Research On Key Technologies For Compressed Sensing Based Spectrum Sensing

Cognitive Radio is designed for taking full use of the so called "Spectrum Hole", in order to alleviate the dilemma of the need and reserve of Radio Frequency resource. As a key technology in cognitive radio, Spectrum Sensing aims at detecting the presence of spectrum holes accurately and in time. Traditional Spectrum Sensing focuses on narrow-band signal sensing, which have limited the sensing ability of CR. As a result, Wideband Spectrum Sensing has been getting more and more attention.Wideband spectrum sensing operates in a very wide spectrum area, which leads to an extremely high sampling frequency. Such high sampling rate restricts the development and application of wideband spectrum sensing. Compressed Sensing makes it come true to sample wideband signal at a very low rate by exploiting the sparse property of wideband signal. Well, traditional Sub-Nyquist samping needs too much prior knowledge of the signal to realize practical application. So in this paper, we propose two methods to improve the Sub-Nyquist Sampling Technologies used in Wideband Spectrum Sensing:Firstly, we systematically investigate and stdudy the revelant knowledge of Modulated Wideband Converter, and then we implement it on the simulation platform of Matlab.Secondly, as an efficient compressed sampling system, the MWC relies too much of ideal low-pass filters. So we propose a compensating scheme, which can enable the MWC to get good reconstruction results with non-ideal filters. We also propose the optimum reconstruction condition by analyzing and deriving. Simulation results demonstrate the efficiency of our method.Thirdly, MWC system needs too much prior knowledge of the signal when used for wideband spectrum sensing. So we propose two strategies to break the limits brought by signal prior knowledge. In our strategies we decide the sampling parameters by using the statistical information instead of accurate band numbers, and then we hire SASP-MMV algorithm for doing signal reconstruction, Simulation experiments verify the efficiency of our strategies.