A Study of Active Queue Management for Congestion Control
-
Upload
cade-mcfadden -
Category
Documents
-
view
14 -
download
0
description
Transcript of A Study of Active Queue Management for Congestion Control
![Page 1: A Study of Active Queue Management for Congestion Control](https://reader030.fdocuments.us/reader030/viewer/2022032605/56812c1f550346895d9084d7/html5/thumbnails/1.jpg)
AQM for Congestion Control 1
A Study of Active Queue Management for Congestion
Control
Victor FiroiuMarty Borden
![Page 2: A Study of Active Queue Management for Congestion Control](https://reader030.fdocuments.us/reader030/viewer/2022032605/56812c1f550346895d9084d7/html5/thumbnails/2.jpg)
AQM for Congestion Control 2
Outline
• Introduction
• Feedback Control System Background
• FCS applied to AQM
• Calculating FCS equations
• Simulation verifications
• RED configuration recommendations
• Conclusion
![Page 3: A Study of Active Queue Management for Congestion Control](https://reader030.fdocuments.us/reader030/viewer/2022032605/56812c1f550346895d9084d7/html5/thumbnails/3.jpg)
AQM for Congestion Control 3
Introduction
• Goal - Determine “best” RED configuration using systematic approach
• Models - queue vs. feedback control system• Mathematical analysis and fundamental
Laws • Simulation verification of model• Recommendations• Future directions
![Page 4: A Study of Active Queue Management for Congestion Control](https://reader030.fdocuments.us/reader030/viewer/2022032605/56812c1f550346895d9084d7/html5/thumbnails/4.jpg)
AQM for Congestion Control 4
Feedback Control systems
• What is it? – Model where a change in input causes system variables to conform to desired values called the reference
• Why this model ? - Can create a stable and efficient system
• Two basic models - Open vs. Closed loop
![Page 5: A Study of Active Queue Management for Congestion Control](https://reader030.fdocuments.us/reader030/viewer/2022032605/56812c1f550346895d9084d7/html5/thumbnails/5.jpg)
AQM for Congestion Control 5
Feedback Control (closed loop)
Actuator
Monitor
reference
control input
controlled variable
manipulatedvariable
Controlled System
+ -
error
controlfunction
Controller
sample
![Page 6: A Study of Active Queue Management for Congestion Control](https://reader030.fdocuments.us/reader030/viewer/2022032605/56812c1f550346895d9084d7/html5/thumbnails/6.jpg)
AQM for Congestion Control 6
How to apply FCS to AQM
• Try to get two equations to derive steady state behavior – in our case queue function (avg. length of queue) and control function (dependent upon architecture –RED)
Control theory stability
• Networks as a feedback system
• Distributed & delayed feedback
![Page 7: A Study of Active Queue Management for Congestion Control](https://reader030.fdocuments.us/reader030/viewer/2022032605/56812c1f550346895d9084d7/html5/thumbnails/7.jpg)
AQM for Congestion Control 7
Model TCP Avg. Queue Size
![Page 8: A Study of Active Queue Management for Congestion Control](https://reader030.fdocuments.us/reader030/viewer/2022032605/56812c1f550346895d9084d7/html5/thumbnails/8.jpg)
AQM for Congestion Control 8
Single flow feedback system
rt,i(p,Ri) = T(p,Ri)
Becomes
rt,i(p,R) ≤ c/n, 1 ≤ i ≤ n
![Page 9: A Study of Active Queue Management for Congestion Control](https://reader030.fdocuments.us/reader030/viewer/2022032605/56812c1f550346895d9084d7/html5/thumbnails/9.jpg)
AQM for Congestion Control 9
Finding the Queue “Law”
![Page 10: A Study of Active Queue Management for Congestion Control](https://reader030.fdocuments.us/reader030/viewer/2022032605/56812c1f550346895d9084d7/html5/thumbnails/10.jpg)
AQM for Congestion Control 10
Non Feedback Queue “Law”
R = R0 + q/c
p0 = T-1p (c/n, R0)
q(p) = { max (B,c (T-1R (p,c/n) - R0)), p ≤ p0
Else 0
u(p) = { 1, p ≤ p0 Else T(p, R0) /(c/n)
![Page 11: A Study of Active Queue Management for Congestion Control](https://reader030.fdocuments.us/reader030/viewer/2022032605/56812c1f550346895d9084d7/html5/thumbnails/11.jpg)
AQM for Congestion Control 11
Verification through simulation
• Using NS run multiple simulations varying link capacity, number of flows, and drop probability p
• Flows are “infinite” FTP sessions with fixed RTT
• Buffer is large enough to prevent packet loss due to overflow
• Graph mathematically predicted average queue size vs. simulation (and do the same with link utilization)
![Page 12: A Study of Active Queue Management for Congestion Control](https://reader030.fdocuments.us/reader030/viewer/2022032605/56812c1f550346895d9084d7/html5/thumbnails/12.jpg)
AQM for Congestion Control 12
One Sample Result
![Page 13: A Study of Active Queue Management for Congestion Control](https://reader030.fdocuments.us/reader030/viewer/2022032605/56812c1f550346895d9084d7/html5/thumbnails/13.jpg)
AQM for Congestion Control 13
Add in Feedback
![Page 14: A Study of Active Queue Management for Congestion Control](https://reader030.fdocuments.us/reader030/viewer/2022032605/56812c1f550346895d9084d7/html5/thumbnails/14.jpg)
AQM for Congestion Control 14
Feedback Control system Equilibrium point
![Page 15: A Study of Active Queue Management for Congestion Control](https://reader030.fdocuments.us/reader030/viewer/2022032605/56812c1f550346895d9084d7/html5/thumbnails/15.jpg)
AQM for Congestion Control 15
RED as a Control Function
![Page 16: A Study of Active Queue Management for Congestion Control](https://reader030.fdocuments.us/reader030/viewer/2022032605/56812c1f550346895d9084d7/html5/thumbnails/16.jpg)
AQM for Congestion Control 16
Simulation with G(p) and H(q)
![Page 17: A Study of Active Queue Management for Congestion Control](https://reader030.fdocuments.us/reader030/viewer/2022032605/56812c1f550346895d9084d7/html5/thumbnails/17.jpg)
AQM for Congestion Control 17
RED convergence point
![Page 18: A Study of Active Queue Management for Congestion Control](https://reader030.fdocuments.us/reader030/viewer/2022032605/56812c1f550346895d9084d7/html5/thumbnails/18.jpg)
AQM for Congestion Control 18
Stable system results
![Page 19: A Study of Active Queue Management for Congestion Control](https://reader030.fdocuments.us/reader030/viewer/2022032605/56812c1f550346895d9084d7/html5/thumbnails/19.jpg)
AQM for Congestion Control 19
Unstable results
![Page 20: A Study of Active Queue Management for Congestion Control](https://reader030.fdocuments.us/reader030/viewer/2022032605/56812c1f550346895d9084d7/html5/thumbnails/20.jpg)
AQM for Congestion Control 20
Unstable results part 2
![Page 21: A Study of Active Queue Management for Congestion Control](https://reader030.fdocuments.us/reader030/viewer/2022032605/56812c1f550346895d9084d7/html5/thumbnails/21.jpg)
AQM for Congestion Control 21
RED configuration Recommendations
• drop-conservative policy: low p, high q
• delay-conservative policy: low q, high p
• Need to estimate:1. Line speed c
2. Min and Max throughput per flow τ or number of flows n
3. Min and Max packet size M
4. Min and Max RRT R0
![Page 22: A Study of Active Queue Management for Congestion Control](https://reader030.fdocuments.us/reader030/viewer/2022032605/56812c1f550346895d9084d7/html5/thumbnails/22.jpg)
AQM for Congestion Control 22
Sample Control Law policy
![Page 23: A Study of Active Queue Management for Congestion Control](https://reader030.fdocuments.us/reader030/viewer/2022032605/56812c1f550346895d9084d7/html5/thumbnails/23.jpg)
AQM for Congestion Control 23
Range of Queue Laws
![Page 24: A Study of Active Queue Management for Congestion Control](https://reader030.fdocuments.us/reader030/viewer/2022032605/56812c1f550346895d9084d7/html5/thumbnails/24.jpg)
AQM for Congestion Control 24
Configuring Estimator of average queue Size
Consists of :
• Queue averaging algorithms
• Averaging interval
• Sampling the queue size
![Page 25: A Study of Active Queue Management for Congestion Control](https://reader030.fdocuments.us/reader030/viewer/2022032605/56812c1f550346895d9084d7/html5/thumbnails/25.jpg)
AQM for Congestion Control 25
Queue Averaging Algorithm
• Low- pass filter on current queue size • Moving average to filter out bursts• Exponential weighting decreasing with age• Estimate is computed over samples from the
previous I time period – recommendations for I to follow
Average weight = w = 1- aδ/I
![Page 26: A Study of Active Queue Management for Congestion Control](https://reader030.fdocuments.us/reader030/viewer/2022032605/56812c1f550346895d9084d7/html5/thumbnails/26.jpg)
AQM for Congestion Control 26
Averaging Interval I
• Should provide good estimate of long term average assuming number of flows is constant
• Should adapt as fast as possible to change in traffic conditions
![Page 27: A Study of Active Queue Management for Congestion Control](https://reader030.fdocuments.us/reader030/viewer/2022032605/56812c1f550346895d9084d7/html5/thumbnails/27.jpg)
AQM for Congestion Control 27
I = P is recommended
![Page 28: A Study of Active Queue Management for Congestion Control](https://reader030.fdocuments.us/reader030/viewer/2022032605/56812c1f550346895d9084d7/html5/thumbnails/28.jpg)
AQM for Congestion Control 28
Sampling the Queue size
• Queue size acts like a step function
• Changes every RTT with adjustments made from information received
• “Ideal” sampling rate is once every RTT
• Recommend sampling = minimum RRT
![Page 29: A Study of Active Queue Management for Congestion Control](https://reader030.fdocuments.us/reader030/viewer/2022032605/56812c1f550346895d9084d7/html5/thumbnails/29.jpg)
AQM for Congestion Control 29
Conclusions
• Feedback control model validated through simulations
• Found instability points and recommended settings to avoid them
• Also developed recommended RED queue size estimator settings
• Many issues still to look at in future
![Page 30: A Study of Active Queue Management for Congestion Control](https://reader030.fdocuments.us/reader030/viewer/2022032605/56812c1f550346895d9084d7/html5/thumbnails/30.jpg)
AQM for Congestion Control 30
Thoughts
• Nice idea using model from a different discipline to analyze networks
• Good simulations to validate predicted data
• Many assumptions made to make math and model work which may make it invalid
• Limited traffic patterns and type of traffic also make the model’s value suspect
![Page 31: A Study of Active Queue Management for Congestion Control](https://reader030.fdocuments.us/reader030/viewer/2022032605/56812c1f550346895d9084d7/html5/thumbnails/31.jpg)
AQM for Congestion Control 31
Questions?