Spectrum Sensing Security In Cognitive Radio Networks

With the rapid development of wireless communication technology,the limited spectrum resources can not satisfy the growing demand of wireless business.Research shows that,the existing static spectrum allocation scheme makes some spectrum resources carry too much business,but most of them are in the idle state,which causes great waste.As an effective technique to ease the contradiction among limited spectrum resource,increasing spectrum requirement and low frequency spectrum utilization,cognitive radio technology has attracted widely attention.Besides traditional security problems,some new kinds of potential safety hazard have been appeared in cognitive radio network because of the openness of the wireless channel.Malicious users rob spectrum resources of cognitive radio network or risk into network by various means of attack.Many researches are focused on the security problems as the increased application value of cognitive radio network.As spectrum sensing is the key core technology in cognitive radio network,it has important significance to study on the safety problems associated with spectrum sensing.On the basis of the principle of spectrum sensing,this paper focuses on its related potential safety hazards.The feasible defense schemes,such as security cooperation spectrum sensing sheme,signal characteristic detection sheme,signal source localization sheme,are put forward.Specific details go as follows:Firstly,the influence factors of cooperative spectrum sensing performance are analyzed.A security cooperative spectrum sensing sheme based on secondary users screening is presented.In order to maximize the detection probability of cooperative spectrum sensing,the sheme filters secondary users through their detection probabilities.To overcome spectrum sensing data falsification attack,an improved security cooperative spectrum sensing sheme is proposed.Fusion center weights the local decisions of secondary users to improve the performance of spectrum sensing and ensure the safety of cognitive radio network.Secondly,according to fractal theory,the defect of fractal box dimension on the recognition of signal modulation mode is analyzed.A new modulation signal identification scheme based on the instantaneous characteristic parameters is proposed,and low computational efforts and simple implement are obtained with the new instantaneous characteristic parameter.However,this new parameter has blind area for identification.This paper puts forward a new detection sheme by combining with box dimension and the new instantaneous parameter.Experiment proves that this method has a high probability to recognize the modulation type of signals even at a low signal noise ratio.It not only can realize primary user emulation attack defense,but also can defend against primary user concealment attack when be used in spectrum sensing stage.Thirdly,the target localization algorithms in wireless sensor networks are introduced to the source localization in cognitive radio network.Aiming at primary user emulation attack,a new source localization algorithm based on compressive sensing is presented.Secondray users regard the signal strength of source received as source fingerprints,and fusion center obtains,which can be converted to the relevance of geographical position between secondray users and signal source,by using compressive sensing technology.The relevance will be weighted to centroid algorithm to estimate the location of signal source.Finally,the sparse target localization algorithm based on orth is analyzed.Although the orth preprocessing make the measurement dictionary meet restricted isometry property,it changes the sparsity of original singal.Based on this,a novel compressive sensing localization algorithm based on singular value decomposition is proposed.In this sheme,the sparsity of original singal remains the same while the measurement dictionary meets restricted isometry property.Therefore,the reconstruction performance can be guaranteed and the localization accuracy is improved.