Dynamic channel allocation based on the game learning in cognitive radio networks

A priority table of available channels in network is structured with the analysis of the channel usability by cooperative spectrum sensing. With this table, we design an information exchange and access mechanism based on TDMA mode. We bring in the EWA (Experience-Weight Attraction) game learning model, and propose a channel selection learning algorithm based on the channel priority and the interference between cognitive radio users (CRUs). Simulation results compared with the learning automata based algorithm and no-regret learning algorithm show that the proposed algorithm can adaptively select channel with the best usability through the learning of historical experience, increase the throughput of system and have a better equity in the resource allocation.