Sensing Data Flasification Attack Defense For Cooperative Spectrum Sensing Schemes In Cognitive Radio Systems

The scarcity of wireless spectrum makes the cognitive radio technology research become increasingly hot. Cognitive users are supposed to be with cognitive ability. Under the premise of not causing interference to authorized users, cognitive user can find and utilize the spectrum hole in the surrounding wireless environment with opportunities, and dynamically access the spare authorized spectrum. To achieve this goal, cognitive users must have the capability of spectrum sensing in the first. The performance of cognitive users' spectrum sensing ability will affect the performance of the whole cognitive radio network. The so-called cooperative spectrum sensing technology combines multiple users to sense the primary user. It improves the performance of cognitive radio network a lot, at the same time, weakens the impact induced by single cognitive user's sensing results. The cognitive users in cooperative spectrum sensing are not required great accuracy as single user spectrum sensing did. Because of the cooperation characteristic, the cooperative spectrum sensing technology more easily suffers the spectrum sensing data falsification (SSDF) attack. Built on the analysis to the cooperative spectrum sensing technology, we research and discuss the SSDF attack problems over cognitive radio network in this thesis.We introduce Bayesian Reputation to reflect cognitive user's service quality to solve the probabilistic SSDF attack problem. Our method changes forgetting factor to re-compute the user's reputation, which through the reputation difference detection. Simulation results show that our method can effectively resist probabilistic SSDF attack which at the price of high computational complexity.In our thesis, we use principal-agent theory to solve SSDF attack problem as well. The primary user's network and cognitive users can be regarded as principal and agents who have principal-agent relationship. To the "OR" rule hard decision fusion, we propose and compute the optimal incentive contract. And above the issue of soft decision fusion, we propose a creative incentive mechanism which combines team incentive and personal incentive. Simulation results indicate that our incentive mechanism can relieve the attack from malicious users who want to access the spare spectrum.