On the Accuracy of the Gaussian Approximation With Linear Cooperative Spectrum Sensing Over Rician Fading Channels

Linear cooperative spectrum sensing has been proposed for cognitive radio networks which are often limited by the achievable complexity of the devices involved. The method is based on the Gaussian approximation of the summary statistics sent to the fusion center. The goodness of this approximation is investigated in this letter in the case of block Rician fading. The impact of the number of secondary users, sampling times, and the Rician factor are assessed by an analytic method proposed to calculate the false-alarm and the missed-detection probabilities.