Secure Cooperative Multi-channel Spectrum Sensing in Cognitive Radio Networks

In this paper, two new counterattacks are proposed to combat Byzantine attacks comprising coalition head and cognitive radio as attackers which target to reduce the number of available channels for sensing in distributed multi-channel cooperative spectrum sensing. In the proposed counterattack for coalition head attack, by using statistical properties of the exchanged SNRs in coalitions, the probability of attack is derived and a new selection formula for coalition head is proposed. In the second proposed counterattack for multi-channel Byzantine attack, the probability of available channel detection (the counterpart of false alarm probability) in the presence of Byzantine attackers is first formulated. Since in practice the coalition heads have no information of the presence of Byzantine attacks, predicting such information is a very challenging issue when the attack takes place. Then, the probability of each cognitive radio changing its local decisions and become Byzantine attacker, called probability of attack, is derived. By applying such probability in the probability of available channel detection, an iterative algorithm is proposed. Based on the actions of the cognitive radios, after each round of the algorithm, the probability of attack is updated. If Byzantine attackers continue attacking the system, their contribution to their associated coalitions decreases and they will be blocked out of coalitions. Simulation results show that the proposed counterattacks can remarkably eliminate Byzantine attackers in the cognitive radio network.