
connection management
 3.6 Principles of
congestion control
 3.7 TCP congestion
control
Transport Layer 3-21
Principles of Reliable data transfer
 important in app., 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 app., 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
Principles of Reliable data transfer
 important in app., 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 3-24
Reliable data transfer: getting started
rdt_send(): called from above,
(e.g., by app.). Passed data to
deliver to receiver upper layer
send
side
udt_send(): called by rdt,
to transfer packet over
unreliable channel to receiver
deliver_data(): called by
rdt to deliver data to upper
receive
side
rdt_rcv(): called when packet
arrives on rcv-side of channel
Transport Layer 3-25
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
state: when in this
“state” next state
uniquely determined
by next event
state
1
event causing state transition
actions taken on state transition
event
actions
state
2
Transport Layer 3-26
Rdt1.0: reliable transfer over a reliable channel
 underlying channel perfectly reliable
 no bit errors
Event causes change of state
Action that occurs
 no loss of packets
 separate FSMs for sender, receiver:
 sender sends data into underlying channel
 receiver read data from underlying channel
Wait for
call from
above
rdt_send(data)
packet = make_pkt(data)
udt_send(packet)
sender
Wait for
call from
below
rdt_rcv(packet)
extract (packet,data)
deliver_data(data)
receiver
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

negative acknowledgements (NAKs): receiver explicitly

that pkt received OK
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->sender
Transport Layer 3-28
rdt2.0: FSM specification
rdt_send(data)
snkpkt = make_pkt(data, checksum)
udt_send(sndpkt)
rdt_rcv(rcvpkt) &&
isNAK(rcvpkt)
Wait for
Wait for
call from
ACK or
udt_send(sndpkt)
above
NAK
rdt_rcv(rcvpkt) && isACK(rcvpkt)
L
sender
receiver
rdt_rcv(rcvpkt) &&
corrupt(rcvpkt)
udt_send(NAK)
Wait for
call from
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
call from
ACK or
udt_send(sndpkt)
above
NAK
rdt_rcv(rcvpkt) && isACK(rcvpkt)
L
rdt_rcv(rcvpkt) &&
corrupt(rcvpkt)
udt_send(NAK)
Wait for
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_send(data)
snkpkt = make_pkt(data, checksum)
udt_send(sndpkt)
rdt_rcv(rcvpkt) &&
isNAK(rcvpkt)
Wait for
Wait for
call from
ACK or
udt_send(sndpkt)
above
NAK
rdt_rcv(rcvpkt) && isACK(rcvpkt)
L
rdt_rcv(rcvpkt) &&
corrupt(rcvpkt)
udt_send(NAK)
Wait for
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
ACK/NAK corrupted?
 sender doesn’t know what
happened at receiver!
 can’t just retransmit:
possible duplicate
Handling duplicates:
 sender retransmits current
pkt if ACK/NAK garbled
 sender adds sequence
number to each pkt
 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
isNAK(rcvpkt) )
ACK or
call 0 from
udt_send(sndpkt)
NAK 0
above
rdt_rcv(rcvpkt)
&& notcorrupt(rcvpkt)
&& isACK(rcvpkt)
rdt_rcv(rcvpkt)
&& notcorrupt(rcvpkt)
&& isACK(rcvpkt)
L
rdt_rcv(rcvpkt) &&
( corrupt(rcvpkt) ||
isNAK(rcvpkt) )
udt_send(sndpkt)
L
Wait for
ACK or
NAK 1
Wait for
call 1 from
above
rdt_send(data)
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)
rdt_rcv(rcvpkt) && (corrupt(rcvpkt)
extract(rcvpkt,data)
deliver_data(data)
sndpkt = make_pkt(ACK, chksum)
udt_send(sndpkt)
rdt_rcv(rcvpkt) && (corrupt(rcvpkt)
sndpkt = make_pkt(NAK, chksum)
udt_send(sndpkt)
rdt_rcv(rcvpkt) &&
not corrupt(rcvpkt) &&
has_seq1(rcvpkt)
sndpkt = make_pkt(ACK, chksum)
udt_send(sndpkt)
sndpkt = make_pkt(NAK, chksum)
udt_send(sndpkt)
Wait for
0 from
below
Wait for
1 from
below
rdt_rcv(rcvpkt) && notcorrupt(rcvpkt)
&& has_seq1(rcvpkt)
rdt_rcv(rcvpkt) &&
not corrupt(rcvpkt) &&
has_seq0(rcvpkt)
sndpkt = make_pkt(ACK, chksum)
udt_send(sndpkt)
extract(rcvpkt,data)
deliver_data(data)
sndpkt = make_pkt(ACK, chksum)
udt_send(sndpkt)
Transport Layer 3-34
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
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
[say “0”]
 duplicate ACK at sender [two ACKs for “0”] results
in same action as NAK: retransmit current pkt [“1”]
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) )
ACK
call 0 from
0
udt_send(sndpkt)
above
sender FSM
fragment
rdt_rcv(rcvpkt) &&
(corrupt(rcvpkt) ||
has_seq1(rcvpkt))
Wait for
0 from
below
sndpkt =
make_pkt(ACK0, chksum)
udt_send(sndpkt)
“Duplicate ACK”
rdt_rcv(rcvpkt)
&& notcorrupt(rcvpkt)
&& isACK(rcvpkt,0)
receiver FSM
fragment
L
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:
underlying channel can
also lose packets (data
or ACKs)

checksum, seq. #, ACKs,
retransmissions will be
of help, but not enough
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
Transport Layer 3-38
rdt3.0 sender
rdt_send(data)
sndpkt = make_pkt(0, data, checksum)
udt_send(sndpkt)
start_timer
rdt_rcv(rcvpkt)
L
Wait for
call 0 from
above
rdt_rcv(rcvpkt)
&& notcorrupt(rcvpkt)
&& isACK(rcvpkt,1)
timeout
udt_send(sndpkt)
start_timer
rdt_rcv(rcvpkt) &&
( corrupt(rcvpkt) ||
isACK(rcvpkt,0) )
L
Wait
for
ACK0
timeout
udt_send(sndpkt)
start_timer
rdt_rcv(rcvpkt)
&& notcorrupt(rcvpkt)
&& isACK(rcvpkt,0)
stop_timer
stop_timer
L
rdt_rcv(rcvpkt) &&
( corrupt(rcvpkt) ||
isACK(rcvpkt,1) )
Wait
for
ACK1
Wait for
call 1 from
above
rdt_send(data)
rdt_rcv(rcvpkt)
L
sndpkt = make_pkt(1, data, checksum)
udt_send(sndpkt)
start_timer
Transport Layer 3-39
rdt3.0 in action
Transport Layer 3-40
rdt3.0 in action
Transport Layer 3-41
Review rdt3.0
ACK Packet
Lets Sender know that data has been received.
Sequence Number
In Header - Lets Receiver know where to place data in
buffer
(to avoid duplication and gaps)
Acknowledgement Number
In ACK - Lets Sender know which data has been
received.
(so it knows what to resend)
Retransmit Timer
Keeps connection from freezing because of a lost packet.
Transport Layer 3-42
Performance of rdt3.0
 rdt3.0 works, but performance stinks
 example: 1 Gbps link, 15 ms e-e prop. delay, 1KB packet:
Ttransmit =



L (packet length in bits)
8kb/pkt
=
= 0.008 millisec
R (transmission rate, bps)
10**9 b/sec
U sender: utilization – fraction of time sender busy sending
1KB pkt every 30 msec -> 33kB/sec thruput over 1 Gbps link
network protocol limits use of physical resources!
Transport Layer 3-43
rdt3.0: stop-and-wait operation
sender
receiver
first packet bit transmitted, t = 0
last packet bit transmitted, t = L / R
RTT
first packet bit arrives
last packet bit arrives, send ACK
ACK arrives, send next
packet, t = RTT + L / R
U is the “channel utilization factor”, or “efficiency”
Transport Layer 3-44
Pipelined protocols
Pipelining: sender allows multiple, “in-flight”, yetto-be-acknowledged pkts


range of sequence numbers must be increased
buffering at sender and/or receiver
 Two generic forms of pipelined protocols:
selective repeat
go-Back-N,
Transport Layer 3-45
Pipelining: increased utilization
first packet bit transmitted,
t=0
last bit transmitted, t = L / R
sender
RTT
receiver
first packet bit arrives
last packet bit arrives, send ACK 1
last bit of 2nd packet arrives, send ACK 2
last bit of 3rd packet arrives, send ACK 3
ACK arrives, send next
packet, t = RTT + L / R
Increase utilization
by a factor of 3 !
If 3L/R < RTT and Window = 3L, then Throughput Rt = R*U = (Window/RTT)
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 each in-flight pkt
 timeout(n): retransmit pkt n and all higher seq # pkts in window

Transport Layer 3-47
Go Back N: Receiver
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 packet with the highest in-order seq #
Transport Layer 3-48
GBN in
action
Transport Layer 3-49
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
Transport Layer 3-50
Selective repeat: sender, receiver windows
send_base
moves forward
to first notack’ed packet
rcv_base
moves
forward to
first nottransferred
packet
(TCP buffers all, but ACK’s first missing byte, = rcv-base)
Transport Layer 3-51
Selective repeat
sender
data from above :
receiver
pkt n in [rcvbase, rcvbase+N-1]
 if next available seq # in
 send ACK(n)
timeout(n):
 in-order: deliver (also
window, send pkt
 resend pkt n, restart timer
ACK(n) in [sendbase,sendbase+N]:
 mark pkt n as received
 if n smallest unACKed pkt,
advance window base to
next unACKed seq #
 out-of-order: buffer
deliver buffered, in-order
pkts), advance window to
next not-yet-received pkt
pkt n in
[rcvbase-N,rcvbase-1]
 ACK(n)
(already received & delivered)
otherwise:
 Ignore
Transport Layer 3-52
Selective repeat in action
TCP
ack1
ack2
ACK(2nd)
ack2 (3rd)
ack2 (4th)
On 4th ack2
The sender
resends only
Pkt2,
Then continues
Sender must keep a list
of pkt’s ACKed
Transport Layer 3-53
Selective repeat:
dilemma
Example:
 seq #’s: 0, 1, 2, 3
 window size=3
 receiver sees no
difference in two
scenarios!
 incorrectly passes
duplicate data as new
in (a)
Q: what relationship
between seq # size
and window size?
Transport Layer 3-54
Reliable Data Transport
Problem
Solution
Packet may arrive with errors.
Add checksum, CRC, or hash.
Packet may not arrive.
Receiver sends “ACK” back. If
ACK not received, packet re-sent.
Sender may wait forever for ACK.
Timeout timer added to sender.
ACK may not arrive, dup. sent.
Add sequence no.s to detect dups.
Packets may arrive out-of-order.
Buffer packets to rearrange order.
Inefficient to send one pkt per RT
Have a “window” to send before ACK
(pipelining).
Missing packet early in window.
“Go-Back-N” to last in-order packet.
“Go-Back-N” inefficient.
“Selective Repeat” to fill in gaps only.
---- Also in TCP ---
----
Packets may be different sizes.
Sequence number for each byte.
Slow down when network
“Slow-Start”, or "Multiplicative
congested (as detected by RTO or
Decrease" to reduce transmit window.
triple duplicate ACKs.
Know when receiver buffer will
be full.
Receiver includes “space left” in every
ACK.
Transport Layer 3-55
Chapter 3 outline
 3.1 Transport-layer
services
 3.2 Multiplexing and
demultiplexing
 3.3 Connectionless
transport: UDP
 3.4 Principles of
reliable data transfer
 3.5 Connection-oriented
transport: TCP




segment structure
reliable data transfer
flow control
connection management
 3.6 Principles of
congestion control
 3.7 TCP congestion
control
Transport Layer 3-56
TCP: Overview
 point-to-point:
 one sender, one receiver
 reliable, in-order
steam:

byte
no “message boundaries”
 pipelined:
 TCP congestion and flow
control set window size

socket
door
send & receive buffers
application
writes data
application
reads data
TCP
send buffer
TCP
receive buffer
RFCs: 793, 1122, 1323, 2018, 2581
 full duplex data:
 bi-directional data flow
in same connection
 MSS: maximum segment
size
 connection-oriented:
 handshaking (exchange
of control msgs) init’s
sender, receiver state
before data exchange
 flow controlled:
 sender will not
socket
door
overwhelm receiver
segment
Transport Layer 3-57
TCP segment structure
32 bits
URG: urgent data
(generally not used)
ACK: ACK #
valid
PSH: push data now
(generally not used)
RST, SYN, FIN:
connection estab
(setup, teardown
commands)
Internet
checksum
(as in UDP)
source port #
dest port #
sequence number
acknowledgement number
head not
UA P R S F
len used
checksum
Receive window
Urg data pnter
Options (variable length)
application
data
(variable length)
counting
by bytes
of data
(not segments!)
# bytes
rcvr willing
to accept
Seg. Size = IP Size
– IP Header Size
Data Size = IP Size
– IP Header Size
- TCP Header Size
Transport Layer 3-58
TCP seq. #’s and ACKs
Seq. #’s:
 byte stream
“number” of first
byte in segment’s
data
ACKs:
 seq # of next byte
expected from
other side
 cumulative ACK
Q: how receiver handles
out-of-order segments
 A: TCP spec
doesn’t say, - up to
implementor
Host A
User
types
‘C’
Host B
host ACKs
receipt of
‘C’, echoes
back ‘C’
host ACKs
receipt
of echoed
‘C’
simple telnet scenario
time
Transport Layer 3-59
TCP Round Trip Time and Timeout
Q: how to set TCP
timeout value?
 longer than RTT

but RTT varies
 if too short: premature
timeout and
 unnecessary
retransmissions
 if too long: slow
reaction to segment
loss
Q: how to estimate RTT?
 SampleRTT: measured time from
segment transmission until ACK
receipt
 ignore measurement if packet
was retransmitted.
 SampleRTT will vary, we want
the estimated RTT “smoother”
 average several recent
measurements, not just
current SampleRTT by
using a "running
average.
Transport Layer 3-60
TCP Round Trip Time and Timeout
EstimatedRTT = (1- )*EstimatedRTT + *SampleRTT
This is like a computer instruction, rather than an algebra. It may
be clearer to write:
EstimatedRTT[new] = (1-)* EstimatedRTT[old]
+ *SampleRTT[new]
where "old" and "new" are array indices, and
new = old + 1
 Exponential weighted moving average
 influence of past sample decreases exponentially fast
 typical value:  = 0.125
Transport Layer 3-61
Example RTT estimation:
Transport Layer 3-62
TCP Round Trip Time and Timeout
Setting the retransmission timeout, RTO:
 use EstimatedRTT plus “safety margin”

larger variation in EstimatedRTT -> larger safety margin
 use a running average, DevRTT, to estimate how much
SampleRTT deviates from EstimatedRTT:
DevRTT[new] = (1-)* DevRTT[old] +
 * |SampleRTT[new]-EstimatedRTT[old*]|
(typically,  = 0.25)
Note the absolute bars,|...|
Then set a new timeout interval, RTO:
Timeout Interval: RTO = EstimatedRTT + 4 * DevRTT
*Note - "old" value used, not "new" value.
Transport Layer 3-63
A[i] is a exponentially-decaying (1-c) weighed
average of x[i], x[i-1], x[i-2], …
A[i] = (1-c) * A[ i – 1 ] + c * x[ i ]
A[ i - 1] = (1-c) * A[ i – 2 ] + c * x[ i - 1 ]
A[ i - 2] = (1-c) * A[ i – 3 ] + c * x[ i - 2 ]
...
A[i] = c * { x[ i ] + (1-c) * x[ i – 1 ] + + (1-c)^2 * x[ i – 2 ] + … }
Advantages: Only the last value of A[i] needs to be remembered.
More recent values of x[i] are more heavily weighted.
0.25
0.20
c=0.25
0.15
c=0.125
0.10
0.05
0.00
1
2
3
4
5
6
7
8
9 10
64
Chapter 3 outline
 3.1 Transport-layer
services
 3.2 Multiplexing and
demultiplexing
 3.3 Connectionless
transport: UDP
 3.4 Principles of
reliable data transfer
 3.5 Connection-oriented
transport: TCP




segment structure
reliable data transfer
flow control
connection management
 3.6 Principles of
congestion control
 3.7 TCP congestion
control
Transport Layer 3-65
TCP reliable data transfer
 TCP creates rdt
service on top of IP’s
unreliable service
 Pipelined segments
Window is in bytes
 Cumulative acks*
 TCP uses single
 Retransmissions are
triggered by:


timeout events (GoBack-N)
3 duplicate acks (like
Selective Repeat*)
retransmission timer*
• Differs from Selective Repeat. If a segment is missing, the receiver
keeps sending the same ACK number (the first missing byte).
• After 3 duplicate ACKs, the sender sends only the missing
segment.
Transport Layer 3-66
Maximum Segment Size (MSS), in bytes
The initial segments (the SYN and SYN-ACK) contain the MSS in an option
field. It stays constant after this.
This tells the other host the maximum size of a segment that can be
handled by their local network (without fragmentation).
Examples, one host may say it's MSS value is 1400, the other may say it's
MSS value is 1420.
Since segments have to transverse both local networks, the smaller MSS
value is used for the connection.
TCP rules, involving Window sizes, are in number of MSS (bytes), not
number of segments.
For simplification, examples may say "the host is sending maximum size
segments," so that 1 MSS = 1 segment. Sometimes this is implied without
being stated in problems.
MSS includes the TCP header bytes (20 to 64) and data bytes, but not the IP
header bytes (20). Since Ethernet and WiFi limit datagram size to 1500
bytes, MSS is never larger than 1480 bytes when either host is on a LAN.
3-67
TCP ACK generation
[RFC 1122, RFC 2581]
Event at Receiver
TCP Receiver action
Arrival of in-order segment with
expected seq #. All data up to
expected seq # already ACKed
Delayed ACK. Wait up to 500ms
for next segment. If no next segment,
send ACK [don’t ACK every segment]
Arrival of in-order segment with
expected seq #. One other
segment has ACK pending
Immediately send single cumulative
ACK, ACKing both in-order segments
[ACK every other segment]
Arrival of out-of-order segment
higher-than-expect seq. # .
Gap detected
Immediately send duplicate ACK,
indicating seq. # of next expected byte
[1st byte in gap]
Arrival of segment that
partially or completely fills gap
Immediate send ACK for highest
continuous byte. If segment does not
start at lower end of gap, it will be
another duplicate ACK.
Transport Layer 3-68
Two windows, separate limits on bytes
sent (Flow Control & Congestion Control).
RcvrWin in each TCP header received (room in receiver buffer)
CongWin calculated by sender (in response to perceived congestion).
Sender Buffer
ACK 92,
Window 40
received
92
100
SendBase is the lowest
unacknowledged byte no.,
and it is also the highest
ACK no. received.
120
+ CongWin (30)
SendBase
= 92
+ RcvrWin (40)
ACK 120,
Window 20
received
122
132
+ CongWin (30)
SendBase+ RcvrWin (20)
= 120
140
160
Sender can not send bytes beyond either window.
MSS limits bytes in a segment, constant after SYN / SYN-ACK.
Transport Layer 3-69
TCP Flow control
LastByte
InBuffer
LastByte
ACKed
<- byte no.
 Rcvr advertises spare room
by including value of
RcvWindow in every
segment (TCP header)
 Sender limits data to
RcvWindow + ACK

guarantees receive
buffer doesn’t overflow
higher byte no.s
Receiver-Window =
 spare room in buffer
= LastByteInBuffer LastByteACKed
flow control
sender won’t overflow
receiver’s buffer by
transmitting too fast
Transport Layer 3-70
TCP sender events: RTO (timeout)
data rcvd from app:
 Create segment with seq #
 seq # is byte-stream
number of first data byte
in segment
 start RTO timer if not
already running (think of
timer as being for oldest
un-ACKed segment)
 expiration interval:
calculate new value for
TimeOutInterval(RTO)
timeout:
 retransmit segment that caused
timeout (Go-Back-N)
 RTO -> 2*RTO temporarily
 CongWin = 1 MSS (maximum
segment size, e.g. 1420 bytes)*
 restart RTO timer
ACK rcvd:
 If acknowledges previously unACKed segments


* CongWin has become very small. It can
double in size every RTT, but it would take
6 RTT to get back to 64 MSS (Slow Start
mode).


update what is known to be
ACKed (move SendBase to
ACK#)
CongWin =CongWin + MSS
RTO -> original value (from 2X)
start RTO timer if there are
outstanding segments
Transport Layer 3-71
TCP: retransmission scenarios
Host A
X
loss
SendBase
= 100
SendBase
= 120
SendBase
= 100
time
SendBase
= 120
lost ACK scenario
Host B
Seq=92 timeout
Host B
Seq=92 timeout
timeout
Host A
time
premature timeout
Transport Layer 3-72
TCP retransmission scenarios (more)
Host A
Host B
timeout
Sender Buffer
92
X
120
+ CongWin
loss
SendBase
SendBase
= 120
100
+ RcvrWin
+ CongWin
+ RcvrWin’
SendBase
time
Cumulative ACK scenario
Sender can not send bytes
beyond either window.
Every ACK has updated value of RcvrWin
Transport Layer 3-73
Fast Retransmit
 Time-out period
often relatively long:

long delay before
resending lost packet
 Detect lost segments
via duplicate ACKs.


Sender often sends
many segments backto-back
If segment is lost,
there will likely be
many duplicate ACKs.
 If sender receives 4 ACKs
for the same data (3
dups), it supposes that
segment after ACKed data
was lost:

fast retransmit: resend
segment before timer
expires
When resent packet is ACKed before a
timeout, go to Fast Recovery Mode:
- Halve "CongWin"
- Increase CongWin by 1 MSS per
CongWin bytes sent and ACKed.
Transport Layer 3-74
Fast retransmit algorithm:
event: ACK received, with ACK field value of y
if (y > SendBase) {
SendBase = y
if (there are currently not-yet-acknowledged segments)
start timer
}
else {
increment count of dup ACKs received for y
if (count of dup ACKs received for y = 3) {
resend segment with sequence number y
then resume sending new segments
CongWin = CongWin/2
}
a duplicate ACK for
fast retransmit
already ACKed segment
Transport Layer 3-75
Fast Retransmit
(after 3 duplicate ACKs)
( not CongWin limited)
Sequence
Number
(kBytes)
Retransmission
Retransmission
Retransmission.
Send time (ms)
Chapter 3 outline
 3.1 Transport-layer
services
 3.2 Multiplexing and
demultiplexing
 3.3 Connectionless
transport: UDP
 3.4 Principles of
reliable data transfer
 3.5 Connection-oriented
transport: TCP




segment structure
reliable data transfer
flow control
connection management
 3.6 Principles of
congestion control
 3.7 TCP congestion
control
Transport Layer 3-77
TCP Flow Control
 receive side of TCP
connection has a
receive buffer:
flow control
sender won’t overflow
receiver’s buffer by
transmitting too fast
<- byte no.
 speed-matching
higher byte no.s
 app process may be
slow at reading from
buffer
service: matching the
send rate to the
receiving app’s drain
rate
Every TCP header sent by the
receiver contains the value of
the receiver's Window. This
decreases as the incoming data
buffer fills up.
Transport Layer 3-78
Chapter 3 outline
 3.1 Transport-layer
services
 3.2 Multiplexing and
demultiplexing
 3.3 Connectionless
transport: UDP
 3.4 Principles of
reliable data transfer
 3.5 Connection-oriented
transport: TCP




segment structure
reliable data transfer
flow control
connection management
 3.6 Principles of
congestion control
 3.7 TCP congestion
control
Transport Layer 3-79
TCP Connection Management
Recall: TCP sender, receiver Three way handshake:
establish “connection”
before exchanging data
segments
 initialize TCP variables:
 seq. #s
 buffers, flow control
info (e.g. RcvWindow)
 client: connection initiator
Socket clientSocket = new
Socket("hostname","port
number");

server: contacted by client
Socket connectionSocket =
welcomeSocket.accept();
Step 1: client host sends TCP SYN
segment to server
 specifies initial seq #
 TCP option sends client’s MSS*
 no data
Step 2: server host receives SYN,
replies with SYN-ACK segment



server allocates buffers
specifies server initial seq. #
TCP option sends server’s MSS
Step 3: client receives SYN-ACK,
replies with ACK segment, normally
has no data
MSS - maximum segment size, usually 1420 bytes.
Transport Layer 3-80
TCP Connection Management (cont.)
Closing a connection:
client closes socket:
clientSocket.close();
client
server
close
Step 1: client end system
close
Step 2: server receives
FIN, replies with ACK.
Closes connection, sends
FIN.
timed wait
sends TCP FIN control
segment to server
closed
Or, either host sends a packet with the RES (reset) flag bit set.
This is frequent between Web servers and browsers,
Transport Layer 3-81
TCP Connection Management (cont)
TCP server
lifecycle
TCP client
lifecycle
Transport Layer 3-82
Chapter 3 outline
 3.1 Transport-layer
services
 3.2 Multiplexing and
demultiplexing
 3.3 Connectionless
transport: UDP
 3.4 Principles of
reliable data transfer
 3.5 Connection-oriented
transport: TCP




segment structure
reliable data transfer
flow control
connection management
 3.6 Principles of
congestion control
 3.7 TCP congestion
control
Transport Layer 3-83
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)
• Sender's CongWin decreases , throughput =
CongWin/RTT decreases

long delays (queueing in router buffers)
• RTT increases, throughput = Window/RTT decreases
Transport Layer 3-84
Causes/costs of congestion: scenario 1
Host A
 two senders, two
receivers
 one router,
infinite buffers
 no retransmission
Host B
lout
lin : original data
unlimited shared
output link buffers
 large delays
when congested
 maximum
achievable
throughput
Transport Layer 3-85
Causes/costs of congestion: scenario 3
H
o
s
t
A
l
o
u
t
H
o
s
t
B
Another “cost” of congestion:
 when packet dropped, any “upstream transmission
capacity used for that packet (to the point where
it was dropped) was wasted!
Transport Layer 3-86
Approaches towards congestion control
Two broad approaches towards congestion control:
End-end congestion
control:
 no explicit feedback from
network
 congestion inferred from
end-system (1) observed
loss, and (2) delay – by two
methods: (a) duplicate
ACKs or (b) RTOs.
 approach taken by TCP
Network-assisted
congestion control:
 routers provide feedback
to end systems
 single bit indicating
congestion (SNA,
DECbit, TCP/IP ECN,
ATM)
 explicit rate sender
should send at
Transport Layer 3-87
Case study: ATM ABR congestion control
ABR: available bit rate:
 “elastic service”
RM (resource management)
cells:
 if sender’s path
 sent by sender, interspersed
“underloaded”:
 sender should use
available bandwidth
 if sender’s path
congested:
 sender throttled to
minimum guaranteed
rate
with data cells
 bits in RM cell set by switches
(“network-assisted”)
 NI bit: no increase in rate
(mild congestion)
 CI bit: congestion
indication
 RM cells returned to sender by
receiver, with bits intact
Transport Layer 3-88
Chapter 3 outline
 3.1 Transport-layer
services
 3.2 Multiplexing and
demultiplexing
 3.3 Connectionless
transport: UDP
 3.4 Principles of
reliable data transfer
 3.5 Connection-oriented
transport: TCP




segment structure
reliable data transfer
flow control
connection management
 3.6 Principles of
congestion control
 3.7 TCP congestion
control
Transport Layer 3-89
TCP congestion control: additive increase,
multiplicative decrease

Approach: increase transmission rate (window size), probing for
usable bandwidth, until loss occurs
 additive increase: increase CongWin by 1 MSS every RTT
until loss detected
 multiplicative decrease: cut CongWin in half after loss
congestion
window
Saw tooth
behavior: probing
for bandwidth
congestion window size
24 Kbytes
Lost packet (3 duplicate ACKs)
16 Kbytes
8 Kbytes
* MSS = maximum segment size (e.g.. , 1280
bytes). specified as a TCP option in SYN and SYN-ACK
packets.
time
time
Transport Layer 3-90
TCP Congestion Control: details
 sender limits transmission:
LastByteSent-LastByteAcked
 CongWin
 Roughly,
rate =
CongWin
Bytes/sec
RTT
 CongWin is dynamic, function
of perceived network
congestion
How does sender perceive
congestion?
 loss event = timeout (RTO) or
3 duplicate ACKs
 TCP sender reduces rate
(CongWin) after loss event
mechanisms:
 AIMD (additive increase,
multiplicative decrease)
 slow start (after RTO)

conservative after timeout
events (CongWin = 1 MSS)
Transport Layer 3-91
TCP Slow Start
 When connection begins,
CongWin = 1 MSS*


Example: MSS = 500
bytes & RTT = 200 msec
initial rate = 20 kbps
 available bandwidth may
be >> MSS/RTT

desirable to quickly ramp
up to respectable rate
 When connection begins,
increase rate
exponentially fast until
first loss event (doubles
every RTT)
 or until CongWin =
Receiver-Window.
* MSS = maximum segment size (e.g.. , 1420 bytes).
specified as a TCP option in SYN and SYN-ACK packets.
Transport Layer 3-92
TCP Slow Start (more)
 When connection begins,


Host B
RTT
increase rate
exponentially until first
loss event:
Host A
double CongWin every RTT
done by incrementing
CongWin for every ACK
received (by d-SendBase)
 Summary: initial rate is
slow but ramps up
exponentially fast
time
Transport Layer 3-93
Q: When should the
exponential
increase switch to
linear?
A: When CongWin
gets to 1/2 of its
value before
timeout.
Implementation:
Congestion Window /mss
Refinement
 Variable Threshold
“Fast Recovery”
(after 3 dup. ACKs)
1/2
“Slow Start”
Transmission Round
(time/rtt)
 At loss event, Threshold is
set to 1/2 of CongWin just
before loss event
Transport Layer 3-94
CongWin <= Threshold: Doubles each RTT (add MSS for each MSS ACKed)
CongWin > Threshold: Adds MSS each RTT
Time Out: Threshold = 1/2 CongWin, CongWin = 1 (Slow-Start)
3-Dup Ack: Threshold = 1/2 CongWin, CongWin = Threshold (Fast Recovery)
CongWin / mss
Time Out
CongWin = 20
3 Dup. ACKs
12
6
Transport Layer 3-96
Refinement: inferring loss
 After 3 dup ACKs:
is cut in half
 window then grows
linearly
 But after timeout event:
 CongWin instead set to
1 MSS;
 window then grows
exponentially
 to a threshold, then
grows linearly
 CongWin
Philosophy:
 3 dup ACKs indicates
network capable of
delivering some segments
 timeout indicates a
“more alarming”
congestion scenario
Transport Layer 3-96
Summary: TCP Congestion Control
 When CongWin is below Threshold, sender in slow-
start phase, window grows exponentially.
 When CongWin is above Threshold, sender is in
congestion-avoidance phase, window grows linearly.
 When a triple duplicate ACK occurs, Threshold set
to CongWin/2 and CongWin set to Threshold
(no slow start, linear mode immediately).
 When timeout occurs, Threshold set to CongWin/2
and CongWin is set to 1 MSS.
Transport Layer 3-97
TCP sender congestion control
State
Event
TCP Sender Action
Commentary
Slow Start
(SS)
ACK receipt
for previously
unacked
data
CongWin = CongWin + MSS,
If (CongWin > Threshold)
set state to “Congestion
Avoidance”
Resulting in a doubling of
CongWin every RTT
Congestion
Avoidance
(CA)
ACK receipt
for previously
unacked
data
CongWin = CongWin+MSS *
(MSS/CongWin)
Additive increase, resulting
in increase of CongWin by
1 MSS every RTT
SS or CA
Loss event
detected by
triple
duplicate
ACK
Threshold = CongWin/2,
CongWin = Threshold,
Set state to “Congestion
Avoidance”
Fast recovery,
implementing multiplicative
decrease. CongWin will not
drop below 1 MSS.
SS or CA
Timeout
Threshold = CongWin/2,
CongWin = 1 MSS,
Set state to “Slow Start”
Enter slow start
SS or CA
Duplicate
ACK
Increment duplicate ACK count
for segment being ACKed
CongWin and Threshold not
changed
Transport Layer 3-98
TCP throughput
 What’s the average throughout of TCP as
a function of window size and RTT?

Ignore slow start
 Let W be the window size when loss occurs.
 When window is W, throughput is W/RTT
 Just after loss, window drops to W/2,
throughput to W/2RTT.
 Average throughout: .75 W/RTT (assuming
loss occurs every time CongWin grows to
W)
Transport Layer 3-99
TCP Futures: TCP over “long, fat pipes”
 Example: With 1500 byte segments and 100ms
RTT, if we want 10 Gbps throughput:
 Requires window size, W = 125,000,000 bytes
(for 83,333 1500-byte in-flight segments)
 New versions of TCP for high-speed needed!
 Present W of 65,000 bytes -> 5 Mbps
The TCP Window-Multiplier option allows the Window to
be incearsed by a exponent of 2.
Transport Layer 3-100
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
TCP
connection 2
bottleneck
router
capacity R
Transport Layer 3-101
Why is TCP fair?
Two competing sessions:
 Additive increase gives slope of 1, as throughout increases
 multiplicative decrease decreases throughput proportionally
Rmax
Finally: equal bandwidth share
loss: decrease windows by factor of 2
congestion avoidance: additive increase
R1,R2 -losses: decrease windows by factor of 2
congestion avoidance: additive increase
R1/2,R2/2
Start: R1 = 5 x R2, R1 +R2 < Rmax
Start
Connection 1 throughput
Rmax
Transport Layer 3-102
Fairness (more)
Fairness and UDP
 Multimedia apps often
do not use TCP

do not want rate
throttled by congestion
control
 Instead use UDP:
 pump audio/video at
constant rate, tolerate
packet loss
 Research area: make
UDP more TCP friendly

Solution: reserve 50% of
router buffer space for
TCP segments (excess
UDP segments dropped).
Fairness and parallel TCP
connections
 nothing prevents app from
opening parallel connections
between 2 hosts.
 Web browsers do this
 Example: link of rate R
supporting 9 connections;


new app asks for 1 TCP, gets
rate R/10
new app asks for 10 TCPs, gets
R/2 !
Browsers using parallel
connections are not fair, but
work better.
Fairness is not an issue if there is
no congestion.
Transport Layer 3-103
Fixed congestion window (1)
First case:
WS > R * RTT (Rate Limited)
ACK for first segment in
window returns before
window’s worth of data
sent (so no idle time)
Data sent at full rate, R
R * RTT is the Bandwidth-Delay Product
Transport Layer 3-104
Fixed congestion window (2)
Second case:
 WS < R*RTT
 (Window Limited)
 wait for ACK after sending
window’s worth of data
sent
Data sent at rate
WS/RTT < R
Transport Layer 3-105
TCP Delay Modeling: Slow Start (2)
Delay components:
• 2 RTT for connection
estab and request
• O/R to transmit object
• time server idles due to
slow start
initiate TCP
connection
request
object
first window
= S/R
RTT
second window
= 2S/R
Server idles: for RTT
third window
= 4S/R
fourth window
= 8S/R
•O is object size,
•R is transmission rate
complete
transmission
object
delivered
•RTT is round-trip-time.
time at
client
time at
server
Transport Layer 3-106
TCP Delay Modeling (3)
As the CongWin grows, the connection can change from being
"Window Limited" (DataRate=W/RTT) to being
"Transmission Rate Limited" (DataRate = R).
initiate TCP
connection
request
object
first window
= S/R
RTT
second window
= 2S/R
third window
= 4S/R
fourth window
= 8S/R
complete
transmission
object
delivered
time at
client
time at
server
Transport Layer 3-107
32 bits
TCP Options
source port #
Options appear in extra
bytes appended to the
end of the TCP header:
number of bytes,
option ID,
parameter.
A receiver can ignore
Options it does not
understand (extensible).
dest port #
sequence number
acknowledgement number
head not
UA P R S F
len used
checksum
Receive window
Urg data pnter
20
Options (variable length)
data
<60
Common Options
Time Stamp (RFC1323)
Window Scale Option (RFC1323) Multiply Window by 2^n.
SACK (RFC2018) Acknowledge non-contiguous blocks received.
NOP - Pads out options to multiple of 4 bytes
3-108
Chapter 3: Summary
 principles behind transport
layer services:
 multiplexing,
demultiplexing
 reliable data transfer
 flow control
 congestion control
 instantiation and
implementation in the
Internet
 UDP
 TCP
Next:
 leaving the network
“edge” (application,
transport layers)
 into the network
“core”
Transport Layer 3-109