Spectrum Sensing For Cognitive Radios Through Differential Entropy
Spectrum sensing for cognitive radio is challenging. In this letter, a spectrum sensing method based on quintiles of Order-Statistics is proposed. We derive the test statistic and evaluate the performance of the proposed method by Monte Carlo simulations. Simulation results show that order statistics based sensing considerably outperforms both energy detection and anderson darling based sensing in an Additive White Gaussian Noise (AWGN) channel; especially in a lower signal to noise ratio region.