Distribution of observed energy level over medium-traffic primary network using improved energy detector

By considering spectrum-sensing periods of secondary users in a cognitive radio network besides the traffic rate of primary users, it is probable to have a medium-traffic primary network in which the primary user may change its state, either idle or active, at most once during the time period of spectrum sensing, resulting in performance degradation of the secondary network. In this paper, we evaluate the accuracy of approximate methods on the probability distribution of measured energy levels observed by a secondary user under Rayleigh-faded channel, when the traffic rate of the primary network is medium. Numerical results show that the gamma approach outperforms the Gaussian approximation.