Research On DOA Estimation Algorithm Based On Compressed Sensing

In recent years,the direction of arrival(DOA)estimation theory based on compressed sensing has been developed rapidly.Compared to the traditional DOA estimation algorithm,compressed sensing DOA estimation has more excellent characteristics.In the condition of the low SNR and the small number of samples compressed sensing DOA estimation principle still can get better estimation performance.Further more,compressed sensing DOA estimation principle can deal with coherent signals directly.To solve the off-grid problem in DOA estimation based on compressed sensing,this paper first applies the DOA estimation algorithm based on Bayesian compressed sensing.Then,combining the EIV mismatch optimization model and the fast iterative shrinkage threshold algorithm(FISTA),the off-grid problem in DOA estimation of compressed sensing is solved.On this basis,a M-FISTA algorithm is proposed.A OMP-FISTA algorithm is proposed to speed up the M-FISTA algorithm.Finally,the M-FISTA algorithm is applied to two-dimensional DOA estimation in the real antenna array.Firstly,to solve the off-grid problem in DOA estimation based on compressed sensing,this paper studies the DOA estimation algorithm based on Bayesian compressed sensing.Applying Bayesian compressed sensing theory to DOA estimation,we obtain a BCS-DOA algorithm which can effectively solve the off-grid problem.Then,the EIV mismatch optimization model and FISTA algorithm are combined to solve the off-grid problem in DOA estimation of compressed sensing.In the case of multiple snapshots DOA estimation model,this paper proposes a M-FISTA algorithm.Further more,the OMP-FISTA algorithm is proposed to speed up the efficiency of the M-FISTA algorithm.Finally,for the two-dimensional DOA estimation problem,the M-FISTA algorithm is applied to the two-dimensional DOA estimation in the real antenna array,and the performance of the proposed algorithm is verified by the actual antenna array and the two-dimensional DOA estimation.