Spectrum Sensing Statistics Based-GLRT Algorithm in Cognitive Radio

摘要

In Cognitive Radio, because secondary users are completely unknown to the primary users, a statistics algorithm about spectrum sensing is constructed in the paper, which overcomes the shortcoming that the characteristics of the primary users signal and channel are completely unknown to secondary users. The statistic, based on generalized likelihood ratio test (GLRT), is only calculated by means of the received signal. The unknown parameters of the channel are also obtained based on maximum likelihood estimates (MLE) by making use of the sampling signal covariance matrix. Theoretical analysis shows that we can compute the detection probability of primary users with the covariance matrix eigenvalue and the simulation proves that the statistics has the characteristics of the simple calculation and practicability.