DefinePK

Abstract

Spectrum Sensing plays an important role incorporating intelligence into cognitive radio networks. There are many sensing algorithms proposed in the literature evaluating the detection performance. However, most of them failed to operate in noisy environment. Due to inaccurate estimation of noise power, it is difficult to sense the primary user signal in the radio frequency band. In this paper, a detector is proposed based on the power spectral density of entropy computed in the frequency domain. In the proposed work of spectrum sensing, the SNR wall was obtained with an improvement of 5 dB for DVB-T signal compared with the published results. Several Monte Carlo simulations were carried under Rayleigh fading and Rician fading channels with additive white guassian noise. Simulation results show the robustness of the proposed method against the noise uncertainty.