Cooperative Estimation of Primary Traffic Under Imperfect Spectrum Sensing and Byzantine Attacks

Abstract

Cognitive radio (CR) systems exploit the accurate knowledge of primary traffic statistics to improve the CR performance and reduce the harmful interference on primary network. It is essential for cooperative CRs to operate in an efficient secured manner while estimating the primary statistics. In this paper, a new reporting mechanism for cooperative estimation of primary traffic is proposed to increase the spectrum and energy efficiency. This is achieved by the reduction of the reporting channel overhead from cooperative users to the fusion center. Simulation results show that the proposed scheme reduces significantly the signaling overhead, thus making the system more spectrum and energy efficient. Moreover, the openness of cooperative CR makes it susceptible to data falsification attacks, also known as Byzantine attacks. This attack poses a series of damages on the reliability of the estimation of primary traffic. In this paper, we define the types of malicious coordinated attacks on CRs and analyze the possibility of estimating the primary traffic statistics under these attacks. Moreover, we provide a simple yet effective countermeasure based on the proposed reporting for cooperative estimation. Simulation evaluation shows that the proposed algorithm provides an excellent countermeasure for spectrum sensing data falsification attacks.