Research Of Spectrum Sensing Security Algorithm In Cognitive Radio Networks

As a promising wireless communication technology,cognitive radio effectively improves spectrum utilization via employing dynamic spectrum allocation strategy and enable secondary users to utilize licensed spectrum resources without interfering the communication of authorized users.Since wireless electromagnetic environment possess open characteristic,the precondition for the realization of cognitive radio—the harmful effect exerted by malicious users lead to great reduction of communication reliability.This paper studies the detection and defensing mechanism concerning spectrum sensing data falsification attack in sensing security problems from the following aspects.First,proposing a comprehensive reputation-based secure collaborative spectrum sensing algorithm.The fusion center assigns reputation to secondary users according to their current and historical sensing behaviors in order to evaluate their sensing reliability.Meanwhile,presenting a punishment strategy to improve comprehensive reputation,introducing reward factor and penalty factor to motivate nodes to report correct sensing data.The main target of the algorithm is to mitigate the harmful influence exerted by malicious users and ensure the correctness of global decision.The simulation results prove that the algorithm can achieve good sensing performance even when there exist high proportion malicious users which all adopt complex and dynamic attack strategy.Second,proposing a collaborative spectrum sensing algorithm based on evidential reasoning and credibility calculation,including four consecutive procedures: basic probability assignment(BPA),holistic credibility calculation,option and amelioration for BPA and evidence combination via reasoning rule.The transmitting bandwidth has been decreased by introducing projection approximation approach to modify local BPA masses.Through evaluating the credibility of SUs from both real-time difference and statistical sensing behavior,malicious SUs can be effectively distinguished from genuine ones.Abundant simulation experiments verify that,under the scenario of high proportion of malicious users adopt multiple complex agressire modes,the performance of the algorithm performs far better than the existing methods.Finally,proposing a dual-layer defense-based cooperative sensing algorithm to protect networks from malicious attacks,including credibility-based first-layer defense and reputationbased second-layer defense.In the first-layer defense,the credibility degree is assigned to SUs within clusters to evaluate their sensing reliability,SUs with low credibility cannot get involved in the intra-cluster level evidence combination,and thus malicious SUs are unable to cause harmful effects on collaborative sensing.In the second-layer defense,cluster heads are assigned with reputation value to assess the trustworthiness of their intra-cluster level reports.Cluster heads with low reputation cannot participate in the final center data mergence.The whole mechanism focuses on alleviating the harmful effect of malicious attack behaviors on network decisions and ensuring the correctness of the global decision continuously.Simulation results validate the effectiveness and robustness of the proposed security mechanism.