Reputation-based collaborative spectrum sensing algorithm in Cognitive Radio networks

In order to resist the attack from malicious Cognitive Radio (CR) users, a reputation-based collaborative spectrum sensing model is proposed in this paper. In this model, each collaborative CR user has a reputation degree, which is set by the central controller and used to weight the sensing result from corresponding CR user in the fusion process at the central controller. We analyze and evaluate the performance of hard decision fusion and soft fusion with the proposed model. And simulation results show that our proposed model weakens the harmful influence caused by malicious CR users and has better detection performance than that of original model both in hard decision fusion and soft fusion.