Chapter 3 v6.01 (1)

Transport Layer 3-1
Chapter 3
Transport Layer
Computer
Networking: A
Top Down
Approach
6th edition
Jim Kurose, Keith Ross
Addison-Wesley
March 2012
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 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:
 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!)
 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
material.
Thanks and enjoy! JFK/KWR
All material copyright 1996-2013
J.F Kurose and K.W. Ross, All Rights Reserved

Transport Layer 3-2
Chapter 3: Transport Layer
our goals:
 understand
principles behind
transport layer
services:
 multiplexing,
demultiplexing
 reliable data
transfer
 flow control
 congestion control
 learn about Internet
transport layer
protocols:
 UDP: connectionless
transport
 TCP: connection-
oriented reliable
transport
 TCP congestion control

Transport Layer 3-3
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 3-4
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
 rcv side: reassembles
segments into
messages, passes to
app layer
 more than one transport
protocol available to apps
application
transport
network
data link
physical
application
transport
network
data link
physical

Transport Layer 3-5
Transport vs. network layer
 network layer:
logical
communication
between hosts
 transport layer:
logical
communication
between
processes
 relies on,
enhances,
network layer
services
12 kids in Ann’s house
sending letters to 12 kids
in Bill’s house:
 hosts = houses
 processes = kids
 app messages = letters
in envelopes
 transport protocol = Ann
and Bill who demux to in-
house siblings
 network-layer protocol =
postal service
household analogy:

Transport Layer 3-6
Internet transport-layer protocols
 reliable, in-order
delivery (TCP)
 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
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
network
data link
physical
network
data link
physical

Transport Layer 3-7
Chapter 3 outline
3.1 transport-layer
services
demultiplexing
3.3 connectionless
transport: UDP
3.4 principles of
reliable data
transfer
transport: TCP
 flow control
 connection
management
3.6 principles of
congestion control
3.7 TCP congestion
control

Transport Layer 3-8
Multiplexing/demultiplexing
process
socket
use header info to deliver
received segments to correct
socket
demultiplexing at receiver:
handle data from multiple
sockets, add transport
header (later used for
demultiplexing)
multiplexing at sender:
transport
application
physical
link
network
P2
P1
transport
application
physical
link
network
P4
transport
application
physical
link
network
P3

Transport Layer 3-9
How demultiplexing works
 host receives IP
datagrams
 each datagram has
source IP address,
destination IP address
 each datagram carries
one transport-layer
segment
 each segment has
source, destination port
number
 host uses IP addresses
& port numbers to direct
segment to appropriate
socket
source port # dest port #
32 bits
application
data
(payload)
other header fields
TCP/UDP segment format

Transport Layer 3-10
Connectionless demultiplexing
 recall: created socket has
host-local port #:
DatagramSocket mySocket1
= new DatagramSocket(12534);
 when host receives
UDP segment:
 checks destination port
# in segment
 directs UDP segment to
socket with that port #
 recall: when creating
datagram to send into
UDP socket, must
specify
 destination IP address
 destination port #
IP datagrams with
same dest. port #, but
different source IP
addresses and/or
source port numbers
will be directed to same
socket at dest

Connectionless demux:
example
DatagramSocket
serverSocket = new
DatagramSocket
(6428);
transport
application
physical
link
network
P3
transport
application
physical
link
network
P1
transport
application
physical
link
network
P4
DatagramSocket
mySocket1 = new
DatagramSocket
(5775);
DatagramSocket
mySocket2 = new
DatagramSocket
(9157);
source port: 9157
dest port: 6428
source port: 6428
dest port: 9157
source port: ?
dest port: ?
source port: ?
dest port: ?

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

Connection-oriented demux:
example
transport
application
physical
link
network
P3
transport
application
physical
link
P4
transport
application
physical
link
network
P2
source IP,port: A,9157
dest IP, port: B,80
source IP,port: B,80
dest IP,port: A,9157
host: IP
address
A
host: IP
address
C
network
P6
P5
P3
source IP,port: C,5775
dest IP,port: B,80
dest IP,port: B,80
three segments, all destined to IP address: B,
dest port: 80 are demultiplexed to different sockets
server:
IP
address
B

Connection-oriented demux:
example
transport
application
physical
link
network
P3
transport
application
physical
link
transport
application
physical
link
network
P2
source IP,port: A,9157
dest IP, port: B,80
source IP,port: B,80
dest IP,port: A,9157
host: IP
address
A
host: IP
address
C
server:
IP
address
B
network
P3
dest IP,port: B,80
dest IP,port: B,80
P4
threaded server

Chapter 3 outline
3.1 transport-layer
services
demultiplexing
3.3 connectionless
transport: UDP
3.4 principles of
reliable data
transfer
transport: TCP
 flow control
 connection
management
3.6 principles of
congestion control
3.7 TCP congestion
control

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
 UDP use:
 streaming multimedia
apps (loss tolerant,
rate sensitive)
 DNS
 SNMP
 reliable transfer over
UDP:
 add reliability at
application layer
 application-specific
error recovery!

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

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

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
1 1 0 1 1 1 0 1 1 1 0 1 1 1 0 1 1
1 1 0 1 1 1 0 1 1 1 0 1 1 1 1 0 0
1 0 1 0 0 0 1 0 0 0 1 0 0 0 0 1 1
wraparound
sum
checksum
Note: when adding numbers, a carryout from the most
significant bit needs to be added to the result

Chapter 3 outline
3.1 transport-layer
services
demultiplexing
3.3 connectionless
transport: UDP
3.4 principles of
reliable data
transfer
transport: TCP
 flow control
 connection
management
3.6 principles of
congestion control
3.7 TCP congestion
control

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)

transfer

Reliable data transfer: getting started
send
side
receive
side
rdt_send(): called from above,
(e.g., by app.). Passed data to
deliver to receiver upper layer
udt_send(): called by rdt,
to transfer packet over
unreliable channel to receiver
rdt_rcv(): called when packet
arrives on rcv-side of channel
deliver_data(): called by
rdt to deliver data to upper

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
1
state
2
event causing state transition
actions taken on state transition
state: when in this
“state” next state
uniquely determined
by next event
event
actions
Reliable data transfer: getting started

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

 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
 receiver feedback: control msgs (ACK,NAK) rcvr-
>sender
rdt2.0: channel with bit errors
How do humans recover from “errors”
during conversation?

 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
rdt2.0: channel with bit errors

rdt2.0: FSM specification
Wait for
call from
above
sndpkt = make_pkt(data, checksum)
udt_send(sndpkt)
extract(rcvpkt,data)
deliver_data(data)
udt_send(ACK)
rdt_rcv(rcvpkt) &&
notcorrupt(rcvpkt)
rdt_rcv(rcvpkt) && isACK(rcvpkt)
udt_send(sndpkt)
rdt_rcv(rcvpkt) &&
isNAK(rcvpkt)
udt_send(NAK)
rdt_rcv(rcvpkt) &&
corrupt(rcvpkt)
Wait for
ACK or
NAK
Wait for
call from
below
sender
receiver
rdt_send(data)
L

rdt2.0: operation with no errors
Wait for
call from
above
snkpkt = make_pkt(data, checksum)
udt_send(sndpkt)
deliver_data(data)
udt_send(ACK)
rdt_rcv(rcvpkt) &&
notcorrupt(rcvpkt)
udt_send(sndpkt)
rdt_rcv(rcvpkt) &&
isNAK(rcvpkt)
udt_send(NAK)
rdt_rcv(rcvpkt) &&
corrupt(rcvpkt)
Wait for
ACK or
NAK
Wait for
call from
below
rdt_send(data)
L

rdt2.0: error scenario
Wait for
call from
above
snkpkt = make_pkt(data, checksum)
udt_send(sndpkt)
deliver_data(data)
udt_send(ACK)
rdt_rcv(rcvpkt) &&
notcorrupt(rcvpkt)
udt_send(sndpkt)
rdt_rcv(rcvpkt) &&
isNAK(rcvpkt)
udt_send(NAK)
rdt_rcv(rcvpkt) &&
corrupt(rcvpkt)
Wait for
ACK or
NAK
Wait for
call from
below
rdt_send(data)
L

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
corrupted
 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

rdt2.1: sender, handles garbled
ACK/NAKs
Wait for
call 0 from
above
sndpkt = make_pkt(0, data, checksum)
udt_send(sndpkt)
rdt_send(data)
Wait for
ACK or
NAK 0 udt_send(sndpkt)
rdt_rcv(rcvpkt) &&
( corrupt(rcvpkt) ||
isNAK(rcvpkt) )
udt_send(sndpkt)
rdt_send(data)
rdt_rcv(rcvpkt)
&& notcorrupt(rcvpkt)
&& isACK(rcvpkt)
udt_send(sndpkt)
rdt_rcv(rcvpkt) &&
isNAK(rcvpkt) )
rdt_rcv(rcvpkt)
&& isACK(rcvpkt)
Wait for
call 1 from
above
Wait for
ACK or
NAK 1
L
L

Wait for
0 from
below
sndpkt = make_pkt(NAK, chksum)
udt_send(sndpkt)
rdt_rcv(rcvpkt) &&
not corrupt(rcvpkt) &&
has_seq0(rcvpkt)
rdt_rcv(rcvpkt) && notcorrupt(rcvpkt)
&& has_seq1(rcvpkt)
deliver_data(data)
sndpkt = make_pkt(ACK, chksum)
udt_send(sndpkt)
Wait for
1 from
below
&& has_seq0(rcvpkt)
deliver_data(data)
udt_send(sndpkt)
rdt_rcv(rcvpkt) && (corrupt(rcvpkt)
udt_send(sndpkt)
rdt_rcv(rcvpkt) &&
not corrupt(rcvpkt) &&
has_seq1(rcvpkt)
rdt_rcv(rcvpkt) && (corrupt(rcvpkt)
udt_send(sndpkt)
sndpkt = make_pkt(NAK, chksum)
udt_send(sndpkt)
rdt2.1: receiver, handles garbled
ACK/NAKs

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
“expected” pkt
should have seq # of
0 or 1
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

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

rdt2.2: sender, receiver fragments
Wait for
call 0 from
above
udt_send(sndpkt)
rdt_send(data)
udt_send(sndpkt)
rdt_rcv(rcvpkt) &&
isACK(rcvpkt,1) )
rdt_rcv(rcvpkt)
&& isACK(rcvpkt,0)
Wait for
ACK
0
sender FSM
fragment
&& has_seq1(rcvpkt)
deliver_data(data)
sndpkt = make_pkt(ACK1, chksum)
udt_send(sndpkt)
Wait for
0 from
below
rdt_rcv(rcvpkt) &&
(corrupt(rcvpkt) ||
has_seq1(rcvpkt))
udt_send(sndpkt)
receiver FSM
fragment
L

rdt3.0: channels with errors and loss
new assumption:
underlying channel
can also lose
packets (data,
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 seq. #’s
already handles this
 receiver must specify
seq # of pkt being
ACKed
 requires countdown timer

rdt3.0 sender
udt_send(sndpkt)
start_timer
rdt_send(data)
Wait
for
ACK0
rdt_rcv(rcvpkt) &&
isACK(rcvpkt,1) )
Wait for
call 1 from
above
udt_send(sndpkt)
start_timer
rdt_send(data)
rdt_rcv(rcvpkt)
&& isACK(rcvpkt,0)
rdt_rcv(rcvpkt) &&
isACK(rcvpkt,0) )
rdt_rcv(rcvpkt)
&& isACK(rcvpkt,1)
stop_timer
stop_timer
udt_send(sndpkt)
start_timer
timeout
udt_send(sndpkt)
start_timer
timeout
rdt_rcv(rcvpkt)
Wait for
call 0from
above
Wait
for
ACK1
L
rdt_rcv(rcvpkt)
L
L
L

sender receiver
rcv pkt1
rcv pkt0
send ack0
send ack1
send ack0
rcv ack0
send pkt0
send pkt1
rcv ack1
send pkt0
rcv pkt0
pkt0
pkt0
pkt1
ack1
ack0
ack0
(a) no loss
sender receiver
rcv pkt1
rcv pkt0
send ack0
send ack1
send ack0
rcv ack0
send pkt0
send pkt1
rcv ack1
send pkt0
rcv pkt0
pkt0
pkt0
ack1
ack0
ack0
(b) packet loss
pkt1
X
loss
pkt1
timeout
resend pkt1
rdt3.0 in action

rdt3.0 in action
rcv pkt1
send ack1
(detect duplicate)
pkt1
sender receiver
rcv pkt1
rcv pkt0
send ack0
send ack1
send ack0
rcv ack0
send pkt0
send pkt1
rcv ack1
send pkt0
rcv pkt0
pkt0
pkt0
ack1
ack0
ack0
(c) ACK loss
ack1
X
loss
pkt1
timeout
resend pkt1
rcv pkt1
send ack1
(detect duplicate)
pkt1
sender receiver
rcv pkt1
send ack0
rcv ack0
send pkt1
send pkt0
rcv pkt0
pkt0
ack0
(d) premature timeout/ delayed ACK
pkt1
timeout
resend pkt1
ack1
send ack1
send pkt0
rcv ack1
pkt0
ack1
ack0
send pkt0
rcv ack1 pkt0
rcv pkt0
send ack0
ack0
rcv pkt0
send ack0
(detect duplicate)

Performance of rdt3.0
 rdt3.0 is correct, but performance stinks
 e.g.: 1 Gbps link, 15 ms prop. delay, 8000 bit
packet:
 U sender: utilization – fraction of time sender busy
sending
U
sender =
.008
30.008
= 0.00027
L / R
RTT + L / R
=
 if RTT=30 msec, 1KB pkt every 30 msec: 33kB/sec
thruput over 1 Gbps link
 network protocol limits use of physical resources!
Dtrans =
L
R
8000 bits
109 bits/sec
= = 8 microsecs

rdt3.0: stop-and-wait operation
first packet bit transmitted, t = 0
sender receiver
RTT
last packet bit transmitted, t = L / R
first packet bit arrives
last packet bit arrives, send ACK
ACK arrives, send next
packet, t = RTT + L / R
U
sender =
.008
30.008
= 0.00027
L / R
RTT + L / R
=

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

Pipelining: increased utilization
first packet bit transmitted, t = 0
sender receiver
RTT
last bit transmitted, t = L / R
first packet bit arrives
last packet bit arrives, send ACK
ACK arrives, send next
packet, t = RTT + L / R
last bit of 2nd packet arrives, send ACK
last bit of 3rd packet arrives, send ACK
3-packet pipelining increases
utilization by a factor of 3!
U
sender =
.0024
30.008
= 0.00081
3L / R
RTT + L / R
=

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

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

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

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 #
Wait
udt_send(sndpkt)
default
rdt_rcv(rcvpkt)
&& notcurrupt(rcvpkt)
&& hasseqnum(rcvpkt,expectedseqnum)
deliver_data(data)
sndpkt = make_pkt(expectedseqnum,ACK,chksum)
udt_send(sndpkt)
expectedseqnum++
expectedseqnum=1
sndpkt =
make_pkt(expectedseqnum,ACK,chksum)
L
GBN: receiver extended FSM

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

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

Selective repeat: sender, receiver
windows

Selective repeat
data from above:
 if next available seq # in
window, send pkt
timeout(n):
 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 #
sender
pkt n in [rcvbase, rcvbase+N-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
receiver

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

Selective repeat:
dilemma
example:
 seq #’s: 0, 1, 2, 3
 window size=3
receiver window
(after receipt)
sender window
(after receipt)
0 1 2 3 0 1 2
0 1 2 3 0 1 2
0 1 2 3 0 1 2
pkt0
pkt1
pkt2
0 1 2 3 0 1 2 pkt0
timeout
retransmit pkt0
0 1 2 3 0 1 2
0 1 2 3 0 1 2
0 1 2 3 0 1 2
X
X
X
will accept packet
with seq number 0
(b) oops!
0 1 2 3 0 1 2
0 1 2 3 0 1 2
0 1 2 3 0 1 2
pkt0
pkt1
pkt2
0 1 2 3 0 1 2
pkt0
0 1 2 3 0 1 2
0 1 2 3 0 1 2
0 1 2 3 0 1 2
X
will accept packet
with seq number 0
0 1 2 3 0 1 2 pkt3
(a) no problem
receiver can’t see sender side.
receiver behavior identical in both cases!
something’s (very) wrong!
 receiver sees no
difference in two
scenarios!
 duplicate data
accepted as new in
(b)
Q: what relationship
between seq # size
and window size to
avoid problem in
(b)?

Chapter 3 outline
3.1 transport-layer
services
demultiplexing
3.3 connectionless
transport: UDP
3.4 principles of
reliable data
transfer
transport: TCP
 flow control
 connection
management
3.6 principles of
congestion control
3.7 TCP congestion
control

TCP: Overview 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) inits sender,
receiver state before
data exchange
 flow controlled:
 sender will not
overwhelm receiver
 point-to-point:
 one sender, one
receiver
 reliable, in-order byte
steam:
 no “message
boundaries”
 pipelined:
 TCP congestion and
flow control set
window size

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

TCP seq. numbers, ACKs
sequence numbers:
byte stream “number”
of first byte in
segment’s data
acknowledgements:
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
sequence number
checksum
rwnd
urg pointer
incoming segment to sender
A
sent
ACKed
sent, not-
yet ACKed
(“in-
flight”)
usable
but not
yet sent
not
usable
window size
N
sender sequence number space
sequence number
checksum
rwnd
urg pointer
outgoing segment from sender

TCP seq. numbers, ACKs
User
types
‘C’
host ACKs
receipt
of echoed
‘C’
host ACKs
receipt of
‘C’, echoes
back ‘C’
simple telnet scenario
Host B
Host A
Seq=42, ACK=79, data = ‘C’
Seq=79, ACK=43, data = ‘C’
Seq=43, ACK=80

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

RTT: gaia.cs.umass.edu to fantasia.eurecom.fr
100
150
200
250
300
350
1 8 15 22 29 36 43 50 57 64 71 78 85 92 99 106
time (seconnds)
RTT
(milliseconds)
SampleRTT Estimated RTT
EstimatedRTT = (1- )*EstimatedRTT + *SampleRTT
 exponential weighted moving average
 influence of past sample decreases
exponentially fast
 typical value:  = 0.125
RTT
(milliseconds)
RTT: gaia.cs.umass.edu to fantasia.eurecom.fr
sampleRTT
EstimatedRTT
time (seconds)

 timeout interval: EstimatedRTT plus “safety
margin”
 large variation in EstimatedRTT -> larger safety margin
 estimate SampleRTT deviation from EstimatedRTT:
DevRTT = (1-)*DevRTT +
*|SampleRTT-EstimatedRTT|
(typically,  = 0.25)
TimeoutInterval = EstimatedRTT + 4*DevRTT
estimated RTT “safety margin”

Chapter 3 outline
3.1 transport-layer
services
demultiplexing
3.3 connectionless
transport: UDP
3.4 principles of
reliable data
transfer
transport: TCP
 flow control
 connection
management
3.6 principles of
congestion control
3.7 TCP congestion
control

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

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

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

TCP: retransmission scenarios
lost ACK scenario
Host B
Host A
Seq=92, 8 bytes of data
ACK=100
X
timeout
ACK=100
premature timeout
Host B
Host A
ACK=100
Seq=92, 8
bytes of data
timeout
ACK=120
ACK=120
SendBase=100
SendBase=120
SendBase=120
SendBase=92

TCP: retransmission scenarios
X
cumulative ACK
Host B
Host A
ACK=100
timeout
ACK=120

TCP ACK generation [RFC 1122, RFC 2581]
event at receiver
arrival of in-order segment with
expected seq #. All data up to
expected seq # already ACKed
arrival of in-order segment with
expected seq #. One other
segment has ACK pending
arrival of out-of-order segment
higher-than-expect seq. # .
Gap detected
arrival of segment that
partially or completely fills gap
TCP receiver action
delayed ACK. Wait up to 500ms
for next segment. If no next segment,
send ACK
immediately send single cumulative
ACK, ACKing both in-order segments
immediately send duplicate ACK,
indicating seq. # of next expected byte
immediate send ACK, provided that
segment starts at lower end of gap

TCP 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
back-to-back
 if segment is lost,
there will likely be
many duplicate
ACKs.
if sender receives 3
ACKs for same data
(“triple duplicate
ACKs”), resend
unacked segment
with smallest seq #
 likely that unacked
segment lost, so
don’t wait for timeout
TCP fast retransmit
(“triple duplicate
ACKs”),

X
fast retransmit after sender
receipt of triple duplicate ACK
Host B
Host A
ACK=100
timeout
ACK=100
ACK=100
ACK=100
TCP fast retransmit

Chapter 3 outline
3.1 transport-layer
services
demultiplexing
3.3 connectionless
transport: UDP
3.4 principles of
reliable data
transfer
transport: TCP
 flow control
 connection
management
3.6 principles of
congestion control
3.7 TCP congestion
control

TCP flow control
application
process
TCP socket
receiver buffers
TCP
code
IP
code
application
OS
receiver protocol stack
application may
remove data from
TCP socket buffers ….
… slower than TCP
receiver is delivering
(sender is sending)
from sender
receiver controls sender, so
sender won’t overflow
receiver’s buffer by
transmitting too much, too fast
flow control

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

Chapter 3 outline
3.1 transport-layer
services
demultiplexing
3.3 connectionless
transport: UDP
3.4 principles of
reliable data
transfer
transport: TCP
 flow control
 connection
management
3.6 principles of
congestion control
3.7 TCP congestion
control

Connection Management
before exchanging data, sender/receiver
“handshake”:
 agree to establish connection (each knowing the other
willing to establish connection)
 agree on connection parameters
connection state: ESTAB
connection variables:
seq # client-to-server
server-to-client
rcvBuffer size
at server,client
application
network
connection state: ESTAB
connection Variables:
seq # client-to-server
server-to-client
rcvBuffer size
at server,client
application
network
Socket clientSocket =
newSocket("hostname","port
number");
Socket connectionSocket =
welcomeSocket.accept();

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

Agreeing to establish a connection
2-way handshake failure scenarios:
retransmit
req_conn(x)
ESTAB
req_conn(x)
half open connection!
(no client!)
client
terminates
server
forgets x
connection
x completes
retransmit
req_conn(x)
ESTAB
req_conn(x)
data(x+1)
retransmit
data(x+1)
accept
data(x+1)
choose x
req_conn(x)
ESTAB
ESTAB
acc_conn(x)
client
terminates
ESTAB
choose x
req_conn(x)
ESTAB
acc_conn(x)
data(x+1) accept
data(x+1)
connection
x completes server
forgets x

TCP 3-way handshake
SYNbit=1, Seq=x
choose init seq num, x
send TCP SYN msg
ESTAB
SYNbit=1, Seq=y
ACKbit=1; ACKnum=x+1
choose init seq num, y
send TCP SYNACK
msg, acking SYN
ACKbit=1, ACKnum=y+1
received SYNACK(x)
indicates server is live;
send ACK for SYNACK;
this segment may contain
client-to-server data
received ACK(y)
indicates client is live
SYNSENT
ESTAB
SYN RCVD
client state
LISTEN
server state
LISTEN

TCP 3-way handshake:
FSM
closed
L
listen
SYN
rcvd
SYN
sent
ESTAB
Socket clientSocket =
newSocket("hostname","port
number");
SYN(seq=x)
Socket connectionSocket =
welcomeSocket.accept();
SYN(x)
SYNACK(seq=y,ACKnum=x+1)
create new socket for
communication back to client
SYNACK(seq=y,ACKnum=x+1)
ACK(ACKnum=y+1)
ACK(ACKnum=y+1)
L

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

FIN_WAIT_2
CLOSE_WAIT
FINbit=1, seq=y
ACKbit=1; ACKnum=y+1
ACKbit=1; ACKnum=x+1
wait for server
close
can still
send data
can no longer
send data
LAST_ACK
CLOSED
TIMED_WAIT
timed wait
for 2*max
segment lifetime
CLOSED
TCP: closing a connection
FIN_WAIT_1 FINbit=1, seq=x
can no longer
send but can
receive data
clientSocket.close()
client state server state
ESTAB
ESTAB

Chapter 3 outline
3.1 transport-layer
services
demultiplexing
3.3 connectionless
transport: UDP
3.4 principles of
reliable data
transfer
transport: TCP
 flow control
 connection
management
3.6 principles of
congestion control
3.7 TCP 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!
Principles of congestion control

Causes/costs of congestion: scenario
1
 two senders, two
receivers
 one router, infinite
buffers
 output link capacity: R
 no retransmission
 maximum per-
connection throughput:
R/2
unlimited shared
output link buffers
Host A
original data: lin
Host B
throughput: lout
R/2
R/2
l
out
lin R/2
delay lin
 large delays as arrival rate,
lin, approaches capacity

 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
finite shared output
link buffers
Host A
lin : original data
Host B
lout
l'in: original data, plus
retransmitted data
‘
2

idealization: perfect
knowledge
 sender sends only
when router buffers
available
link buffers
lin : original data
lout
retransmitted data
copy
free buffer space!
R/2
R/2
l
out
lin
2
Host B
A

lin : original data
lout
retransmitted data
copy
no buffer space!
Idealization: known
loss packets can be
lost, dropped at router
due to full buffers
 sender only resends if
packet known to be
lost
2
A
Host B

lin : original data
lout
retransmitted data
free buffer space!
2
Idealization: known
loss packets can be
lost, dropped at router
due to full buffers
 sender only resends if
packet known to be
lost
R/2
R/2
lin
l
out
when sending at R/2,
some packets are
retransmissions but
asymptotic goodput
is still R/2 (why?)
A
Host B

A
lin
lout
l'in
copy
free buffer space!
timeout
R/2
R/2
lin
l
out
some packets are
retransmissions
including duplicated
that are delivered!
Host B
Realistic: duplicates
 packets can be lost,
dropped at router due to
full buffers
 sender times out
prematurely, sending two
copies, both of which are
delivered
2

R/2
l
out
some packets are
retransmissions
including duplicated
that are delivered!
“costs” of congestion:
 more work (retrans) for given “goodput”
 unneeded retransmissions: link carries multiple copies
of pkt
 decreasing goodput
R/2
lin
2
Realistic: duplicates
 packets can be lost,
dropped at router due to
full buffers
 sender times out
prematurely, sending two
copies, both of which are
delivered

 four senders
 multihop paths
 timeout/retransmit
Q: what happens as lin and
lin
’ increase ?
link buffers
Host A lout
3
Host B
Host C
Host D
lin : original data
retransmitted data
A: as red lin
’ increases, all
arriving blue pkts at upper
queue are dropped, blue
throughput g 0

another “cost” of congestion:
 when packet dropped, any “upstream
transmission capacity used for that packet
was wasted!
3
C/2
C/2
l
out
lin
’

Approaches towards congestion
control
two broad approaches towards congestion control:
end-end
congestion
control:
 no explicit feedback
from network
 congestion inferred
from end-system
observed loss,
delay
 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 for
sender to send at

Case study: ATM ABR congestion
control
ABR: available bit
rate:
 “elastic service”
 if sender’s path
“underloaded”:
 sender should use
available bandwidth
 if sender’s path
congested:
 sender throttled to
minimum
guaranteed rate
RM (resource
management) cells:
 sent by sender,
interspersed 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

Case study: ATM ABR congestion
control
 two-byte ER (explicit rate) field in RM cell
 congested switch may lower ER value in cell
 senders’ send rate thus max supportable rate on
path
 EFCI bit in data cells: set to 1 in congested
switch
 if data cell preceding RM cell has EFCI set, receiver
sets CI bit in returned RM cell
RM cell data cell

Chapter 3 outline
3.1 transport-layer
services
demultiplexing
3.3 connectionless
transport: UDP
3.4 principles of
reliable data
transfer
transport: TCP
 flow control
 connection
management
3.6 principles of
congestion control
3.7 TCP congestion
control

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
cwnd:
TCP
sender
congestion
window
size
AIMD saw tooth
behavior: probing
for bandwidth
additively increase window size …
…. until loss occurs (then cut window in half)
time

TCP Congestion Control:
details
 sender limits
transmission:
 cwnd is dynamic,
function of perceived
network congestion
TCP sending rate:
 roughly: send cwnd
bytes, wait RTT for
ACKS, then send
more bytes
last byte
ACKed sent, not-
yet ACKed
(“in-
flight”)
last byte
sent
cwnd
LastByteSent-
LastByteAcked
< cwnd
sender sequence number space
rate ~
~
cwnd
RTT
bytes/sec

TCP Slow Start
 when connection
begins, increase rate
exponentially until first
loss event:
 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
Host A
RTT
Host B
time

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)

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
TCP: switching from slow start to
CA

Summary: TCP Congestion
Control
timeout
ssthresh = cwnd/2
cwnd = 1 MSS
dupACKcount = 0
retransmit missing segment
L
cwnd > ssthresh
congestion
avoidance
cwnd = cwnd + MSS (MSS/cwnd)
dupACKcount = 0
transmit new segment(s), as allowed
new ACK
.
dupACKcount++
duplicate ACK
fast
recovery
cwnd = cwnd + MSS
duplicate ACK
ssthresh= cwnd/2
cwnd = ssthresh + 3
dupACKcount == 3
timeout
ssthresh = cwnd/2
cwnd = 1
dupACKcount = 0
ssthresh= cwnd/2
cwnd = ssthresh + 3
dupACKcount == 3
cwnd = ssthresh
dupACKcount = 0
New ACK
slow
start
timeout
ssthresh = cwnd/2
cwnd = 1 MSS
dupACKcount = 0
cwnd = cwnd+MSS
dupACKcount = 0
new ACK
dupACKcount++
duplicate ACK
L
cwnd = 1 MSS
ssthresh = 64 KB
dupACKcount = 0
New
ACK!
New
ACK!
New
ACK!

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
W
W/2
avg TCP thruput =
3
4
W
RTT
bytes/sec

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]:
➜ 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
TCP throughput =
1.22 . MSS
RTT L

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 Fairness
TCP connection 2

Why is TCP fair?
two competing sessions:
 additive increase gives slope of 1, as throughout
increases
 multiplicative decrease decreases throughput
proportionally
R
R
equal bandwidth share
Connection 1 throughput
congestion avoidance: additive increase
loss: decrease window by factor of 2
congestion avoidance: additive increase
loss: decrease window by factor of 2

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

Chapter 3: summary
 principles behind
transport layer services:
 multiplexing,
demultiplexing
 flow control
 instantiation,
implementation in the
Internet
 UDP
 TCP
next:
 leaving the
network “edge”
(application,
transport layers)
 into the network
“core”

Chapter 3 v6.01 (1)

More Related Content

What's hot (20)

Similar to Chapter 3 v6.01 (1) (20)

Recently uploaded (20)

Chapter 3 v6.01 (1)