Overflow management with multipart packets

We study an abstract setting, where the basic information units (called "superpackets") do not fit into a single packet, and are therefore spread over multiple packets. We assume that a superpacket is useful only if the number of its delivered packets is above a certain threshold. Our focus of attention is communication link ingresses, where large arrival bursts result in dropped packets. The algorithmic question we address is which packets to drop so as to maximize goodput. Specifically, suppose that each superpacket consists of k packets, and that a superpacket can be reconstructed if at most s ? k of its packets are lost, for some given parameter 0 ≤ s < 1 . We present a simple online distributed randomized algorithm in this model, and prove that in any scenario, its expected goodput is ? opt k ( 1 - s ) ? , where opt denotes the best possible goodput by any algorithm, and ? denotes the size of the largest burst (the bound can be improved as a function of burst-size variability). We also analyze the effect of buffers on goodput under the assumption of fixed burst size, and show that in this case, when the buffer is not too small, our algorithm can attain, with high probability, ( 1 - ? ) goodput utilization for any ? 0 . Finally, we present some simulation results that demonstrate that the behavior of our algorithm in practice is far better than our worst-case analytical bounds.

[1]  Boaz Patt-Shamir,et al.  Buffer overflow management in QoS switches , 2001, STOC '01.

[2]  Johan Håstad,et al.  Clique is hard to approximate within n/sup 1-/spl epsiv// , 1996, Proceedings of 37th Conference on Foundations of Computer Science.

[3]  Michael H. Goldwasser A survey of buffer management policies for packet switches , 2010, SIGA.

[4]  J. Håstad Clique is hard to approximate withinn1−ε , 1999 .

[5]  Boaz Patt-Shamir,et al.  Competitive Router Scheduling with Structured Data , 2011, WAOA.

[6]  Alexander Schrijver,et al.  On the Size of Systems of Sets Every t of Which Have an SDR, with an Application to the Worst-Case Ratio of Heuristics for Packing Problems , 1989, SIAM J. Discret. Math..

[7]  Boaz Patt-Shamir,et al.  Competitive buffer management with packet dependencies , 2009, 2009 IEEE International Symposium on Parallel & Distributed Processing.

[8]  Jan Kratochvíl,et al.  Independent Sets with Domination Constraints , 2000, Discret. Appl. Math..

[9]  Boaz Patt-Shamir,et al.  Optimal smoothing schedules for real-time streams (extended abstract) , 2000, PODC.

[10]  Boaz Patt-Shamir,et al.  Online set packing and competitive scheduling of multi-part tasks , 2010, PODC '10.

[11]  Oded Schwartz,et al.  On the Complexity of Approximating k-Dimensional Matching , 2003, RANDOM-APPROX.

[12]  Boaz Patt-Shamir,et al.  Optimal smoothing schedules for real-time streams , 2004, PODC '00.

[13]  Piotr Berman,et al.  A d/2 Approximation for Maximum Weight Independent Set in d-Claw Free Graphs , 2000, Nord. J. Comput..

[14]  Jill M. Boyce,et al.  Packet loss effects on MPEG video sent over the public Internet , 1998, MULTIMEDIA '98.

[15]  Gabriel Scalosub,et al.  Buffer Management for Aggregated Streaming Data with Packet Dependencies , 2010, IEEE Transactions on Parallel and Distributed Systems.

[16]  Michael Luby,et al.  A simple parallel algorithm for the maximal independent set problem , 1985, STOC '85.