Congestion Control

1
Congestion Control

• Outline
• Queuing Discipline
• Reacting to Congestion
• Avoiding Congestion

2
Issues

• Two sides of the same coin
• pre-allocate resources in order to avoid
  congestion
• control congestion if (and when) is occurs
• Two points of implementation
• hosts at the edges of the network (transport
  protocol)
• routers inside the network (queuing discipline)
• Underlying service model
• best-effort (assume for now)
• multiple qualities of service (later)

Source
1
10-Mbps Ethernet
Router
Destination
1.5-Mbps T1 link
100-Mbps FDDI
Source
2
3
Framework

• Connectionless flows
• sequence of packets sent between
  source/destination pair
• maintain soft state at the routers
• Taxonomy
• router-centric versus host-centric
• reservation-based versus feedback-based
• window-based versus rate-based

Source
1
Router
Destination
1
Router
Source
2
Router
Destination
2
Source
3
4
Evaluation

• Fairness
• Power (ratio of throughput to delay)

Throughput/delay
Optimal
Load
load
5
Queuing Discipline

• First-In-First-Out (FIFO)
• does not discriminate between traffic sources
• Fair Queuing (FQ)
• explicitly segregates traffic based on flows
• ensures no flow captures more than its share of
  capacity
• variation weighted fair queuing (WFQ)
• Problem?

Flow 1
Flow 2
Round-robin
service
Flow 3
Flow 4
6
FQ Algorithm

• Suppose clock ticks each time a bit is
  transmitted
• Let Pi denote the length of packet i
• Let Si denote the time when start to transmit
  packet i
• Let Fi denote the time when finish transmitting
  packet i
• Fi Si Pi
• When does router start transmitting packet i?
• if before router finished packet i - 1 from this
  flow, then immediately after last bit of i - 1
  (Fi-1)
• if no current packets for this flow, then start
  transmitting when arrives (call this Ai)
• Thus Fi MAX (Fi - 1, Ai) Pi

7
FQ Algorithm (cont)

• For multiple flows
• calculate Fi for each packet that arrives on each
  flow
• treat all Fis as timestamps
• next packet to transmit is one with lowest
  timestamp
• Not perfect cant preempt current packet
• Example

Flow 1
Flow 2
Flow 1
Flow 2
Output
Output
(arriving)
(transmitting)
F 10
F 10
F 8
F 5
F 2
(a)
(b)
8
TCP Congestion Control

• Idea
• assumes best-effort network (FIFO or FQ
  routers)each source determines network capacity
  for itself
• uses implicit feedback
• ACKs pace transmission (self-clocking)
• Challenge
• determining the available capacity in the first
  place
• adjusting to changes in the available capacity

9
Additive Increase/Multiplicative Decrease

• Objective adjust to changes in the available
  capacity
• New state variable per connection
  CongestionWindow
• limits how much data source has in transit
• MaxWin MIN(CongestionWindow,
  AdvertisedWindow)
• EffWin MaxWin - (LastByteSent -
  LastByteAcked)
• Idea
• increase CongestionWindow when congestion goes
  down
• decrease CongestionWindow when congestion goes up

10
AIMD (cont)

• Question how does the source determine whether
  or not the network is congested?
• Answer a timeout occurs
• timeout signals that a packet was lost
• packets are seldom lost due to transmission error
• lost packet implies congestion

11
AIMD (cont)
Source
Destination

• Algorithm
• increment CongestionWindow by one packet per RTT
  (linear increase)
• divide CongestionWindow by two whenever a timeout
  occurs (multiplicative decrease)

• In practice increment a little for each ACK
• Increment (MSS MSS)/CongestionWindow
• CongestionWindow Increment

12
AIMD (cont)

• Trace sawtooth behavior

13
Slow Start
Source
Destination

• Objective determine the available capacity in
  the first
• Idea
• begin with CongestionWindow 1 packet
• double CongestionWindow each RTT (increment by 1
  packet for each ACK)

14
Slow Start (cont)

• Exponential growth, but slower than all at once
• Used
• when first starting connection
• when connection goes dead waiting for timeout
• Trace
• Problem lose up to half a CongestionWindows
  worth of data

70
60
50
KB
40
30
20
10
1.0
2.0
3.0
4.0
5.0
6.0
7.0
8.0
9.0
15
Fast Retransmit and Fast Recovery

• Problem coarse-grain TCP timeouts lead to idle
  periods
• Fast retransmit use duplicate ACKs to trigger
  retransmission

Sender
Receiver
Packet 1
Packet 2
ACK 1
Packet 3
ACK 2
Packet 4
ACK 2
Packet 5
Packet 6
ACK 2
ACK 2
Retransmit
packet 3
ACK 6
16
Results
70
60
50
40
KB
30
20
10
1.0
2.0
3.0
4.0
5.0
6.0
7.0

• Fast recovery
• skip the slow start phase
• go directly to half the last successful
  CongestionWindow (ssthresh)

17
Congestion Avoidance

• TCPs strategy
• control congestion once it happens
• repeatedly increase load in an effort to find the
  point at which congestion occurs, and then back
  off
• Alternative strategy
• predict when congestion is about to happen
• reduce rate before packets start being discarded
• call this congestion avoidance, instead of
  congestion control
• Two possibilities
• router-centric DECbit and RED Gateways
• host-centric TCP Vegas

18
DECbit

• Add binary congestion but to each packet header
• Router
• monitors average queue length over last busyidle
  cycle
• set congestion bit if average queue length gt 1
• attempts to balance throughout against delay

Queue length
Current
time
T
ime
Current
Previous
cycle
cycle
A
veraging
interval
19
End Hosts

• Destination echoes bit back to source
• Source records how many packets resulted in set
  bit
• If less than 50 of last windows worth had bit
  set
• increase CongestionWindow by 1 packet
• If 50 or more of last windows worth had bit set
  
• decrease CongestionWindow by 0.875 times

20
Random Early Detection (RED)

• Notification is implicit
• just drop the packet (TCP will timeout)
• could make explicit by marking the packet
• Early random drop
• rather than wait for queue to become full, drop
  each arriving packet with some drop probability
  whenever the queue length exceeds some drop level

21
RED Details

• Compute average queue length
• AvgLen (1 - Weight) AvgLen
• Weight SampleLen
• 0 lt Weight lt 1 (usually 0.002)
• SampleLen is queue length each time a packet
  arrives

MaxThreshold
MinThreshold
A
vgLen
22
RED Details (cont)

• Two queue length thresholds
• if AvgLen lt MinThreshold then
• enqueue the packet
• if MinThreshold lt AvgLen lt MaxThreshold then
• calculate probability P
• drop arriving packet with probability P
• if ManThreshold lt AvgLen then
• drop arriving packet

23
RED Details (cont)

• Computing probability P
• TempP MaxP (AvgLen - MinThreshold)/
  (MaxThreshold - MinThreshold)
• P TempP/(1 - count TempP)
• Drop Probability Curve

P(drop)
1.0
MaxP
A
vgLen
MinThresh
MaxThresh
24
Tuning RED

• Probability of dropping a particular flows
  packet(s) is roughly proportional to the share of
  the bandwidth that flow is currently getting
• MaxP is typically set to 0.02, meaning that when
  the average queue size is halfway between the two
  thresholds, the gateway drops roughly one out of
  50 packets.
• If traffic id bursty, then MinThreshold should be
  sufficiently large to allow link utilization to
  be maintained at an acceptably high level
• Difference between two thresholds should be
  larger than the typical increase in the
  calculated average queue length in one RTT
  setting MaxThreshold to twice MinThreshold is
  reasonable for traffic on todays Internet
• Penalty Box for Offenders

25
TCP Vegas

• Idea source watches for some sign that routers
  queue is building up and congestion will happen
  too e.g.,
• RTT grows
• sending
• rate flattens

70
60
50
40
KB
30
20
10
0.5
1.0
1.5
4.0
4.5
6.5
8.0
2.0
2.5
3.0
3.5
5.0
5.5
6.0
7.0
7.5
8.5
Time (seconds)
1100
900
700
Sending KBps
500
300
100
0.5
1.0
1.5
4.0
4.5
6.5
8.0
2.0
2.5
3.0
3.5
5.0
5.5
6.0
7.0
7.5
8.5
Time (seconds)
10
Queue size in router
5
0.5
1.0
1.5
4.0
4.5
6.5
8.0
2.0
2.5
3.0
3.5
5.0
5.5
6.0
7.0
7.5
8.5
Time (seconds)
26
Algorithm

• Let BaseRTT be the minimum of all measured RTTs
  (commonly the RTT of the first packet)
• If not overflowing the connection, then
• ExpectRate CongestionWindow/BaseRTT
• Source calculates sending rate (ActualRate) once
  per RTT
• Source compares ActualRate with ExpectRate
• Diff ExpectedRate - ActualRate
• if Diff lt a
• increase CongestionWindow linearly
• else if Diff gt b
• decrease CongestionWindow linearly
• else
• leave CongestionWindow unchanged

27
Algorithm (cont)

• Parameters
• a 1 packet
• b 3 packets
• Even faster retransmit
• keep fine-grained timestamps for each packet
• check for timeout on first duplicate ACK

70
60
50
40
KB
30
20
10
0.5
1.0
1.5
2.0
2.5
3.0
3.5
4.0
4.5
5.0
5.5
6.0
6.5
7.0
7.5
8.0
T
ime (seconds)
240
200
160
CAM KBps
120
80
40
0.5
1.0
1.5
2.0
2.5
3.0
3.5
4.0
4.5
5.0
5.5
6.0
6.5
7.0
7.5
8.0
T
ime (seconds)
28
Quality of Service

• Outline
• Realtime Applications
• Integrated Services
• Differentiated Services

29
Realtime Applications

• Require deliver on time assurances
• must come from inside the network
• Example application (audio)
• sample voice once every 125us
• each sample has a playback time
• packets experience variable delay in network
• add constant factor to playback time playback
  point

Sampler
,
Microphone
Buffer
,
A D
D A
converter
Speaker
30
Playback Buffer
Packet
arrival
Packet
generation
Playback
Sequence number
Buffer
Network
delay
T
ime
31
Example Distribution of Delays
90
97
98
99
3
2
Packets ()
1
150
200
100
50
Delay (milliseconds)
32
Taxonomy
Applications
Real time
Elastic
Interactive
Asynchronous
Interactive
T
olerant
Intolerant
bulk
Adaptive
Nonadaptive
Rate-adaptive
Nonadaptive
Delay-
Rate-
adaptive
adaptive
33
Integrated Services

• Service Classes
• guaranteed
• controlled-load
• Mechanisms
• signalling protocol
• admission control
• policing
• packet scheduling

34
Flowspec

• Rspec describes service requested from network
• controlled-load none
• guaranteed delay target
• Tspec describes flows traffic characteristics
• average bandwidth burstiness token bucket
  filter
• token rate r
• bucket depth B
• must have a token to send a byte
• must have n tokens to send n bytes
• start with no tokens
• accumulate tokens at rate of r per second
• can accumulate no more than B tokens

35
Per-Router Mechanisms

• Admission Control
• decide if a new flow can be supported
• answer depends on service class
• not the same as policing
• Packet Processing
• classification associate each packet with the
  appropriate reservation
• scheduling manage queues so each packet receives
  the requested service

36
Reservation Protocol

• Called signaling in ATM
• Proposed Internet standard RSVP
• Consistent with robustness of todays
  connectionless model
• Uses soft state (refresh periodically)
• Designed to support multicast
• Receiver-oriented
• Two messages PATH and RESV
• Source transmits PATH messages every 30 seconds
• Destination responds with RESV message
• Merge requirements in case of multicast
• Can specify number of speakers

37
RSVP Example
38
RSVP versus ATM (Q.2931)

• RSVP
• receiver generates reservation
• soft state (refresh/timeout)
• separate from route establishment
• QoS can change dynamically
• receiver heterogeneity
• ATM
• sender generates connection request
• hard state (explicit delete)
• concurrent with route establishment
• QoS is static for life of connection
• uniform QoS to all receivers

39
Differentiated Services

• Problem with IntServ scalability
• Idea support two classes of packets
• premium
• best-effort

P(drop)

• Mechanisms
• packets in and out bit
• edge routers tag packets
• core routers RIO (RED with In and Out)

1.0
MaxP
A
vgLen
Min
Max
Max
Min
in
in
out
out