Research On Spectrum Sensing Of Cognitive Radio

As an effective way to solve the shortage of spectrum, cognitive radio technology is being paid on more and more attention. And the spectrum sensing technology as the key technology of cognitive radio has become a hot research topic. Spectrum sensing is the procedure of the cognitive radio terminal detecting the signals through a variety of sensing technology within a given frequency band, and manage the spectrum according to spectrum sensing situation. The cooperative spectrum sensing technology can improve the inference accuracy and sensitivity using the location diversity of cognitive users.This paper introduces existing single point spectrum sensing algorithms, studies the cyclic spectrum of modulation signal, and proposes a spectrum sensing algorithm based on cyclic spectrum section. Modulation signals have peak value at zero frequency section, and the noise was no significant peak in the section, we can use this characteristic to distinguish the noise and modulation signal. By calculating the ratio of maximum and average signal to determine whether exist signals. Simulation results show that the proposed algorithm relative to the existing spectrum sensing algorithm has better anti-noise performance and low computational complexity.The cyclic spectrum of common modulation signals contains a lot of information, to take full advantage of perceptual information and conducive to the subsequent spectrum management of cognitive radio system. Based on summarizing and analyzing the cyclic spectrum section of signal, for a single signal we propose the algorithm which based on cyclic spectrum to estimate the carrier frequency and symbol period and modulation. The simulation shows that compared with existing algorithms, the algorithm has higher accuracy. Aiming at time-frequency overlapped two-signals in single-channel, this paper proposed a method to estimate carrier frequency based on cyclic autocorrelation function, compared with similar algorithms, the proposed algorithm does not require signal separation, and also has high accuracy.This paper summarizes the existing cooperative spectrum sensing algorithms, studies cognitive radio system may exists the depth fading users and malicious users, and proposes a cooperative spectrum sensing algorithm based on the historical performance. The credibility is used to measure the historical performance of each user, and it is also used to weight the sensing result. Use this algorithms, the impact of the depth fading users and malicious users on the detection performance is effectively alleviated. NiosⅡsoft-core are used widely because of its flexibility and configurability, in this paper we build cooperative spectrum sensing simulation system with the NiosⅡsoft-core on DE2 platform, and simulate the proposed cooperative spectrum sensing algorithm and analysis results.