The Key Technology Of Wideband Compressed Spectrum Sensing In Cognitive Radio Network

Cognitive radio (CR), a key technology in the next-generation wireless network, is an effective way to solve the shortage of spectrum, realizing dynamic management of spectrum and enhancing the usage of spectrum. CR technology has a good ability to perceive the surrounding environment and learn by itself, as well as access the spectrum adaptively. As a result, CR realizes spectrum sharing and management of spectrum. As a key technology of CR, spectrum sensing has become a hot research topic in recent years because of its fundamental role.If we apply the traditional narrow-band spectrum sensing technology which can detect only one channel at a time to detect a wide band that is divided into several sub-band, the detection efficiency will be too low, obviously. In order to improve the detection efficiency, the wideband spectrum sensing technology comes into being. For a long time, Nyquist sampling theorem has been dominated the signal sampling and coding procedures of narrow-band spectrum sensing. While in the sensing process of wide band, the usage of Nyquist sampling theorem is bound to bring severe tests to the digital signal processor. Fortunately, the compressed sensing (CS) theory brings the gospel to it. CS theory breaks through the limitation of Nyquist sampling theorem and puts forward a completely new method of signal acquisition. And then, distributed compressed sensing (DCS) that develops the compression and sampling of signals from single-signal to multi-signal comes up to expand the application of compressed sensing theory further.This paper does some researches on wideband spectrum sensing technology and applies CS theory on it and then proposes two wideband spectrum compressive sensing models for multiple cooperative users. The specific works are as follows:Firstly, this paper has an overview of its research background and current development status of spectrum sensing technology in cognitive radio. In addition, it briefly introduces several common single user spectrum sensing algorithms. Then, after analyzing the limitations of single user spectrum sensing technology, the paper launches a statement of cooperative spectrum sensing technology. The comprehensive comparison and simulation analysis of fusion schemes are done as well.Secondly, this paper introduces and narrates the theory of CS from three aspects: the sparse representation of signal, coding and measurement as well as reconstruction of signal. Then, the existing CS reconstruction algorithms are classified and compared; matching pursuit (MP) series algorithms are well studied, especially.Finally, by doing deep search on wideband compressed spectrum sensing key technology in accordance with wideband spectrum sensing and cooperative spectrum sensing applying CS theory to CR networks, this paper puts forward two cooperative spectrum sensing models respectively based on CS and DCS in CR networks. Furthermore, the performances of algorithms based on two models are investigated in simulations.