On the Evaluation of an Entropy-Based Spectrum Sensing Strategy Applied to Cognitive Radio Networks

In this paper, the evaluation of a spectrum sensing strategy based on the frequency domain entropy applied to cognitive radio networks is presented. Entropy estimation is performed using Bartlett periodogram. A tradeoff between variance and the spectral resolution for Bartlett periodogram is presented. This tradeoff affects the probability of detection and false alarm of the spectrum sensing strategy in environments with low signal-to-noise ratio and noise uncertainty. The Entropy detector is optimal when the product of the number of segments and the number of points used is equal to the number of available samples of the received signal.