Spectrum Sensing for Cognitive Radio using S-method based Joint Time-Frequency Representation

Abstract

Cognitive Radio (CR) has emerged as a solution for spectrum crunch due to rise in large wireless digital ecosystem. Spectrum sensing(SS) is considered to be the brain beyond efficient utilization of spectrum without interfering the licensed Primary User (PU). In this paper, Time-Frequency Tool (TFT) based spectrum sensing is investigated as the novel non-parametric detection method. Here, variant quadratic time-frequency distribution, is obtained from Short time Fourier transform (STFT) called S-method time frequency tool is used for spectrum sensing. S-method is novel T-F distribution that overcomes the drawbacks of quadratic Wigner distribution and obtains energetic distribution. In this work spectrum sensing performance of S-method is compared with well researched Energy Detection (ED) technique in terms of detection probability. This paper also highlights importance of noisy signal analysis in the both time and frequency compared to the time or frequency domain only.