Efficient Collaborative Spectrum Sensing under the Smart Primary User Emulation Attacker Network

In this paper, collaborative spectrum sensing to detect random signals corrupted by Gaussian noise in the presence of Primary User Emulation Attackers (PUEAs) is studied. We consider smart PUEAs which aims at increasing the false alarm probability and constitute a PUEA network on a Cognitive Radio (CR) network by impersonating Primary Users (PUs). In addition, we propose two security schemes in which sensing nodes get assistance from the Secondary Users (SUs) using two different approaches. In the first approach, the proposed scheme requires having some knowledge about the PUEA network similar to most of the schemes available in the literature. In our second proposed scheme, information about the PUEA network is not required yielding a scheme which is robust to the strategy of attackers. In both proposed approaches, we propose an algorithm to incorporate the SUs assistance in spectrum sensing. The final collaborative decision is made through solution of an optimization problem in order to achieve the best performance and protect the CR predefined requirements. Furthermore, in order to evaluate the performance of the proposed detector at the SUs and at the employed detector in the Fusion Center (FC), the closed form expressions for detection and false alarm probabilities are computed analytically. The provided closed-form analytical results in addition to simulation results show that the proposed schemes significantly outperform the existing secure spectrum sensing schemes.