The increasing number of wireless devices operating in unlicensed spectrum motivates the development of intelligent adaptive approaches to spectrum access that go beyond traditional carrier sensing. We develop a novel distributed implementation of a policy gradient method known as Proximal Policy Optimization modelled on a two stage Markov decision process that enables such an intelligent approach, and still achieves decentralized contention-based medium access. In each time slot, a base station (BS) uses information from spectrum sensing and reception quality to autonomously decide whether or not to transmit on a given resource, with the goal of maximizing proportional fairness network-wide. Empirically, we find the proportional fairness reward accumulated by the policy gradient approach to be significantly higher than even a genie-aided adaptive energy detection threshold. This is further validated by the improved sum and maximum user throughputs achieved by our approach.
View on arXiv