Maximum Mean Discrepancy Based Secure Fusion Strategy for Robust Cooperative Spectrum Sensing

Abstract

Collaborative spectrum sensing (CSS) in Cognitive Radio based Networks (CRNs) is vulnerable to Spectrum Sensing Data Falsification (SSDF) attack. Many existing defense mechanisms assume the number of malicious users are in minority or the attackers' flip rates are identical and fixed. However, such assumption doesn't hold when some intelligent attacks such as Sybil attack are launched successfully, wherein one dedicated attacker can pretend to be multiple attackers. Besides, most existing approaches adopting hard decision approach by identifying the attackers first and ignore their sensing reports in the fusion operation. Thus the overall system performance is degraded since some intelligent attackers' sensing reports are still possible to be genuine. On the other hand, representative existing work using soft decision method still cannot distinguish the malicious users and honest users correctly under certain condition. The defense mechanism doesn't perform properly for the case when the distributions of two sensing reports are different but have equal mean and variance. In this paper, we propose a secure fusion strategy which adopts soft decision method and can distinguish malicious users and honest users under any distribution of sensing reports using maximum mean discrepancy (MMD). Our proposed CSS scheme is suitable for any general CRN application scenarios. The simulation results show our proposed defense mechanism outperforms the existing works.