Decentralized Reinforcement Learning Based MAC Optimization

In this paper, we propose a novel decentralized framework for optimizing the transmission strategy of Irregular Repetition Slotted ALOHA (IRSA) protocol in sensor networks. The proposed method is inspired by reinforcement learning algorithms. To deal with sensor nodes limited lifetime and communication range, we allow sensor nodes to decide how many packet replicas to transmit and when to transmit them considering only their own buffer state. We show that this information is sufficient and can help avoiding packets' collisions and improving the throughput significantly. We solve the problem using the decentralized partially observable Markov Decision Process (POMDP) framework, where we allow each node to decide independently of the others how many packet replicas to transmit and when. The performance of the proposed method is compared with the native IRSA protocol. The results make clear that our method leads to large throughput gains without imposing nodes coordination, in particular, when network traffic is heavy.