Improving Adaptability and Fairness
in Internet Congestion Control
May 30, 2001
Seungwan Ryu
PhD Student of IE Department
University at Buffalo
I. Internet Congestion Control




Internet Congestion Control
Mathematical Modeling and Analysis
Adaptive AQM and User Response
Future Study Plan
2
I. Internet Congestion Control






What is Congestion ?
Congestion Control and Avoidance
Implicit vs. Explicit feedback
TCP Congestion Control
Active Queue management (AQM)
Explicit Congestion Notification (ECN)
3
What is congestion ?

What is congestion ?


The aggregate demand for bandwidth exceeds the
available capacity of a link.
What will be occur ?

Performance Degradation
•
•
•
•
Multiple packet loss
Low link utilization (low Throughput)
High queueing delay
Congestion collapse
4
Congestion Control and Avoidance

Two approaches for handling Congestion

Congestion Control (Reactive)
• Play after the network is overloaded

Congestion Avoidance (Proactive)
• Play before the network becomes overloaded
5
Implicit vs. Explicit feedback

Implicit feedback Congestion Control


Network drops packets when congestion occur
Source infer congestion implicitly
• time-out, duplicated ACKs, etc.


Example: end-to-end TCP congestion Control
Simple to implement but inaccurate
• implemented only at Transport layer (e.g., TCP)
6
Implicit vs. Explicit feedback - 2

Explicit feedback Congestion Control

Network component (e.g., router) Provides
congestion indication explicitly to sources




use packet marking, or RM cells (in ATM ABR control)
Examples: DECbit, ECN, ATM ABR CC, etc.
Provide more accurate information to sources
But is more complicate to implement


Need to change both source and network algorithm
Need cooperation between sources and network
component
7
TCP Congestion Control

Use end-to-end congestion control

use implicit feedback
• e.g., time-out, triple duplicated ACKs, etc.

use window based flow control
• cwnd = min (pipe size, rwnd)
• self-clocking
• slow-start and congestion avoidance

Examples:
• TCP Tahoe, TCP Reno, TCP Vegas, etc.
8
TCP Congestion Control - 2

cwnd
Slow-start and Congestion Avoidance
Slow Start
Congestion Avoidance
W
4
W+1
2
1
RTT
RTT
Time
9
Active Queue Management (AQM) - 1

Performance Degradation in current TCP
Congestion Control




Multiple packet loss
Low link utilization
Congestion collapse
The role of the router (i.e., network)


Control congestion effectively with a network
Allocate bandwidth fairly
10
AQM - 2

Problems with current router algorithm



Use FIFO based tail-drop (TD) queue management
Two drawbacks with TD: lock-out, full-queue
Possible solution: AQM



Drop packets before buffer becomes full
Examples: RED, BLUE, ARED, SRED, FRED,….
Use (exponentially weighted) average queue length
as an congestion indicator
11
AQM - 3

Random Early Detection (RED)


use network algorithm to detect incipient
congestion
Design goals:
•
•
•
•

minimize packet loss and queueing delay
avoid global synchronization
maintain high link utilization
removing bias against bursty source
Achieve goals by
• randomized packet drop
• queue length averaging
12
RED
P
1
maxp
minth
maxth
K
13
Active Queue Management (AQM) - 4

Problems with existing AQM Proposals




Mismatch between macroscopic and microscopic
behavior of queue length
Insensitivity to the change of input traffic load
Configuration (parameter setting) problem
Reasons:



Queue length averaging
use inappropriate congestion indicator
Use inappropriate control function
14
Explicit Congestion Notification (ECN)

Current congestion indication



Use packet drop to indicate congestion
source infer congestion implicitly
ECN




to give less packet drop and better performance
use packet marking rather than drop
need cooperation between sources and network
need two bits in IP header: ECT-bit, CE-bit
15
ECN - 2
ECT
IP Header
CE
1
ECT
0
CE
1
1
1
TCP Header
0
0
CWR
CWR
2
1
ACK TCP
Header
ECN-Echo
TCP Header
3
1
CWR
Source
4
Router
Destination
16
Contents




Internet Congestion Control
Mathematical Modeling and Analysis
Adaptive AQM and User Response
Future Study Plan
17
II. Mathematical Modeling and
Analysis

An Overview

Mathematical Modeling of AQM



Window based packet switching and the Internet
Mathematical modeling and analysis of AQM
Problems with existing AQMs


Problems with existing AQMs
Adaptive congestion indicator and control function
18
Overview

Goal of mathematical modeling




see system dynamics (in steady state)
capture main factors influence to performance
provide design and/or operational
recommendations
Two approaches

Modeling steady state TCP behaviors
• the square root law, PFTK
• assume TD queue management at the router

Mathematical modeling and analysis of AQM (RED)
19
Overview - 2

AQM modeling and analysis




Analytic modeling and analysis
Control Theoretic Analysis
Window based modeling and Analysis
Assumptions



Poisson assumption for input traffic
Fixed number of persistent TCP traffics
Steady state window size saturation
20
Mathematical Modeling of AQM

Window based packet switching Model (Yang 99)

If link j is not congested
nsj  0, Q j  0 s  S ( j ), j
C j  s  s

j
If link j is congested
nsj  0, Q j  0 s  S ( j ), j
C j  s  s
j
21
Mathematical Modeling of AQM - 2

Window size of an individual connection

Since
nsj
Qj

s
Cj
Q j  0
     (1)


Qj

Ws  s Rs   jJ ( s ) nsj   s  j
 Rs  s  S ( j )


C
j



Limitation of this model

Assume infinite buffer size
• No buffer overflow
• No packet drop
• No queue management algorithm at routers
22
Mathematical Modeling of AQM - 3
A simple AQM model
Sources
s1
1
S2
AQM Router
2
Bottleneck
Link
Destination
C
S
K
Min_th
SS
23
Mathematical Modeling of AQM - 4

Extend Yang’s Model to AQM model



Finite buffer capacity K
The router use AQM to control congestion
When congested
• Our Model:
• Yang’s Model:

s
s
 C , s  s (1  pd )

s
s
 C,
s  s
24
Mathematical Modeling of AQM - 5

Case 1: Tail drop

We obtain two relationship
nsj
Qj
W


s
Cj
Q j  0 ,
 s s ( Rs 

s
s  C
      ( 2)
Qs
 Q

)  R 
C
C
Finally, packet drop probability Pd:
W

1 
pd    ( R  Q )
C

0

if
 C
if
 C
25
Mathematical Modeling of AQM - 6

Case 2: AQM



Q  Q  min th  s ns
Let
Then
Q
W   (1  pd )( R  )
C
Packet drop prob. Pd:
W

1


pd    ( R  Q )
C

0

if
  C , Q  min th
O.w.
26
Mathematical Modeling of AQM - 7

Congestion Indicator
 Input traffic load  should be the congestion

Indicator
Current AQMs
• Use queue length Q as an alternative
• Assume that the input traffic load  is fixed in
equilibrium

Reason
• can not measure(or estimate)  exactly for on line
implementation of packet drop function
27
Mathematical Modeling of AQM - 8

Packet drop function
p d  f ( )

Reason
• The traffic load

fluctuate, NOT stay in equilibrium
• queue length is a function of input traffic

Alternatively:
pd  f ( , Q)
28
Problems with existing AQMs

Mismatch between macroscopic and
microscopic behavior of queue length

Insensitivity to the input traffic load variation

parameter configuration problem
29
Problems with existing AQMs - 2

Mismatch problem
Internet Traffic Generation
40
35
25
20
15
10
5
time
31
28
25
22
19
16
13
10
7
4
0
1
Window size
30
30
Problems with existing AQMs - 3
Mismatch between macroscopic and
microscopic behavior of queue length
Rho
Actual
Wq=0.02
Wq=0.1
25
20
Queue Length

15
10
5
0
1
6
11
16
21
26
31
Time
31
Problems with existing AQMs - 4

Insensitivity to the input traffic load variation

With light traffic (i.e.,   0.3,   0.5 )
32
Problems with existing AQMs - 5

Insensitivity to the input traffic load variation

With medium traffic (i.e.,   0.7,   0.9 )
33
Problems with existing AQMs - 6

Insensitivity to the input traffic load variation

With heavy traffic (i.e.,   1.1,   1.4 )
34
Problems with existing AQMs - 7

Parameter configuration problem


Has been a main design issue since 1993
many modified AQMs has been proposed
• Verified with simple simulation or simple experiment
• good for particular traffic conditions
• Real traffic is totally different.

Need adaptive congestion indicator and control
function
• Adaptive to input traffic load variation
• Avoid congestion NOT based on current state (i,e,. Q)
35
Contents




Internet Congestion Control
Mathematical Modeling and Analysis
Adaptive AQM and User Response
Future Study Plan
36
III. Adaptive AQM and User Response




Input traffic load Prediction
Adaptive AQM algorithms
Adaptive parameter configuration
Adaptive User response algorithm
37
Input traffic load Prediction

Consider time-slotted model



Time is divided into unit time slots, t, t=0,1,…
calculate parameters at the end of each slot
estimate Qt+1 to detect congestion proactively
Qt 1  (t 1  C )  Qt
• Predict ˆt  1 from measured input traffic t-1, t of past
two time slots
• Then, predict Q̂t  1 of next time slot t
38
Adaptive AQM algorithms

Algorithm I: E-RED and E-GRED

Enhanced-RED

0,

Q̂  minth

p   max p t  1
maxth  minth


1

Q̂t  1  minth
minth  Q̂t  1  maxth
maxth  Q̂t  1
E-GRED: similar to E-RED
39
Adaptive AQM algorithms - 2

Algorithm II:


Use both predicted traffic intensity
buffer utilization t=Qt/K
Possible algorithms:
 1 ˆ t  1t ,

 2 ˆ t2 1t ,
ˆ tand
current
1
 3 2 ˆ t
t 1
Example:
• If t is low and ˆ t  1 is high: more penalty to incoming packets
• If t is high and ˆ t  1 is low: more penalty on existing packets
• Only High penalty for both packets when t and ˆ t  1 are high
40
Adaptive AQM algorithms - 3

Algorithm III: E-BLUE

BLUE Algorithm
• uses packet drops and link idle for adjusting packet drop
probability
• Can not avoid some degree of performance degradation

Enhancement
• Use Virtual lower/upper bound (VL, VU)
• Combine predicted queue length Q̂t  1 with BLUE
• Impose penalty according to the traffic situation ( Qt, Q̂t 1)
41
Adaptive AQM algorithms - 4

E-BLUE


If max(Q̂t  1 ,0 )  0, then pd = pd- 
Else if VL < Q̂t  1<VU,
 pd ( Q̂t  1 Qt )
pd  
 pd ( Qt Q̂t  1 )
for
arriving
for
existing
• Else (Q̂t  1>VU)
1

pd  
min pd ( Qt Q̂t  1 ),1


for
for
arriving
existing
• pd=pd+
42
Adaptive parameter configuration

Adaptive queue length sampling interval t


Previous recommendations
• In [22], minimum RTT was recommended
• In [65], static and link speed independent value was
recommended
• However, models of [22, 65] were assumed to have
persistent fixed N TCP traffics
Our recommendation
• The amount of incoming traffic fluctuate with time
• Adjust t according to the varying traffic situation
(i.e., adjust t according to the amount of input traffic)
43
Adaptive parameter configuration - 2
Q
(i-1)
i
(i+1)
(i+2)
Time
44
Adaptive parameter configuration - 3

Adaptive filtering weight wq


In RED, wq was recommended with 0.02 for long-term
(macroscopic) performance goal
Fixed small value of wq shows problems
• Parameter setting problem
• Insensitivity of control function to the change of traffic
• Fairness problem: impose penalty to innocent packets


Need to have adaptive wq to the change of traffic load
One possible method:
• Set wq as a function of current queue utilization,
•
e.g., wq =  Qt/C , 0 <  < 1
45
Adaptive User response algorithm


AQM need work with intelligent source
response for better performance
Enhanced-ECN

If receive ECN feedback in (t-1)
• If No ECN feedback in t

If received ACK > 0

Else
W  W  M /W  M
W  W  M /W
• Else, Continue usual response to ECN feedback

Else, Continue TCP Congestion Avoidance
46
Contents




Internet Congestion Control
Mathematical Modeling and Analysis
Adaptive AQM and User Response
Future Study Plan
47
IV. Future Study Plan


Future Study plan: a schedule
Mathematical Modeling and Analysis




Stability and Control Dynamics
Alternative Modeling
Control Theoretic Consideration
Simulation plan


Traffics
Performance Metrics
48
Future Study plan: a schedule




Documentation:
Mathematical Modeling and Analysis
Simulation plan
Performance Metrics
49
Mathematical Modeling and Analysis

Since p=f(,q) ,

Then find equilibrium point (*,p*)
p
=g(p)
T ( , q)     (1  p) 
C (1  p)
R p
P=f()
(*,p*)

50
Mathematical Modeling and Analysis - 2
Alternative Modeling:


State dependent service M/M/1 queueing model


0
1
C




L-1
C
L
C


L-1
(C+p1)

K-1
(C+pK’-1)
K
C+
L=minth, K’=K-minth
51
Mathematical Modeling and Analysis - 3

Service rates

Steady state probabilities
Qi  min th
 C,

S  C  pi, min th  Qi  K
 C  ,
K  Qi




( ) i 0 ,
i  min th

C

 i  min th
  min th


i  
0 ,
min th  i  K

 j1
 ( C )
(

C

p

)

i 

1
min th 
K


  min th

  ( )i   ij1min th
i0
 ,
 ( C )
(

C

p

)
 i 1 C
i  min th 1 
i 

52
Mathematical Modeling and Analysis - 3

Control Theoretic Consideration
Control
Function
S
t
Queue
dynamics
Router
Buffer
t(1-p)
D
ACK (or NACK)
53
Simulation plan

Goal of simulation study



See dynamics and performance of our AQM
Compare results with other AQM such as RED
Use realistic traffic


previous studies has been done with simple and
unreal traffic (fixed number of persistent TCPs)
Generate realistic Internet traffic
• Long-lived (FTP) and short-lived (web-like) TCP traffic
• UDP traffic: CBR and/or ON/OFF
54
Performance Metrics

TCP traffics

Network-centric: for aggregate traffic
• Throughput (or goodput)
• Packet dropping (marking) probability
• Link utilization (or queueing delay)

User-centric: for Individual traffic
• goodput (or throughput)
• mean response time (RTT)

UDP traffic
• individual packet drop probability and its distribution
55