Robust Spectrum Sensing Algorithm Based on Generalized Similarity Detection

The rapid development of wireless communication technology has led to the existing fixed spectrum allocation methods difficult to meet the increasing service demand. Spectrum sensing (SS) technology enables the second users to dynamically access a shared spectrum, which can maximize the utilization rate of spectrum resources. In this paper, a spectrum sensing algorithm based on generalized similarity detection is proposed to address the problem that the performance of traditional spectrum sensing algorithms degrades seriously in the impulsive noise environment. By analyzing the influence of impulse noise, a generalized similarity detector (SD) is constructed based on the difference between the received data when the primary user signal is present or not. Then, a signal detector is developed according to the SD. Theoretical analysis indicates that the proposed SD can significantly weaken the bad effect of impulsive noise and improve sensing accuracy. In addition, the probability distribution function of the test statistic constructed by the similarity detector is presented based on the U-statistical theory. Accordingly, the false alarm probability, the detection probability, and the detection threshold corresponding to a predetermined false alarm probability are analyzed. Finally, experimental simulations are carried out to confirm the good detection performance and the effectiveness of the proposed spectrum sensing algorithm.