Broadband Spectrum Detection Technology Oriented To Compressed Sensing

Cognitive radio(CR) that can solve the spectrum shortage by enhancing the usage of spectrum is a key technology in the wireless network. CR technology access the spectrum adaptively by perceiving the surrounding environment and realizes spectrum sharing. As a fundamental technology of CR, spectrum sensing has become a hot research topic in recent years.The main content of this thesis is broadband spectrum detection technology oriented to compressed sensing. First, recovery algorithm for compressed sensing is studied and improved algorithms are proposed, Then compressed sensing theory is introduced into the broadband spectrum sensing according to its feature, Also the data fusion algorithm is analyzed and studied. Details are as follows:Firstly, The research background and current development status of spectrum sensing technology is introduced, and its model, performance indicators and common spectrum sensing methods are also given. The simulation by energy detection algorithm points out the problem of single user detection. Then the model, fusion rule is given of the multi-user cooperative spectrum sensing, soft fusion and hard fusion are compared by simulation analysis.Secondly, The theory of CS is introduced from three aspects: the sparse representation of signal, coding and measurement as well as reconstruction of signal, and mainly focused on the CS recovery algorithm. SL0 algorithm is fast, but it convergences slowly and the accuracy is not high. Focused on the defects of SL0 algorithm, an improved algorithm MSL0 is given. First, the standard Gaussian function is replaced by hyperbolic tangent function that can better fit, the initial point is obtained by steepest descent method, Then we use Newton method to search. Simulation results show that compared with the original algorithm, the improved algorithm has faster convergence speed and higher reconstruction accuracy. The performance of IHT depends much on the start point. In order to overcome the defect of IHT, backtracking is introduced into it so that the support set is optimized for each iteration, and also reduce the situation of certain atoms being repeatedly selected and discarded. it also converges the idea of NSL0, searching along the newton direction at each iteration to make sure that we can find the best solution. Through the simulation analysis,we can see that the reconstruction precision of the improved method is improved largely.In this thesis, some typical compressed sensing reconstruction algorithms are applied to the broadband spectrum detection in cognitive radio, and broadband spectrum detection technology oriented to compressed sensing is realized. Through the analysis of the fusion process of cooperative spectrum detection based on energy, The original method is to integrate all the CR users. However, in actual wireless communication environment, not all the CR users are very good, for this part CR users, their integration can't bring enhance the overall detecting accuracy. Adaptive mode to select the users is used in this thesis. Also the detection energy are fused by weight to adapt to the changes in the channel conditions. The simulation results show that the improved algorithm can improve the detection probability,since the amount of data in fusion center is reduced, so the time performance is also improved.