Spectrum sharing in cognitive networks

Abstract

Cognitive Radio Networks (CRNs) provide an efficient solution to address the increasing demand for spectrum resources. The cooperation among Secondary Users (SUs) improves the sensing performance and spectrum efficiency. A traffic-demand based cooperation strategy of SUs in multichannel cognitive networks is analysed.When a SU has high traffic demand it can choose to sense multiple channels in the sensing period and obtain more chances to use spectrum resources. This problem has been formulated as a non-transferable utility (NTU) overlapping coalitional game. In this game each SU implements a cooperation strategy according to its expected payoff, which takes into account the expected throughput and energy efficency.Two algorithms Overlapping coalitional game (OCF) and Sequential coalition game (SCF) have been analyzed and a proposed algorithm has been introduced. In proposed coalition formation algorithm SUs are given priority according to data in the buffer and a coalition is developed based on the channel conditions. Each SU chooses a coalition based on its utility history and based on largest uncertainty reduction the SU joins the coalition. Proposed algorithm guarantees less complexity and more expected throughput than all other algorithms. Simulation results prove that our proposed coalition formation algorithm provides better throughput than OCF and SCF algorithms.