Computer and

Communication Networks
EE-357
By Dr Mir Yasir Umair
Associate Professor, MCS, NUST
Transport Layer
Congestion Control
 Chapter 3 outline
3.1 transport-layer services 3.5 connection-oriented
3.2 multiplexing and demultiplexing transport: TCP
▪ segment structure
3.3 connectionless transport: UDP ▪ reliable data transfer
3.4 principles of reliable data transfer ▪ flow control
▪ connection management

3.6 principles of congestion

control
3.7 TCP congestion control

Transport Layer 3-3

 Principles of congestion control
• Congestion: “too many sources sending too much data too fast
for network to handle”
• Sender can sense congestion through receipt of 3 duplicate acks or
timeout events.
• Congestion Window cwnd is dynamic, function of perceived
network congestion
o Max window size = min (cwnd, rwnd)
• Challenge
o Determine available capacity in the first place (slow start)
o Adjust to changes in the available capacity (AIMD)
 Principles of congestion control
congestion:
• informally: “too many sources sending too
much data too fast for network to handle”
• different from flow control!
• manifestations:
o lost packets (buffer overflow at routers)
o long delays (queueing in router buffers)
• a top-10 problem!

Transport Layer 3-5

 Approaches towards congestion control
two broad approaches towards congestion control:

end-end congestion network-assisted

control: congestion control:
❖ no explicit ❖routers provide
feedback from feedback to end
network
systems
❖ congestion inferred ▪ single bit indicating
from end-system congestion (SNA, DECbit,
TCP/IP ECN, ATM)
observed loss, delay
▪ explicit rate for sender to
❖ approach taken by send at
TCP

Transport Layer 3-6

 Chapter 3 outline
3.1 transport-layer services 3.5 connection-oriented
3.2 multiplexing and demultiplexing transport: TCP
▪ segment structure
3.3 connectionless transport: UDP ▪ reliable data transfer
3.4 principles of reliable data transfer ▪ flow control
▪ connection management

3.6 principles of congestion

control
3.7 TCP congestion control

Transport Layer 3-7

TCP congestion control: additive increase multiplicative
decrease
❖ approach: sender increases transmission rate (window
size), probing for usable bandwidth, until loss occurs
▪ additive increase: increase cwnd by 1 MSS every
RTT until loss detected
▪ multiplicative decrease: cut cwnd in half after loss
additively increase window size …
…. until loss occurs (then cut window in half)

congestion window size

cwnd: TCP sender
AIMD saw tooth
behavior: probing
for bandwidth

time
Transport Layer 3-8
 TCP Congestion Control: details
sender sequence number space
cwnd TCP sending rate:
❖ roughly: send cwnd
bytes, wait RTT for
last byte last byte ACKS, then send
ACKed sent, not-
yet ACKed
sent
more bytes
(“in-
flight”) cwnd
❖ sender limits transmission: ~
rate~ bytes/sec
RTT
LastByteSent- < cwnd
LastByteAcked
❖ cwnd is dynamic, function of
perceived network
congestion

Transport Layer 3-9

 TCP Slow Start
Host A Host B
❖ when connection
begins, increase rate
exponentially until first

RTT
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
time

Transport Layer 3-10

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

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

Transport Layer 3-11

 Fast Recovery
• Start with slow start until any loss occurs
• Then go back directly to half the last successful Cong Window
and continue with AIMD
 TCP Flavours
TCP Tahoe
• Uses SS + AIMD + Fast retransmit
• For Loss indicated by
o timeout: set cwind to 1
o 3 duplicate ACKS: set cwind to 1
• Problem: Window size is set to minimum value after packet loss

TCP Reno:
• Uses SS + AIMD + Fast retransmit + Fast Recovery
• Loss indicated by
o timeout: set cwind to 1
o 3 duplicate ACKS: set cwind to 50% since network is
capable of sending packets
 TCP Tahoe
If congestion window is
• below threshold: use SS
• above threshold: Use AIMD

If loss occurs, set

• Threshold = half of the
current congestion window
• congestion window = 1.
 TCP Reno
If congestion window is
• below threshold: use SS
• above threshold: Use
AIMD

If timeout occurs, same as

TCP Tahoe

For dup acks, Threshold =

new congestion window =
half of the current
congestion window
 Example – TCP Reno
• When SS is in use?
1-6, 23-26
• When AIMD is in use?
6-23
• Identify type of segment
loss at 16th round ?
(timeout, 3 dup ACKs) –
ans = dupAck
• Value of ssthreshold at 1st
trx round?
32
• Value of ssthreshold at
18th trx round?
21
 Summary: TCP Congestion Control
New
New ACK!
ACK! new ACK
duplicate ACK
dupACKcount++ new ACK
.
cwnd = cwnd + MSS (MSS/cwnd)
dupACKcount = 0
cwnd = cwnd+MSS transmit new segment(s), as allowed
dupACKcount = 0
L transmit new segment(s), as allowed
cwnd = 1 MSS
ssthresh = 64 KB cwnd > ssthresh
dupACKcount = 0 slow L congestion
start timeout avoidance
ssthresh = cwnd/2
cwnd = 1 MSS duplicate ACK
timeout dupACKcount = 0 dupACKcount++
ssthresh = cwnd/2 retransmit missing segment
cwnd = 1 MSS
dupACKcount = 0
retransmit missing segment
timeout
New
ACK!
ssthresh = cwnd/2
cwnd = 1 New ACK
dupACKcount = 0
cwnd = ssthresh dupACKcount == 3
dupACKcount == 3 retransmit missing segment dupACKcount = 0
ssthresh= cwnd/2 ssthresh= cwnd/2
cwnd = ssthresh + 3 cwnd = ssthresh + 3
retransmit missing segment retransmit missing segment
fast
recovery
duplicate ACK
cwnd = cwnd + MSS
transmit new segment(s), as allowed

Transport Layer 3-17

Thank you