Constrained Stochastic Game in Licensed-Assisted Access for Dynamic Contention Window Adaptation

Licensed-assisted access (LAA) allows operations on unlicensed spectrum for long term evolution. Enhanced listen-before-talk (LBT) scheme is required to improve LAA performance. In this letter, we investigate the LAA dynamic contention window adaptation scheme for individual and cooperative throughput optimization under buffer length constraint. We formulate a constrained stochastic game for this competitive scenario. The difference of buffer status is leveraged for collision alleviation and throughput improvement. A linear programming-based method is introduced to obtain the optimal policy, which is proved to be the constrained Nash equilibrium. Simulation results validate that the proposed scheme can achieve higher throughput than the existing LBT Cat-3 and Cat-4 schemes.