Performance analysis of eigenvalue based spectrum sensing under frequency selective channels

Detecting primary signals with very low Signal-to-Noise Ratio (SNR) is a very important problem in cognitive radio (CR) systems. Small parameter uncertainties are unavoidable in any practical system, and especially the noise variance uncertainty has great effect on the performance of the most basic spectrum sensing method, energy detection. This has motivated the need for advanced spectrum sensing algorithms, like eigenvalue based spectrum sensing, which can be used to overcome the effects of parameter uncertainty in very low SNR cases. In this paper we study the effects of channel frequency selectivity, in combination with noise uncertainty, in case of energy detector and eigenvalue based spectrum sensing. It is demonstrated that channel frequency selectivity significantly enhances the performance of eigenvalue based spectrum sensing techniques.