A Multiple Antennas Spectrum Sensing Method by Using the Information of Covariance Matrix

In order to reduce the impacts of multiple antennas signals’ frequency offset、random phase and the nonuniformity、uncertainty of noise on spectrum sensing performance, a spectrum sensing method is proposed, whose detection statistics is the ratio of the sum of the squares of the main diagonal elements’ absolute values of the covariance matrix to its mean variance, because the main diagonal elements of the covariance matrix include primary users’ major information, as well as, the mean variance of covariance matrix elements reveals the degree of elements’ fluctuation. The probability density function of the test statistics is derived and a detection threshold with a given false alarm probability is obtained. The simulation results in Additive White Gaussian Noise (AWGN) channel and Rayleigh fading channel show that the proposed algorithm is better than the local variance algorithm. As the number of the receiving antennas and the sampling number increase, the proposed algorithm has a larger performance improvement than the local variance algorithm.