GLRT based spectrum sensing with oversampling

Abstract

In order to protect primary users from harmful interference, the proposed spectrum sensing scheme in cognitive radio (CR) is required to perform well even in extremely low signal-to-noise ratio (SNR) environments. Meanwhile, the sensing period is usually required to be short enough so that secondary (unlicensed) users can fully utilize the available spectrum. To meet the requirements of spectrum sensing, we proposed a novel generalized likelihood ratio test (GLRT) based spectrum sensing method in this paper. For the proposed spectrum sensing scheme, enough samples can be obtained during a very short sensing period through oversampling technology. Moreover, the time coherence characteristic of the fading channel can be utilized to improve the sensing performance without any prior knowledge about the variance of noise and channel information between primary and secondary users. Extensive performance comparison between the proposed GLRT based detector and other existing detectors, such as the existing GLRT based detectors and the energy detector (ED), has been performed. Simulation results show that the proposed novel GLRT based detector performs much better than the other existing detection algorithms, especially in the low SNR region.