A multiaccess tree algorithm with free access, interference cancellation and single signal memory requirements

Tree algorithms are a well studied class of collision resolution algorithms for solving multiple access control problems. Successive interference cancellation, which allows one to recover additional information from otherwise lost collision signals, has recently been combined with tree algorithms with blocked access [Y. Yu, G.B. Giannakis, SICTA: A 0.693 contention tree algorithm using successive interference cancellation, in: INFOCOM 2005. 24th Annual Joint Conference of the IEEE Computer and Communications Societies, Miami, USA, 2005, pp. 1908–1916], providing a substantially higher maximum stable throughput (MST): 0.693 for Poisson arrivals, given an infinite number of memory locations for storing signals. We propose a novel tree algorithm for a similar problem, but with two relaxed model assumptions: free access is supported and a single signal memory location suffices. A study of the maximal stable throughput of this algorithm is provided using matrix analytical methods; as a result, an MST of 0.5698 for Poisson arrivals is achieved. Our methodology also allows us to investigate the MST when the multiple access channel is subject to Markovian arrival processes.