Unified Sensing Algorithm : a Smart Full Exploitation of Detection Methods

• Published in: 2019 International Conference on Wireless and Mobile Computing, Networking and Communications (WiMob) pp. 1 - 6
• Main Authors: Bouallegue, Kais, Baudais, Jean-Yves, Crussiere, Matthieu
• Format: Conference Proceeding
• Language: English
• Published: IEEE 01.10.2019
• Subjects: Clustering algorithms; Cognitive radio; Covariance matrices; covariance matrix; Detectors; eigenvalue decomposition; Eigenvalues and eigenfunctions; Maximum likelihood estimation; Signal to noise ratio; spectrum sensing
• ISSN: 2160-4894
• DOI: 10.1109/WiMOB.2019.8923570

In this paper, we propose a new spectrum sensing (SS) algorithm called unified sensing algorithm (USA). This method exploits the results of different algorithms developed in the literature. The considered methods are especially those based on the covariance matrix of the received signal and those which further compute the eigenvalue decomposition of such matrix. More precisely, the USA algorithm implements some covariance matrix based methods, and in the case where the primary user is miss-detected, the eigenvalue decomposition is performed. This approach fully exploits all detection methods and avoids computational costs, as eigenvalue decomposition, when it is not necessary. Finally, the performance of the proposed method is provided and compared with existing state of the art algorithms. Simulation results show the performance improvement obtained with this new approach, namely the contribution in term of detection under very low signal-to-noise ratio (SNR).