A game approach for dynamic resource allocation in cognitive radio networks

Resource allocation is an important issue in cognitive radio networks. Optimizing this process is crucial for primary users (PUs) and secondary users (SUs) to maximize the performance of the whole network. In this paper, we investigate the RA-PS problem in cognitive radio networks, in which each PU acts as a relay for multiple SUs. We formulate this problem as a two-tier game. In the first RA-Game, each PU selects the expected quantities of sub-channels to achieve the highest payoff. Then in the second tier game, namely PS-Game, each PU sells his unused radios to SUs nearby, which is performed as an auction game. In addition, we investigate the existence of Nash equilibria of the two games, which shows our mechanism can efficiently achieve load balance. Furthermore, we propose distributed algorithms including radio allocation and PU selection, which are performed by PUs and SUs independently in a distributed manner. Simulation results show that our proposed mechanism can not only effectively converge to Nash equilibrium but also obtain a capacity gain with the help of relay by PU.