Random capture algorithms fluid limits and stability

We introduce a distributed stateless MAC protocol referred to as Random Capture Algorithm (RCA) and analyze its performance in networks where interference is modeled by a contention graph. RCA does not require any message passing, nor transmitters to be aware of the content of their respective buffers. Yet, it achieves at least the same stability region as that obtained with maximal scheduling. We prove that RCA is actually throughput optimal in networks with N-partite interference graphs. We present simulation results that suggest that RCA is also throughput optimal on simple networks whose interference graphs are not N-partite. From there, it is tempting to conjecture that RCA are throughput optimal in all networks.

[1]  R. Srikant,et al.  Distributed Link Scheduling With Constant Overhead , 2006, IEEE/ACM Transactions on Networking.

[2]  J. Dai On Positive Harris Recurrence of Multiclass Queueing Networks: A Unified Approach Via Fluid Limit Models , 1995 .

[3]  Devavrat Shah,et al.  Distributed algorithm and reversible network , 2008, 2008 42nd Annual Conference on Information Sciences and Systems.

[4]  Eytan Modiano,et al.  Maximizing throughput in wireless networks via gossiping , 2006, SIGMETRICS '06/Performance '06.

[5]  Jean C. Walrand,et al.  Distributed Random Access Algorithm: Scheduling and Congestion Control , 2009, IEEE Transactions on Information Theory.

[6]  Leandros Tassiulas,et al.  Stability properties of constrained queueing systems and scheduling policies for maximum throughput in multihop radio networks , 1992 .

[7]  Alexandre Proutière,et al.  Performance of random medium access control, an asymptotic approach , 2008, SIGMETRICS '08.

[8]  Jean C. Walrand,et al.  A Distributed CSMA Algorithm for Throughput and Utility Maximization in Wireless Networks , 2010, IEEE/ACM Transactions on Networking.

[9]  Mark H. Davis Markov Models and Optimization , 1995 .

[10]  Alexandre Proutière,et al.  Throughput of random access without message passing , 2008, 2008 42nd Annual Conference on Information Sciences and Systems.

[11]  Devavrat Shah,et al.  Network adiabatic theorem: an efficient randomized protocol for contention resolution , 2009, SIGMETRICS '09.

[12]  Nick McKeown,et al.  Maximum size matching is unstable for any packet switch , 2003, IEEE Communications Letters.