Research Of Algorithm Against Collusive SSDF Attack Based On Binary Analysis In Cooperative Spectrum Sensing

With the rapid development of wireless communication technology,the problem of scarce spectrum resource has become increasingly prominent and limited spectrum resources is unable to meet the needs of most users.The emergence of cognitive radio(CR)technology has effectively alleviated this problem.Multi-user cooperative spectrum sensing(CSS)technology is often used in cognitive radio networks to improve the accuracy of spectrum detection.However,CSS introduces new security threats.By conducting spectrum sensing data falsification(SSDF)attack,malicious CR users often send falsified sensing data to the fusion center(FC)to disturb data fusion process,which results that the primary user's communication is affected or other secondary users can not effectively use the free spectrum resources.In this thesis,the problems of collusive SSDF attack in cooperative spectrum sensing technology are analyzed in depth.Mining the characteristics of secondary users submitting sensing data in "0-1" mode in hard decision fusion,a collusive SSDF attack defense mechanism based on binary analysis –XDA is designed.The main work and innovations of this paper are as follows:(1)Noting the duality of sensing data,we propose a defense mechanism called XDA from the perspective of binary analysis to suppress a collusive SSDF attack.In the XDA scheme,the XOR distance calculation in line with the type of "0" and "1" historical sensing data is used to measure the similarity between any two secondary users.Noting that collusive malicious users hold high trust value and the minimum XOR distance,the algorithm to detect collusive malicious users is designed.By reducing the trust value of collusive malicious users' in the process of trust evaluation,XDA scheme can enhance the accuracy of decision results of data fusion center.(2)In order to verify the effectiveness of XDA defense mechanism against collusive SSDF attack,the changes of trust values and network trust errors of collusive malicious users under Bayesian trust and XDA scheme are simulated and compared.The results show that the trust value of collusive malicious users in XDA scheme continuous attenuation as the increase of the number of perceived activities,and soon falls below the threshold,resulting in a smaller network trust error.(3)Introducing XDA defense mechanism to improve trust weighting of traditional hard decision criteria such as "AND" 、 "OR" and "Majority",a hard decision data fusion algorithm against collusive SSDF attacks is proposed.Meanwhile,the specific implementation process of the algorithm is given.From the perspective of detection probability and false alarm probability,the ROC operating characteristics curves of the improved "AND" 、 "OR" and "Majority" hard decision criteria of XDA defense mechanism as well as the attack success rate of collusive SSDF attackers in the hard decision data fusion algorithm are simulated and compared.Simulation results show that XDA defense mechanism can make hard decision data fusion algorithm has better anti-collusion SSDF attack characteristics.