An Adaptive Wavelet-based Scale Space Filtering Algorithm for Spectrum Sensing in Cognitive Radio

摘要

This paper introduces a novel application of an enhanced Wavelet-based Scale Space Filtering (WSSF) algorithm called Adaptive WSSF (AWSSF). The AWSSF concept was conceived to improve Spectrum Sensing (SS) in Cognitive Radio (CR). The algorithm is based on a novel adaptation of the WSSF and Otsu's algorithm (from Image Processing). The AWSSF decomposes the estimated signals into different scale levels by using Wavelet Transformation (WT) theory. Thereafter, it directly multiplies samples from adjacent scales towards reducing the noise samples, while simultaneously increasing the true Licensed User (LU) signals. Furthermore, we adapted Otsu's multi-threshold algorithm for use in the AWSSF to compute the optimum threshold value for the different decomposition levels towards filtering the wavelet coefficients. During evaluation in the low Signal to Noise Ratio (SNR) region, the AWSSF algorithm was compared to the traditional ED, and shown to perform better. We also compared with other WT based approaches at SNR = -10dB, and the AWSSF achieved better results. The AWSSF met the performance requirement of the IEEE 802.22 standard as compared to other approaches, and thus considered viable for application in CR.