A Simple Pseudo-Bayesian Backoff Algorithm for Unsaturated Slotted CSMA Systems

In unsaturated slotted carrier sense multiple access (CSMA) systems, the queue states of the stations (STNs) in uplink vary over time in general. This phenomenon is the main difficulty that complicates the efficient system operation. In this paper, we propose an estimation algorithm to estimate the backlog size, i.e., the number of STNs who have packets to transmit in each slot. Then, based on the estimation, the transmission probability is controlled to maximize the system throughput. One key feature of the proposed algorithm is that it only applies the information about the idle and busy channel slots to estimate the backlog size. Through extensive computer simulations, we show that the proposed algorithm achieves near- optimal system performance in term of mean and variance of queuing delay.