Research Of Cognitive Radio Network Security Key Technology

Cognitive radio using opportunistic spectrum access realizes dynamic use of the idle spectrum. It doesn’t ease the tension of spectrum resource but also improve the utilization of existing spectrum resources. Although, cognitive radio may be as a possible development direction of wireless communication. It brings in some new security risks. These potential security risks are mainly concentrated in the physical layer and MAC layer. This paper will focus on two issues of the two layers.At the physical layer, the main security threat to cognitive radio is primary user emulation attack. MAC layer security threat is spectrum sensing data falsification. The two threats greatly affect the normal work of the cognitive radio technology, the paper will focus on these two security problem.When solving the PUE attack, we first analyze some of the current solution, and according to the principle of development of cognitive radio and the most likely to be application scenario. Adopt the transmitter location authentication solution, which is proposed by scholar Ruiliang Chen and Jung-Min Park. The solution distinguishes PUE attack using comparing the location of the transmitter to the primer user’s. We adopt RSSI-based improved robust maximum likelihood localization algorithm. The algorithm improves the calculation principle of the traditional ML estimation, and using the coordinate correction algorithms modify coordinate, finally using robust ML estimation to amend the revised initial estimate value. Simulation result shows the algorithm can improve the positioning performance and reduce gross error.When solving spectrum sensing data falsification. The paper analyzes conventional algorithm of cooperative spectrum sensing, then analyze the scene that with the presence of malicious nodes. Finally, we adopt the weight-based cooperative detection to solve the spectrum sensing data falsification. The principle of this algorithm is to assign a reputation value to each node of cooperative detection. And update the reputation value and weight value by reward and punishment mechanism. Then weighted the sensing data of each node, and send it to fusion center. The algorithm is simple with low computational complexity, Simulation results show that the algorithm can effectively resist spectrum sensing data falsification during the cooperative detection.