Performance analysis of various eigenvalue-based spectrum sensing algorithms for different types of primary user signals

Abstract

Spectrum sensing plays a very essential role in the implementation of cognitive radio networks. Eigenvalue-based spectrum sensing algorithms have been comprehensively discussed in the literature, for detection of primary user signal in the case of uncertain noise. For detection of signals, the test statistics of these algorithms depend on the eigenvalues of the covariance matrix of the received signal. Eigenvalues generally capture the correlation among the signal samples. In this context, we have examined the sensing performance of various eigenvalue-based spectrum sensing techniques for different types of primary user signals having different levels of correlation. In results, it has been noticed that the sensing performance of the algorithms relies on the type of primary user signal transmitted.