Research On Direction Of Arrival Estimation Of Compressed Sensing In Nonideal Case

Direction of Arrival(DOA)estimation is an important topic and hotspot of array signal processing.It is widely used in civil and military fields such as radar,broadband communication,wireless sensor positioning,and passive positioning.With the large-scale application of antenna arrays in various application fields,more and more complex application scenarios are faced with the problem of DOA estimation of array signals under non-ideal conditions such as low signal-to-noise ratio,fewer snapshots,and dense target distribution.The DOA estimation method based on compressed sensing theory provides a new idea for solving this problem.However,the traditional compressed sensing algorithm also faces some limitations,such as the high correlation caused by mesh partitioning,and it still needs to be improved in terms of estimation accuracy and complexity.Based on the theory of compressed sensing,this paper conducts research on non-ideal situations such as low signal-to-noise ratio and dense DOA,aiming to improve the performance of the dense DOA estimation algorithm.The research content of this article is as follows:(1)Aiming at the problem of high correlation between fixed grid and off-grid in compressed sensing dense DOA estimation,a movable off-grid model based on off-grid offset parameters is designed.The proposed method continuously adjusts the off-grid offset by iteratively,reducing the correlation of the preset uniform grid and off-grid parameters on the grid position offset in the original off-grid method.This method reduces the correlation between the uneven grids.This method is compared with existing off-grid compressed sensing methods,including off-grid parameter model based on first-order Taylor expansion,off-grid parameter model based on covariance matrix,and off-grid parameter model based on linear interpolation.Through Monte Carlo simulation,the performance of the proposed method is compared with the existing on-grid DOA estimation methods and off-grid DOA estimation methods.(2)In the mobile off-grid model,in order to optimize the iterative process of off-grid parameters,a weighted off-grid sparse Bayesian off-grid parameter method is proposed.This improved method designs weighted vectors based on the different characteristics of the signal and noise feature subspaces,enhances the convergence speed of the optimal iteration,and improves the convergence speed and estimation performance of off-grid compressed sensing DOA estimation.In simulation experiments,this method is compared with the existing off-grid sparse Bayesian algorithms.(3)Aiming at the grid partitioning problem of dense DOA estimation,gridless compressed sensing methods for solving dense DOA parameters are studied.First,a dense DOA estimation method based on the sparse parameter method(SPA)is designed.This method establishes an optimization model by using the covariance matching criterion.This method replaces the original parameters solution of grid compressed sensing by calculating the covariance matrix and DOA parameters.Secondly,a dense DOA estimation method based on continuous compressed sensing(CCS)is designed.This method establishes an optimization model by using the atomic norm principle.This method avoids presetting the grid in the compressed sensing model.Through Monte Carlo simulation,the performance characteristics of these two algorithms in non-ideal situations such as low signal-to-noise ratio and dense distribution of DOA parameters are analyzed.