Research On Low Overhead Wideband Spectrum Compressed Sensing Methods

With the development of mobile communication technology,the current static spectrum allocation strategy leads to a serious shortage of spectrum utilization.Cognitive radio can improve the utilization of spectrum resources through opportunistic spectrum access.Spectrum sensing is a key technology in cognitive radio.It can locate the idle component in the spectrum while avoiding interference to the primary user.However,the wideband spectrum sensing scheme implemented in the high frequency band has a problem of high sampling rate.Wideband compressed spectrum sensing can randomly sub-sample wideband signals to reduce the sampling rate of secondary users.In wideband compressed spectrum sensing,the signal sparsity parameter is very important.Due to the wideband signal is dynamic,it is of great significance to obtain the sparsity information of the signal in real time to improve the reliability and efficiency of the sensing scheme.The main work of this thesis is as follows1.In the current wideband compressed spectrum sensing,there are problems of unknown signal sparsity and too many iterations of the reconstruction algorithm.Therefore,this thesis proposes a wideband compressed spectrum sensing scheme based on sparsity interval estimation.Firstly,the current classical sparse degree estimation models are deeply analyzed,and a sparse degree interval estimation model based on the precise confidence interval of binomial distribution is presented.Secondly,the sparsity adaptive matching pursuit algorithm is improved by using the estimated sparsity information.Furthermore,based on the sparsity estimation model and the improved reconstruction algorithm,a wideband compressed spectrum sensing scheme is proposed.Finally,the simulation results show that the proposed scheme can accurately estimate the upper and lower bounds of signal sparsity at the same time,improve the efficiency and reliability of spectrum sensing,reduce the number of iterations of the sparsity adaptive matching pursuit algorithm,and accelerate the convergence speed of the algorithm.2.In order to further reduce the sampling rate and reconstruction delay of secondary users,a cooperative wideband compressed spectrum sensing scheme based on supervised learning is proposed.First,by analyzing the existing sparsity learning prediction model,using the sampling vector second norm and the sub-spectrum block length as the feature vectors of the supervised learning algorithm,an adaptive sparsity prediction model is presented.Secondly,based on the adaptive prediction model,a wideband spectrum screening algorithm is proposed.The cooperation between secondary users is used to find the sparse spectrum set to be reconstructed in the wideband spectrum.Then,a cooperative wideband compressed spectrum sensing scheme is proposed based on the prediction model and the screening algorithm.Finally,Simulation results show that the scheme can effectively reduce the sampling rate and spectrum reconstruction delay of secondary users,improve the fitting effect of the prediction model,and enhance the model's adaptability.