Adaptive Null Spatial Filter Design And DOA Estimation Based On Convex Optimization

Due to the complex and varied marine environment,the detection and identification of underwater targets has always been one of the important problems in underwater acoustic signal processing.Obtaining high signal-to-noise ratio target echo data in a complex marine environment is an urgent problem that needs to be solved to improve underwater target detection and recognition capabilities.In this paper,we focus on the underwater target detection problem,and study the interference and noise of the underwater array receiving signal and the target data of high SNR.The idea of this paper is based on the array signal processing method;the influence of interference and the noise of the interference region are reduced by the spatial matrix filter.Then,through the robust adaptive nulling widening beamforming algorithm,suppressing the strong interference near the target signal,and the signal-to-noise ratio of the target signal is improved by realizing the secondary suppression of interference and noise in the array region and the beam region.In this paper,the target echo signal form involved in the sonar detection target is introduced and summarized in the signal model and signal characteristics.And also introduce several common target azimuth estimation methods,and two spatial matrix filter designs are studied.According to the design criteria of two kinds of spatial matrix filters are studied,and give the corresponding solving methods.It can be seen from the simulation that the spatial matrix filter can make the signal of the direction of interest pass through without distortion,and suppress the interference and noise of other directions,thereby reducing the SNR threshold for depart signals,and at the same time,the MUSIC direction estimation method using pre-filtering algorithm is compared with the conventional MUSIC direction estimation algorithm,and the parameters such as resolution and low SNR Estimated effect are obviously improved.Secondly,in order to balance the pass region and stop region performance of the spatial matrix filter,the signal in the pass region is processed by the optimal weight vector adaptive beamforming algorithm,combined with the depth-adjustable nulling widening method to make the pass region the interference closer to the target directions forms a deeper null,which effectively reduces the impact on the target directions estimate.For the adaptive method in the paper,the constrained optimization problem is transformed into a second-order cone programming problem with double objective function by regularization transformation,whichis more convenient to solve.Aiming at the problem that the traditional null widening algorithm will lead to null shallow while widening the nulling,a method of expanding the spatial spectral function with Hanning window to widen the null is proposed.By weighting the spatial spectral function,you can specify the depth of null.At the same time,the robustness of the algorithm is analyzed by output SINR.The simulation results show that the proposed algorithm has good suppression ability for interference,and the algorithm shows good robustness at different input SNR.Finally,the method of water-suspended target echo in Lake is used to verify the method in this paper.Under different experimental conditions and experimental parameters,the spatial filtering method and the robust optimal weight vector adaptive beamforming algorithm are used for data processing.Compared with the conventional beamforming and high resolution algorithm processing effects,the effectiveness and reliability of the algorithm in practical applications are examined.