A Consensus-State-Aware Adaptive Algorithm for Tuning CSMA Parameters in Multi-Agent Networks

Reaching consensus to realize a group objective is one of the fundamental problems for collaborative control of multi-agent networks, over which the convergence speed is a momentous performance index. However, in many multi-agent networks that using CSMA/CA mechanism, it is difficult to reach consensus quickly when there are tremendous devices to send packets in the network under communication constraints, due to the decline of communication reliability caused by the channel competition mechanism. This paper proposes a distributed adaptive parameters tuning algorithm for CSMA/CA mechanism, which can make the state of all devices in a common multi- agent network converge to an identical value as soon as possible with less energy consumption, especially in the situation where the channel competition is fiercer. The algorithm is light-weight and has lower complexity without any modification to existing communication protocols. Simulation results demonstrate the excellent superiority of the algorithm in convergence speed and energy consumption.