Research On Array Signal Parameter Estimation Method Based On Sparse Bayesian Theory

The development of array signal processing started earlier and was applied in the 1930 s.It receives and processes signals by placing different elements in different positions in space and obtains valuable information from the analysis.It has numerous uses in both military and nonmilitary contexts.(such as vehicle positioning,electronic reconnaissance,etc.).Among them,spatial spectrum estimation and source location are the key problems in array signal processing.However,with the development of low interception probability technology and the increasing complexity of electromagnetic environment,array signal processing has met some practical challenges such as low SNR and small sample,and with the gradual increase of information,data collection,processing and storage and other links have put forward higher requirements for hardware equipment.Compressed sensing is proposed under this background.According to this theory,if the signal is sparse or sparse in a certain domain,it can be collected by using the sampling rate far lower than Nyquist and achieve accurate reconstruction of the signal.In recent years,the theory of compressed sensing has provided new ideas and solutions for solving the problems encountered in array signal processing.According to the different reconstruction algorithms,it can be subdivided into three categories,namely greedy algorithm,convex optimization algorithm and probabilistic inference algorithm.Among them,probabilistic inference algorithms represented by sparse Bayes have obvious advantages in reconstruction accuracy and reconstruction speed.Based on a national defense research project,with spatial spectrum estimation and signal source location as the background,this essay aims to improve the performance under low signal-to-noise ratio and small sample conditions by combining sparse Bayesian compressed sensing algorithm with array signal processing.The main work of this essay is as follows:1.In order to improve the accuracy of direction finding estimation using sparse Bayesian algorithm,and to solve the problem that classical subspace fitting algorithm needs multidimensional search,a sparse Bayesian departure wave direction estimation algorithm based on subspace fitting is proposed.Firstly,the covariance matrix of the received signal is solved and the eigenvalue is solved.After the signal subspace is obtained,the equivalent sparse representation model is constructed by using the fitting relation between the signal subspace and the space formed by the guide vector,and the Bayesian algorithm is used to solve it.At the same time,the error caused by the separation is taken into account to further fit the actual scene.The simulation results show that compared with the traditional subspace-like DOA estimation method,the estimation accuracy is higher.2.There are numerous errors in the actual array direction finding system,including amplitude phase and mutual coupling,which seriously impair the performance of array direction finding.In this case,by introducing the spatial sparsity of the signal,the sparse Bayesian reconstruction technology is used to solve the passive correction and the joint estimation of the array signal azimuths when the amplitude-phase and mutual coupling errors exist simultaneously.Firstly,a supercomplete model of the received signal when the error exists is constructed,and the posterior probability density function of the received signal is obtained.The EM algorithm is used to optimize the probability density function iteratively,and the corresponding parameters are solved.At the same time,the CRLB of the array error and signal orientation is derived.Simulation results show that the algorithm can effectively improve the performance of the estimation of the direction of reach when multiple errors exist simultaneously.3.A grid array localization method based on moving single station is proposed to solve the problem of limited accuracy caused by the mismatch between the radiation source position and the initial grid.By moving the array to collect the radiation source signal in different time slots,Multiple location dictionaries are combined,sparse signal reconstruction and mesh adaptive adjustment are realized by sparse Bayesian reconstruction algorithm.First of all,in the region where the radiation source is located,the dictionary is carried out first-order Taylor expansion at the real location of the radiation source in the form of grid division,and the sparse location model is built.Then,the sparse Bayesian algorithm is used to reconstruct the signal,and the location of the radiation source is obtained.The simulation results show that compared with the traditional two-step positioning and off-grid positioning methods,the proposed algorithm can effectively improve the positioning accuracy.