Research On SSDF Attack And Defense Schemes Based On Belief Propagation

Cooperative spectrum sensing improves the sensing performance of cognitive radio network?However,in distributed cooperative spectrum sensing networks,the information sent by spectrum sensors(such as mobile phones,tablets,etc.)is often unreliable due to equipment failures,channel shadow fading and noise.Malicious users in collaborative spectrum sensing network also send erroneous perceptions to confuse and interfere with the decisions of honest users,this kind of intentionally sending false sensing information to mislead honest users is called Spectrum Sensing Data Falsification(SSDF)attack.The transmission of unreliable messages or erroneous messages between neighbor nodes will inevitably lead to the deviation and error of the perceived results and greatly reduce the efficiency of cooperative spectrum sensing.In addition,when there are a large number of malicious users in secondary users,malicious users will launch large-scale attacks,which will seriously affect the CSS results.In order to solve the above problems,this thesis expands the SSDF attack and defense scheme in cognitive radio networks,the details are as follows:Firstly,spectrum sensing technology of cognitive radio are introduced.Then,the cooperative spectrum sensing process in centralized scene and distributed scene and the common SSDF attack types and defense algorithms are elaborated.Secondly,the SSDF typical attack model is studied,a defense scheme combining belief propagation algorithm with reputation model is proposed.The scheme filters out unreliable information in cooperative spectrum sensing with two phases: firstly,in spectrum sensing phase,grouping the secondary users according to belief propagation algorithm can filter out the users which are unreliable due to equipment failure and other factors.The remaining users will be regarded as normally working users for data fusion.Then,in data fusion phase,reputation values will be used as weighting factors in the belief propagation algorithm to calculate the final decision value.The scheme proposed in this thesis adopts defensive measures respectively in the spectrum sensing phase and the data fusion phase,which can effectively filter the unreliable information in the network and reduce the impact of the harsh spectrum environment on the secondary users.Simulation results show that the proposed scheme has fewer iterations and faster convergence.And it can effectively reduce the damage caused by SSDF attacks,improve the accuracy of sensing results and enhance the security of cognitive radio networks.Lastly,a user selection and belief propagation based dual defense scheme for large-scale intrusion is proposed,in which the mean of the belief value will be used to calculate the income of the honest behavior in user selection.In spectrum sensing phase,a user selection algorithm is used to pick out reliable secondary users and only the selected secondary users enter CSS.The selected secondary users include honest users and malicious users which pretend to be honest.Then,the belief values of secondary users are calculated by the belief propagation algorithm,and the mean of belief values is calculated with the reliabilities as the weights and compared with the preset threshold to further detect the malicious users.The combination of user selection and belief propagation in this thesis make all secondary users endure double-detection,which will effectively defend against largescale malicious user attacks.Simulation results show that the proposed scheme has better detection and convergence performance than the existing three defense mechanisms of SRS?MS?LCDA.