Research On Spectrum Sensing Algorithm Of Cognitive Radio System

With the rapid development of human society, the spectrum resources have been too rareto be used for the new wireless business. However, a large part of the spectrum utilization isvery low. There is an urgent need to develop a new technique to solve the current problem oflow spectrum utilization, so the Cognitive Radio (CR) technology is proposed, which canfind free spectrum dynamically to allocate to other users, and does not interfere the normalcommunication of the primary user, meanwhile. To realize Cognitive Radio, the technologyof spectrum detection is an important precondition.Firstly, this paper described the current research in cognitive radio, and discussed indetail the theoretical model and the advantages and disadvantages of the spectrum detectionalgorithm based on the transmitter and receiver, and focused on the theoretical analysis ofenergy detection. Aimed on the lack of traditional single-user detection technology,multi-user collaboration spectrum detection technology is proposed, and it discussedcooperation spectrum detection fusion rules and the advantages and disadvantages of fusionrules. Also it analysis the theoretical model and actual solution of the EGC detection anddual-threshold detection technology, which pave the way for the following studies.Traditional cooperative spectrum sensing based on majority rule can not only ensure theservice quality of primary user, but also increase the opportunities of cognitive user to accessthe spectrum dynamically, however, it treats the local decisions of all users equally, whichmaybe cause performance degradation if some cognitive node located in low SNRenvironment. To overcome the shortcoming, a cooperative spectrum sensing based onimproved majority rule is proposed, which gives different weights to different cognitive usersaccording to the SNR, and changes the local decisions from the hard decisions to similar softdecisions to reduce the malicious node interference and improve detection performance.Finally, simulation experiments are provided to validate the algorithm and discuss the effectof the specific parameters to the algorithm performance.In addition, this paper presented a cooperative spectrum sensing algorithm based onimproved evidence theory. This algorithm based on the dual-threshold detection model ofenergy detection, and used special conversion function to obtain the basic probabilityassignment (BPA) of the detected energy belonging to different modes, and then used the DS evidence fusion theory to fuse different user information. At last according to some guidelineand the fusion result, it made a final decision. It discussed in detail two structures of theimproved DS evidence fusion theory, which increased the sampling frequency and optimizedthe fusion structure to improve the detection performance, and finally verified the algorithmperformance and the specific parameters’ affect to the algorithm performance by simulationexperiments.