Multiband Spectrum Sensing via Edge Detection Using a Wavelet Approach

Abstract

In cognitive radio (CR), the sensed aggregate bandwidth could be as large as several GHz. This is especially challenging if the bandwidths and central frequencies of the sensed signals are unknown and need to be detected. This work discusses a new method for multi-band spectrum-sensing based in edge detection. The proposed method uses a Welch power-spectrum-density (PSd) estimate and a multi-scale Wavelet approach to reveal the spectrum transition (edges), which is deployed to characterize the spectrum occupancy in CR scenarios where the operating frequency limits of the primary users are unknown. The focus in this work is to improve the performance of the multiband spectrum sensor by refining the edge location and error correcting misleading detections. In order to do so, a comprehensive analytical description and numerical analysis have been carried out by focusing on orthogonal-frequencydivision-multiplexing (OFDM) signals. Also, numerical results corroborate and give support to the effectiveness of the proposed multiband spectrum sensing method.