Distributed coalition formation game for enhancing cooperative spectrum sensing in cognitive radio ad hoc networks

摘要

Cooperative spectrum sensing (CSS) among secondary users (SUs) in cognitive radio ad hoc networks exhibits an essential trade-off minimising the misdetection probability of the primary user and maintaining a reasonable false alarm probability to obtain efficient spectrum utilisation. In this study, distributed coalition formation game with a genetic algorithm is proposed to select the optimal coalitional head. It provides an incentive to SUs that improve their detection performance and cooperation among SUs. Since spectrum sensing will incur a cost for SUs, an incentive mechanism is necessary to motivate active participation. Consider the CSS problem as a non-transferable coalition game and it is used to allow the SUs to self-organise into disjoint coalitions while accounting for the CSS. The authors calculate payoff for each SU to form minimum winning coalition and to analyse their strategy profile for joining in coalition based on its utility maximisation. Simulation results reveal that the proposed algorithm has significantly improved sensing accuracy, minimised energy consumption and thereby enhanced CSS.