Distributed Spectrum Sharing in Cognitive Radio Networks - Game Theoretical View

Consumer communication networking allows connections among consumer devices for possible Ad Hoc or heterogeneous networking on licensed and unlicensed bands. To support maximum flexibility of consumer communication networks, cognitive radio (CR) technology by spectrum sharing can well serve the purpose. Realistic spectrum sharing for cognitive radio networks (CRN) shall be distributed and based on partially available information of spectrum sensing, due to a possible good number of cognitive radios and impossible to perfectly exchange channel availability information. However, existing study assumes either centralized or perfect spectrum sensing. In this paper, we pioneer explore spectrum sensing as a side information and take imperfection of spectrum sensing results into consideration. We therefore develop spectrum sharing algorithms for spectrum access strategy of CRs under two general scenarios: public spectrum information broadcast from the base station and private spectrum information via individual spectrum sensing. Under the public spectrum in-formation, CRs are aware of the strategy of its opponents and therefore game theory model reaches Nash eqUilibrium as a solution. On the other hand, CRs have to follow maximin criterion with only local spectrum information. Difference in the behaviors of CRs between the two scenarios is identified and the optimal spectrum access strategy is proposed accordingly for both cases. Numerical results demonstrate that the proposed algorithms work effectively and prevent the system from collision even in a large network.