Enhanced Energy Detectors Utilizing Wavelet and RLS De-Noising Filters in Cognitive Radio

Abstract

Spectrum resources are limited due to the fast developing technology in wireless communications. Various techniques, for instance, tackled this problem by giving permissions to unlicensed users to utilize the various licensed bands. Spectrum sensing is one of the most utilized technique in cognitive radio system, which has a reliable technique called energy detection. It has reduced the computational and complexities of usage. In order to differentiate between the primary and secondary users in low signal-to-noise ratio (SNR), noise interference is eliminated using denoising filters. In this paper, a new energy detection technique for spectrum sensing is developed. Besides that, a comparison between common de-noising filters is introduced, which are Recursive Least Square (RLS), 1-D and 2-D wavelet de-noising filters. The performance of the wavelet packet transform algorithm is analyzed under many Signal to Noise Ratios, different number of samples, probability of false alarm levels, and number of samples. Moreover, the technique is analyzed for different level of decomposition and different wavelet families. Simulation results revealed that utilizing RLS de-noising filter outperforms the technique used by wavelet de-noising filters under all analyzed circumstances.