A Spectrum Sensing Method Based on Fusion Features and Online Learning in Time-Varying Environments

Abstract

In practical spectrum sensing scenarios, the parameters of environment are time varying. However, most of the traditional methods are not suitable for such time-varying scenarios. In order to overcome this bottleneck, a novel spectrum sensing approach based on fusion features and online model is developed in this article. In the first step, a 3-D feature extraction method is proposed, which uses the complementarity of signals in time and frequency domains to construct the fusion features. Then the Gaussian mixture model is trained by fusion features. In the second step, in order to improve the sensing performance in time-varying scenarios, an online Gaussian mixture model is proposed, which enhances the spectrum sensing in the time-varying environment. Simulation results show that the proposed algorithm achieves better spectrum sensing performance than the traditional method.