Energy detection performance enhancement for cognitive radio using noise processing approach

The performance of energy detection (ED) technique is highly affected by the noise fluctuation which leads to degrade the cognitive radio (CR) system accuracy. Therefore, many methods have been proposed and designed to enhance the CR system performance. In this paper, an improvement for ED performance is achieved by using digital signal processing (DSP) for the primary user (PU) signal before taking the final decision by the secondary user (SU). Firstly, the uncertain noise power will be estimated (signal to noise ratio (SNR) estimation) by the SU to find the quantity of the noise that is affected on the PU signal. Secondly, signal denoising technique will be applied on the noisy PU signal to reduce the noise effect on ED technique accuracy. Consequently, hybrid slantlet transform (HST) has been used to implement these two proposed solutions of noise impact alleviation. Moreover, HST is a set of digital filters which separate the noise from the non-stationary PU signals. Simulation results compare the detection performance of the proposed approach with wavelet approach based signal denoising technique which was used in ED before. Furthermore, the results show the obvious superiority for our proposed approach in comparison with wavelet approach in the detection probability versus the SNR and false alarm probability for different number of PU signal samples.