Co-operative Hybrid Blind Spectrum Sensing for Cognitive Radio

The Eigen values obtained from the sample covariance matrix formed from the received signal has the property to entrap noise featuresand signal relations. This property of Eigen values is utilizedin Cognitive Radio for Spectrum Sensing. This kind of Spectrum Sensing is referred as Blind Spectrum Sensing, because it does not require any prior information of primary user signal and channel. This paper presents a co-operative hybrid blind Spectrum Sensing algorithm, which can detect well under lower SNR. This algorithm is anintermingledcombination of Maximum Eigen value to Trace of the sample covariance matrix (MET) and Arithmeticaverage of Eigen values to Geometric average of Eigen values of the sample covariance matrixmethods (AGM). The test signals of both the methods are considered for the test signal of the proposed algorithm. For deriving the analytical expressions for detection probability and threshold, in the proposed method, the Tracy–Widom distribution of maximum Eigenvalue is used. A novel Eigen value fusion is employed at Fusion Center (FC) in this proposed model. The detection performance of the proposed algorithm is effectively improved, and it is verified using simulation results. For simulation, MATLAB software is used.