Abstract: In order to solve the waste of resources caused by the static spectrum allocation policy, cognitive radio technology can be used to dynamically and efficiently manage spectrum, and spectrum sharing becomes the key issue. To make secondary users to share spectrum fairly, an evolutionary game-based spectrum sharing scheme in cognitive radio network was proposed. Because cognitive radio network was dynamic and heterogeneous, the wireless channel was unreliable and affected by environment easily, the knowledge of fuzzy mathematics was introduced to characterize the parameters related to the secondary user satisfaction degree, and the Gas Brownian Motion Optimization algorithm was used to solve the problem of spectrum access of secondary users. The fairness of spectrum sharing was considered in utility function, and the dynamic equilibrium was solved according to replicator dynamic model. Simulation results show that secondary users can get high utilities under the proposed scheme.

Key words: cognitive radio network, evolutionary game, spectrum sharing, Gas Brownian Motion Optimization algorithm, replicator dynamic model