A Probability-Based Spectrum Sensing Scheme for Cognitive Radio

Spectrum sensing in cognitive radio (CR) typically assumes that the primary signal appears before the sensing block. In this paper, we develop a spectrum sensing technique for CR systems where the primary signal may appear at anytime within a sensing block. We first establish the probability model regarding the appearance of the primary signal at each sampling point of a CR user frame. While the conventional spectrum sensing scheme allocates the same weight to each sample, we propose a probability-based energy detection scheme, in which the weight for each sample is based on the probability of the presence of the primary signal at the corresponding sampling point. Numerical results indicate that the probability-based scheme exhibits better performance than the conventional scheme.