A class of low complexity spectrum sensing algorithms based on statistical covariances

The evaluation of signal detection algorithm involves two aspects: computational complexity and performance. Based on the statistical covariances of the signal, the well-known spectrum sensing algorithm named as maximum-to-minimum ratio eigenvalue (MME) algorithm was proposed in [1]. MME is a blind signal detection algorithm and it has good performance. The main advantage of MME is that it does not related to the noise power. However, due to involving eigenvalue decomposition, MME has a high computational complexity. MME is not the best signal detection algorithm based on statistical covariance matrix. Therefore there may be other algorithm can perform better than MME. In this paper, based on the idea of the approximation of the eigenvalue of the matrix, we proposed three spectrum sensing algorithms with lower complexity. These algorithms are also blind spectrum sensing algorithms, and they are not sensitive to the noise power. Simulation results demonstrate that their performances are better than that of the MME algorithm.