Accuracy of Link Capacity Estimates using Passive and Active Approaches with CapProbe
description
Transcript of Accuracy of Link Capacity Estimates using Passive and Active Approaches with CapProbe
Accuracy of Link Capacity Estimates Accuracy of Link Capacity Estimates using using
Passive and Active Approaches with Passive and Active Approaches with CapProbeCapProbe
Rohit Kapoor, Ling-Jyh Chen, M. Y. Sanadidi, Mario GerlaRohit Kapoor, Ling-Jyh Chen, M. Y. Sanadidi, Mario GerlaDept. of Computer Science, University of California at Los AngelesDept. of Computer Science, University of California at Los Angeles
June 2004 ISCC 2004 2
Packet pair Techniques
TbTaTb
Ideal Case:Ideal Case:bT
pksizeC
June 2004 ISCC 2004 3
Packet Pair and Train Dispersion
June 2004 ISCC 2004 4
Packet Pairs Bandwidth Histogram
(a) Light load conditions(20%) (b) heavy load conditions(80%)
• SCDR is caused by dispersion expansion• PNCM is caused by dispersion compression
Packet-pair estimates: multimodality with cross traffic:
June 2004 ISCC 2004 5
As trains get longer, get “Asymptotic Dispersion Rate” or ADR ADR is not equal to Residual (available) Capacity We found and proved a physical interpretation (to be published): ADR is
the flow share, when merges are proportional to arrival rates at each link
Dovrolis’ results obtained for non-responsive cross traffic flows
Packet Train Bandwidth Histogram
June 2004 ISCC 2004 6
CapProbe: The Main Idea
Observation: Both expansion and compression of dispersion involve queuing due to cross traffic:
Dispersion expansion => second packet queued more Dispersion compression => first packet queued more
Packet pair with minimal end-to-end delay sum, is likely to be dispersed corresponding to narrow link capacity
Looking for packet pair with minimal delay sum is inexpensive
CapProbe appears accurate in most of our experiments, simulations and measurements
CapProbe fails under heavy (~>75%) utilization by non-responsive (UDP) traffic
June 2004 ISCC 2004 7
CapProbe Ideal Case: no cross traffic Real Case: dispersion may be compressed or
expanded by the cross traffic
Under-estimation due to expansion
Over-estimation due to compression
June 2004 ISCC 2004 8
CapProbe Both expansion and compression are due to
queuing A Packet-Pair sample with Minimal Delay Sum
can be used for Capacity Estimation
June 2004 ISCC 2004 9
Wireless Measurements
Bad channel retransmission larger dispersions lower estimated capacity
Results for Bluetooth-interfered 802.11b, TCP cross-traffic
Experiment No.
Capacity Estimated by
CapProbe (kbps)
Capacity Estimated by Strongest Mode
(kbps)
1 5526.68 4955.02
2 5364.46 462.8
3 5522.26 4631.76
4 5369.15 5046.62
5 5409.85 449.73
June 2004 ISCC 2004 10
ToUCLA-2 UCLA-3 UA NTNU
TimeCapacit
yTime Capacity Time Capacity Time Capacity
CapProbe
0’03 5.5 0’01 96 0’02 98 0’07 97
0’03 5.6 0’01 97 0’04 79 0’07 97
0’03 5.5 0’02 97 0’17 83 0’22 97
0’07 5.6 0’01 98 0’09 98 0’04 99
0’03 5.6 0’02 99 0’09 95 0’04 96
Pathrate
6’10 5.6 0’16 98 5’19 86 0’29 97
6’14 5.4 0’16 98 5’20 88 0’25 97
6’5 5.7 0’16 98 5’18 133 0’25 97
6’14 6.8 0’16 98 5’19 88 0’26 97
6’20 5.8 0’16 98 5’19 132 0’25 97
Pathchar
21’12 4.0 22’49 18 3 hr 34 3 hr 34
21’21 4.0 22’53 18 3 hr 31 3 hr 35
21’45 4.0 22’48 18 3 hr 32 3 hr 34
20’43 3.9 27’41 18 3 hr 34 3 hr 35
21.18 4.0 29’47 18 3 hr 30 3 hr 35
Comparison to Earlier Tools
June 2004 ISCC 2004 11
Implementation Issues
User vs. Kernel Mode generation of probes and measurements
End systems processing speed Probe packet size
June 2004 ISCC 2004 12
Testbed
User mode and kernel mode implementations
Slow system: Pentium II 500MHz CPU; Fast system: Pentium IV 1.8 GHz CPU
Probe packet sizes varied from 500 Bytes to 5K Bytes
June 2004 ISCC 2004 13
Measurement Experiments on Internet
PKSize(bytes)
YAHOO NTNU WLSH
1 2 3 1 2 3 1 2 3
User mode 1(slow machine)
500 571.4 285.7 452.8 133.3 571.4 67.8 1.45 1.45 1.41
1000 421.1 61.5 160.0 1000 381.0 216.2 1.51 1.47 1.5
3000 338.0 237.6 452.8 79.2 827.6 444.4 1.47 1.48 1.47
5000 231.2 156.2 239.5 209.4 210.5 243.9 1.48 1.47 1.47
User mode 2(fast machine)
500 64.5 148.1 400 54.1 56.3 40.8 1.26 1.48 1.28
1000 156.9 156.9 150.9 160 82.5 95.2 1.45 2.17 1.50
3000 103.4 112.1 117.0 111.6 123.7 93.1 1.47 1.49 1.46
5000 99.5 100.8 102.6 89.9 102.3 107.8 1.47 1.47 1.48
Kernel mode(slow machine)
500 9.2 102.6 10.2 89.4 90.9 74.0 1.39 1.42 1.5
1000 69.6 79.3 78.6 93.0 85.1 91.6 1.48 1.44 1.47
3000 92.0 86.3 90.0 96.2 98.4 95.6 1.45 1.46 1.45
5000 95.0 86.8 91.3 93.0 96.4 96.4 1.45 1.47 1.46
Narrow Link Capacity 100 100 1.5
Unit: Mbps
June 2004 ISCC 2004 14
Discussion
Packet SizePacket SizeNarrow Link CapacityNarrow Link Capacity
100 Mbps100 Mbps 10 Mbps10 Mbps 1 Mbps1 Mbps
500 bytes500 bytes 0.04 ms 0.4 ms 4 ms
1000 bytes1000 bytes 0.08 ms 0.8 ms 8 ms
3000 bytes3000 bytes 0.24 ms 2.4 ms 24 ms
5000 bytes5000 bytes 0.40 ms 4 ms 40 ms
Required time resolution for accurate estimation
For high speed networks, either a high time resolution machine or a large probing packet size is needed for accuracy
Fine resolution may not be possible in user mode A large packet size increases the chances expansion of dispersion Required time resolution T = pksize / C:
June 2004 ISCC 2004 15
Passive CapProbe CapProbe is an active approach and using ICMP
packets.
Passive approach is less intrusive, thus more scalable
Passive CapProbing within TCP requires back to back TCP packet transmission
Simulation => 15~20% of TCP data packets are sent back-to-back
June 2004 ISCC 2004 16
Simulation
The network topology used in our simulations consists of a six-hop path with capacities {10, 7.5, 5.5, 4, 6 and 8} Mbps.
DelACK is disabled in the simulation.
Different cross traffic are used, with packet size 1000 bytes and 200 bytes.
June 2004 ISCC 2004 17
Passive CapProbePassive CapProbe Active CapProbeActive CapProbe
June 2004 ISCC 2004 18
Conclusion Either a high time resolution machine or a large
probing packet size is necessary for accurate capacity estimation
Passive CapProbing within TCP is feasible, minor TCP sender modification helps a lot (future work)
Other Future work: Experiments at speeds higher than 100 Mbps Passive CapProbing in TFRC, other applications Use of capacity estimates in TCPW and overlays
construction
June 2004 ISCC 2004 19
T h a n k sT h a n k s
Improving Wireless Link Throughput via Interleaved FEC