The Wideband Spectrum Sensing Technologies Based On Compressed Sensing

Radio technology has become part of our living life, the spectrum resources available can't meet with the actual demanding. The fixed methods that are used to distribute the spectrum allocations and a low efficiency in the use of the spectrum resources caused great obstacles to the development of radio technology. The Cognitive Radio(CR) technology successfully crossed this obstacle, this technology aims to improve the efficiency of the use of spectrum resources by finding the spectrum hole to share using in this environment.With the gradually realization of the information society, the cognitive radio technology needs to find more potential spectrum band available more quickly and accurately,thus,the wideband spectrum sensing gradually become a hot research field of wireless communications. According to Nyquist sampling theorem, in order to achieve spectrum sensing on the wideband needs a very high sampling rate which have brought a severe test to the sampling overhead and the hardware complexity. The compressed sensing theory using non-correlation measurements to sample the wideband signals at a low speed and restore the original signal from a small observation data, which provide an available way to the development of wideband spectrum sensing technology.In this paper, we study the wideband spectrum sensing technology deeply based on the theory and applications of compressed sensing theory. The main works and results of this paper are as follows:The first chapter describes the background of this research and described the principles,the research process and the academic value of cognitive radio spectrum sensing technology specifically.The second chapter mainly describes the cognitive radio spectrum sensing technology and takes a comparison of the current single-user spectrum sensing algorithms. By analyzing the defects of the single-user spectrum sensing algorithm, we introduces the multi-user cooperative sensing methods with describing of the corresponding decision criterion.The third chapter combines the compressive sensing theory with the low efficiency of the spectrum resources, and analyzes the traditional reconstruction algorithm for wideband spectral sensing, including the iterative greedy algorithms like OMP, SP, ROMP, Co Sa MP algorithms and the1 l norm algorithms like ADM, BPDN and BP algorithms. However,the scarcity of the signals in the radio environments can't be known in advance, in order to solve this problem, this paper analyzes the SAMP(Scarcity Adaptive Matching Pursuit)algorithm, and propose an improved adaptive matching pursuit(IMAP) algorithm based on the SAMP algorithm. The IMAP algorithm doesn't need any, which solves a big problem that the greedy algorithm commonly required informations about the primary users in advance. Simulation results show that IAMP algorithm is superior than the other greedy algorithms in reconstruction capability while maintaining the execution efficiency that can compete with the classic greedy algorithm like Co Sa MP / SP algorithms, the IAMP algorithms is a real-time and efficient algorithms.Chapter IV aims to solve the problem that common single-user spectrum sensing algorithms can't resist the attacks by vicious users. We propose a wideband spectrum sensing model which use the reward-punishment mechanism, this model base on the traditional wideband cooperative spectrum sensing models and the compressive sensing theory. Each sensing users utilize the compressed sensing theory to sample the signals, use the IAMP algorithm reconstruct the signals to acquire decisions which are sent to the fusion center to make final decisions using the reward-punishment mechanism to update the weighting number. The simulation results show that this mechanism can achieve efficient and low consumption wideband spectrum sensing.At the end of the paper, we sums up the current scientific research, pointing out the future research goals.