Collaborative spectrum sharing based on information pooling for cognitive radio networks with channel heterogeneity

In this paper, we investigate the problem of multiuser channel selection in cognitive radio networks (CRN) with heterogenous channel availability. We consider a practical scenario where the channel availability statistics are unknown. Due to limited sensing capabilities, secondary users (SUs) can only sense partial of the channels. In this case, secondary users can collaboratively share their respective sensing results to gather enough information about the statistics of the unknown channels. By taking into account the competition among SUs, We formulate the problem of multiuser channel selection as an evolutionary game in which SUs make their decisions under replicator dynamics. Furthermore, we develop a decentralized channel selection algorithm based on learning automata, and prove that the replicator dynamics can converge to the evolutionary stable strategy efficiently. Simulation results show that our proposed learning algorithm helps SUs share the spectrum efficiently and the system throughput obtained outperforms that of the random access scheme.