An efficient and fair explicit congestion control protocol for high bandwidth-delay product networks

ABSTRACT

XCP and VCP can achieve excellent performance under high bandwidth-delay product networks, but they all have some defects. In XCP, router needs to calculate a feedback for each departing packet, the cost will be un-negligible in high-speed networks. In VCP, router classifies the degree of congestion into three regions and encodes it in ECN bits to guide sender to update its congestion window (cwnd). Obviously, the feedback in VCP is still not precise enough so that VCP has slow fairness convergence speed. Moreover VCP does not consider the TCP unfairness problem induced by heterogeneous packet sizes. Furthermore, both XCP and VCP cannot solve unfairness problem led by multiple bottleneck routers. To overcome drawbacks in XCP and VCP, we propose a new efficient and fair explicit congestion control protocol in this paper, for convenience, we call it EFXCP. In a sample interval, EFXCP router calculates an expected fair throughput and feedbacks it to senders. EFXCP sender updates its cwnd to make its' throughput converge to expected fair throughput so that the system can rapidly converge to the optimal state. On the one hand, EFXCP calculates feedback only once in a sample interval, the calculation cost is much less than that in XCP. On the other hand, EFXCP adopts the precise feedback, the fairness convergence speed is much faster than that in VCP. Furthermore, one outstanding advantage of EFXCP is that it can achieve excellent fairness in all cases, all competing flows have almost the same throughput when system is stable. Using control theory, we demonstrate that EFXCP is global asymptotic stable regardless of bottleneck capacity, round trip delay time (RTT) and number of sources. Moreover, we quantificationally analyze the rise time and the fairness convergence time of EFXCP. Further, we evaluate the performance of EFXCP using extensive NS2 simulations over a wide range of network scenarios, the results show that EFXCP retains the advantages of XCP and VCP, and discards their disadvantages, and achieves excellent performance. Moreover, like VCP and XCP, the router does not maintain per-flow state.

REFERENCES

• [1]: V. Jacobson, Congestion avoidance and control, Symposium proceedings on Communications architectures and protocols, p.314-329, August 16-18, 1988, Stanford, California, United States
• [2]: Sally Floyd, Highspeed TCP for large cwnds, RFC 3649 December 2003.
• [3]: Tom Kelly, Scalable TCP: improving performance in highspeed wide area networks, ACM SIGCOMM Computer Communication Review, v.33 n.2, April 2003 [doi>10.1145/956981.956989]
• [4]: R.N. Shorten, D.J. Leith, H-TCP: TCP for high-speed and long-distance networks, in: Proceedings of PFLDnet'04, Chicago, IL, USA, 2004.
• [5]: Bhandarkar, Sumitha, Jain, Saurabh, Narasimha Reddy, A.L. and Kohler,, Improving TCP performance in high bandwidth high RTT links using layered congestion control. PFLDNet.
• [6]: Lisong Xu, Khaled Harfoush, Injong Rhee, Binary increase congestion control for fast long-distance networks, in: Proceedings of IEEE Infocom 2004, March 2004.
• [7]: Dina Katabi , Mark Handley , Charlie Rohrs, Congestion control for high bandwidth-delay product networks, Proceedings of the 2002 conference on Applications, technologies, architectures, and protocols for computer communications, August 19-23, 2002, Pittsburgh, Pennsylvania, USA
• [8]: Yong Xia , Lakshminarayanan Subramanian , Ion Stoica , Shivkumar Kalyanaraman, One more bit is enough, ACM SIGCOMM Computer Communication Review, v.35 n.4, October 2005
• [9]: Dina Katabi, Mark Handley, Using precise feedback for controlling congestion in the Inter-net, MIT-LCS Technical Report 820.
• [10]: Athuraliya, S., Li, V., Low, S. and Yin, Q., REM: active queue management. IEEE Network. v15 i3. 48-53.
• [11]: Srisankar Kunniyur , R. Srikant, Analysis and design of an adaptive virtual queue (AVQ) algorithm for active queue management, Proceedings of the 2001 conference on Applications, technologies, architectures, and protocols for computer communications, p.123-134, August 2001, San Diego, California, United States
• [12]: Available from: .
• [13]: UCN/LBL/VINT, Network Simulator-NS2, Available from: .
• [14]: Mark E. Crovella , Azer Bestavros, Self-similarity in World Wide Web traffic: evidence and possible causes, IEEE/ACM Transactions on Networking (TON), v.5 n.6, p.835-846, Dec. 1997 [doi>10.1109/90.650143]
• [15]: Sally Floyd, Connections with multiple congested gateways in packet-switched networks part 1: one-way traffic, ACM SIGCOMM Computer Communication Review, v.21 n.5, p.30-47, Oct. 1991 [doi>10.1145/122431.122434]
• [16]: Sally Floyd, Specifying new congestion control algorithms, RFC 5033, August 2007.

INDEX TERMS

Keywords:
Congestion control, Fairness, High-speed network