Relay Selection for Wireless Cooperative Networks using Adaptive Q-learning Approach

Relay selection is an effective method to improve the system performance of co-operative communication, and thus has received significant attention. In this paper, by assuming that the instantaneous channel state information (CSI) is unknown at the source and relays, we propose a Q-learning (QL) based on relay se-lection method, which can select the relays with maximum cumulative reward to obtain the maximum throughput of the cooperative networks. Besides, Boltzmann learning rule is adopted to achieve well counterpoise between action exploration and exploitation. Simulation results show that the proposed QL algorithm can select the optimal relay adaptively and improve the system performance significantly in comparison with random selection algorithm. Furthermore, it can be found that as the number of relay nodes increases, the QL algorithm can still adaptively select the optimal relay without increasing the computational load.