Agent-based Channel Selection scheme against location inference attack in cognitive radio networks

Location inference attack may cause location privacy leakage of secondary users SUs in database-driven cognitive radio networks. Malicious entities could infer SUs´ location with channel usage information. For the unpredictable return of primary users, there is no solution that is effective to location inference attack. To prevent location inference attack, we propose an Agent-based Channel Selection(ACS) scheme in this paper. In this scheme, the database allocates available spectrum with self-coexistence mechanism to avoid interference among SUs, and each base station registers spectrum information as an agent in the database SUs. Thus, the proposed ACS scheme can decrease the probability of being geo-located for SUs effectively. The variation tendency of the average probability of geo-location with different factors is shown in simulation.