Access Link Capacity Monitoring with TFRC Probe LingJyh

Access Link Capacity Monitoring with TFRC Probe Ling-Jyh Chen, Tony Sun, Dan Xu, M. Y. Sanadidi, Mario Gerla Computer Science Department, University of California at Los Angeles Oct. 3, 2004 E 2 EMON 2004

Motivation l Knowledge of link capacity is important for network management, pricing, and Qo. S support. l The link capacity of a network connection may vary dramatically due to vertical handoff, dynamic channel allocation, and wireless channel quality. l Knowing the link capacity will permit the source to rapidly and appropriately adapt the outbound data transmissions rate. Oct. 3, 2004 E 2 EMON 2004 2

Access Link Capacity Monitoring l Requirements l l l provide correct information work passively without adding excess overhead promptly react to occurrences in network events maintain end-to-end semantics Our approach: TFRC Probe l l l embed Cap. Probe algorithm within TFRC simple, accurate, passive, timely, and end-to-end extensible to other protocols and applications Oct. 3, 2004 E 2 EMON 2004 3

Potential Applications l Adaptive multimedia streaming l Congestion control l Overlay network structuring l Wireless link monitoring l Mobility detection Oct. 3, 2004 E 2 EMON 2004 4

The Capacity Estimation Problem l Estimate minimum link capacity on an Internet path, as seen at the IP level 100 Mbps l 50 Mbps 10 Mbps (Link Capacity) 100 Mbps Design Goals l l l Oct. 3, 2004 End-to-end: assume no help from routers Inexpensive: Minimal additional traffic and processing Fast: converges to capacity fast enough for the application E 2 EMON 2004 5

Packet Pair Dispersion T 1 T 3 Narrowest Link T 2 20 Mbps Oct. 3, 2004 10 Mbps T 3 5 Mbps. T 3 10 Mbps E 2 EMON 2004 T 3 20 Mbps 8 Mbps 6

Ideal Packet Dispersion l No cross-traffic Capacity = (Packet Size) / (Dispersion) Oct. 3, 2004 E 2 EMON 2004 7

Compression and Expansion l l First packet queueing → compressed dispersion → Overestimation Second packet queueing → expanded dispersion → Underestimation Oct. 3, 2004 E 2 EMON 2004 8

Cap. Probe l Filter PP samples that do not have minimum queuing time l Dispersion of PP sample with minimum delay sum reflects capacity l Cap. Probe combines both dispersion and e 2 e transit delay information l Cap. Probe is simple, fast, and accurate Oct. 3, 2004 E 2 EMON 2004 9

TFRC: TCP-Friendly Rate Control l l TFRC is an equation based unicast multimedia streaming protocol. TFRC mimics the TCP long-term throughput by utilizing the function: The receiver is responsible for calculating the loss event rate p and sending the information back to the sender once per round-trip time. The sender is responsible for adjusting its sending rate Tactual to be close to T. Oct. 3, 2004 E 2 EMON 2004 10

TFRC Probe Embedding Cap. Probe within TFRC l Three design issues: 1. Accurate capacity estimation 2. Fast estimation process 3. Minimal traffic overhead and modification to the original TFRC Two design options: l 1. 2. Oct. 3, 2004 One-way estimation Round-trip estimation E 2 EMON 2004 11

TFRC Probe l Accurate Capacity Estimation Embed Cap. Probe algorithm within TFRC by sending two packets back-to-back every n packets Oct. 3, 2004 E 2 EMON 2004 12

TFRC Probe l Fast Link Capacity Estimation l Fast in estimating link capacities from samples Cap. Probe has been shown to be a fast and accurate technique for link capacity estimation. l Fast in getting samples The speed of sampling will increase/decrease when the packet sizes decrease/increase. Oct. 3, 2004 E 2 EMON 2004 13

TFRC Probe Packet size adaptation l l l Rsend is the sending rate of data packets S is the number of samples needed to get a reliable capacity estimation P is the data packet size t is the expected time to get one capacity estimation. n is the number of data packets between samples Oct. 3, 2004 E 2 EMON 2004 14

Simulation l l The monitoring ability of TFRC Probe is verified using NS -2 simulator The bottleneck link (between node 3 and 4) is shared by all the data flows and configured as an asymmetric link with various capacities in the forward direction and fixed capacity (100 Kbps) in the backward direction. Oct. 3, 2004 E 2 EMON 2004 15

Simulation Cross Traffic l l Description Type I 4 FTP flows (from node 7 to 10, 8 to 9, 11 to 14, and 12 to 13); 1500 bytes/packet Type II 4 CBR flows (from node 7 to 10, 8 to 9, 11 to 14, and 12 to 13); 500 bytes/packet; 80% load on the bottleneck Type III 16 Pareto flows with alpha = 1. 9 (4 flows from node 7 to 10, 4 flows from 8 to 9, 4 flows from 11 to 14, and 4 flows from 12 to 13); 1000 bytes/packet; 80% load on the bottleneck Three type of cross traffic are employed in the simulaiton. The link capacity estimation results are collected after 20 and 50 samples. Oct. 3, 2004 E 2 EMON 2004 16

Simulation Results TFRC Probe Cap. Probe no cross traffic 20 samples 100 K 500 K 100 K 1 M 100 K 5 M 100 K 50 samples 100 K 500 K 100 K 1 M 100 K 5 M 100 K cross traffic type II 20 samples 100 K 500 K 1 M 100 K 50 samples 100 K 500 K 1 M 100 K 5 M 100 K cross traffic type III 20 samples 100 K 500 K 1 M 100 K 50 samples 100 K 500 K 1 M 100 K 5 M 100 K bottleneck capacity of the forward direction link 100 Kbps 500 Kbps 1 Mbps 5 Mbps bottleneck capacity of the backward direction link 100 Kbps Oct. 3, 2004 E 2 EMON 2004 17

Experiments l l Implementation: based on the original TFRC codes (Linux platform) Experiments: l Without packet size adaptation l l Evaluate the effectiveness of TFRC Probe in wired Internet links (symmetric and asymmetric links) and wireless links (1 x. RTT and 802. 11 b) With packet size adaptation l Oct. 3, 2004 Evaluate the feasibility of monitoring wireless link capacity using TFRC Probe. E 2 EMON 2004 18

Experiment Results 1 l Wired Internet links (no pkt size adaptation) Run 1 Ethernet 100 Mbps DSL Down. Link 2 Mbps DSL Up. Link 128 Kbps Oct. 3, 2004 Run 2 Run 3 Run 4 Run 5 Capacity Time Capacity Time 20 samples 92. 18 5. 5 94. 92 5. 5 94. 25 5. 6 107. 99 5. 5 94. 07 5. 6 50 samples 98. 17 5. 8 104. 31 5. 7 94. 25 5. 9 94. 02 5. 7 94. 07 5. 8 100 samples 98. 17 6. 1 104. 31 6. 0 94. 25 6. 2 94. 02 6. 1 94. 07 6. 2 20 samples 1. 982 7. 0 1. 807 8. 6 1. 710 8. 8 1. 782 6. 7 1. 835 6. 7 50 samples 1. 982 12. 7 1. 865 12. 4 1. 835 12. 6 2. 145 12. 6 1. 652 10. 5 100 samples 1. 982 20. 9 1. 865 18. 9 1. 835 18. 9 2. 145 19. 0 1. 652 16. 9 20 samples 0. 114 70. 9 0. 122 69. 5 0. 115 72. 7 0. 110 81. 5 0. 119 71. 4 50 samples 0. 114 145. 0 0. 122 137. 6 0. 115 146. 0 0. 115 147. 5 0. 112 142. 3 100 samples 0. 115 265. 6 0. 122 258. 5 0. 119 258. 6 0. 115 266. 0 0. 112 257. 5 E 2 EMON 2004 19

Experiment Results 2 l Wireless links (no pkt size adaptation) Run 1 Run 2 Run 3 Run 4 Run 5 Capacity Time Capacity Time 20 samples 0. 186 115. 9 0. 187 87. 1 0. 186 68. 3 0. 140 116. 0 0. 186 61. 6 50 samples 0. 140 156. 4 0. 140 122. 3 0. 187 113. 2 0. 140 188. 5 0. 140 113. 1 100 samples 0. 140 230. 0 0. 140 182. 9 0. 140 187. 0 0. 140 286. 8 0. 140 212. 3 20 samples 0. 65 17. 0 0. 91 17. 8 0. 85 18. 7 0. 94 17. 7 0. 92 17. 9 50 samples 0. 93 25. 6 0. 93 27. 8 0. 91 28. 2 0. 94 26. 3 0. 93 27. 8 100 samples 0. 91 39. 6 0. 93 41. 6 0. 91 42. 6 0. 92 41. 2 0. 93 42. 3 20 samples 1. 48 16. 7 1. 80 16. 8 1. 81 16. 9 1. 77 17. 9 1. 80 18. 0 50 samples 1. 48 21. 8 1. 74 21. 2 1. 68 21. 6 1. 09 23. 9 1. 69 23. 5 100 samples 1. 49 29. 1 1. 47 29. 6 1. 80 29. 8 1. 72 32. 3 1. 73 32. 5 20 samples 3. 73 14. 8 3. 86 15. 3 4. 36 14. 9 3. 83 15. 1 4. 45 14. 6 50 samples 4. 18 18. 4 3. 86 19. 7 3. 80 19. 5 4. 06 19. 3 3. 87 18. 6 100 samples 4. 18 23. 2 3. 92 24. 4 3. 80 24. 06 24. 0 3. 87 22. 8 20 samples 8. 09 7. 60 7. 5 7. 91 7. 5 7. 00 7. 6 6. 26 7. 9 50 samples 8. 09 9. 0 7. 60 10. 1 6. 23 9. 6 7. 00 10. 2 6. 26 10. 2 Oct. 3, 2004 100 samples 7. 84 12. 7 7. 60 E 2 EMON 11. 8 2004 6. 23 11. 6 6. 60 12. 0 6. 26 20 11. 9 1 x. RTT 150 Kbps 802. 11 b 1 Mbps 802. 11 b 2 Mbps 802. 11 b 5. 5 Mbps 802. 11 b 11 Mbps

Experiment Results 3 l 802. 11 b link with varying data rate (with pkt size adaptation) Oct. 3, 2004 E 2 EMON 2004 21

Conclusion l l TFRC Probe is simple, accurate, passive, timely, and end-to-end. The evaluation of TFRC Probe is performed by both simulation and testbed experiments. The same concept can be applied to other UDP based application protocols (e. g. RAP, RTP, UDP based FTP and P 2 P file downloading) and other emerging data transmission protocols (e. g. DCCP). Cap. Probe website: http: //nrl. cs. ucla. edu/Cap. Probe Oct. 3, 2004 E 2 EMON 2004 22

Thanks! Oct. 3, 2004 E 2 EMON 2004 23