An autonomous pareto optimality achieving algorithm beyond Aloha games with spatial reuse

Aloha games with spatial reuse study the interactions among a group of selfish transmit-receive pairs which share a common collision channel using slotted-Aloha-like protocols. These Tx-Rx pairs are allowed to reuse the channel if they cause negligible interference to each other. Our work in [1] has proved the existence of a Least Fixed Point (LFP) which is the most energy-efficient operating point as well as the unique Nash Equilibrium (NE) in such games. Based on the earlier derived conditions for the stability of this NE and the way to converge to this NE, it is possible to design a self-adaptive algorithm for the players to self-adjust their target rates based on a set of pre-installed rules so that the network always achieves Pareto optimal bandwidth utilization. In this paper, we implement such an algorithm in a fully distributed manner, which requires no information exchange among the players. Each player repeatedly measures its current throughput and uses the measured value to make myopic best response to the current channel idle rate. Our simulations show that the system indeed achieves close to Pareto optimal performance while guaranteeing a certain degree of fairness. The algorithm is robust and can handle various practical issues such as the dynamic arrival/departure of players, parameter estimation errors, etc.