Research Of DOA Estimation Algorithm Based On Constrained Blind Separation

Blind source separation is an important research issue in the field of signal processing.Traditional blind source separation algorithms usually need large amount of computation and the performance of these algorithm degrades under certain separation conditions.The DOA estimation is an important problem in array signal processing.Existing algorithms often cannot achieve the required accuracy requirements in increasingly complex electromagnetic environments and special engineering application scenarios.It is of great theoretical significance and engineering application value to study blind source separation algorithm for special scenes and apply it to special engineering application.This dissertation studies the problem of blind source separation for specific target signals.Based on the prior information in the problem,it improves the traditional algorithm and proposes a blind source separation algorithm based on correlation constraints.By avoiding to recover all source signals,the amount of the computation in proposed algorithm is reduced.Based on above studies,the problem of DOA estimation under strong interference is investigated.Using the oblique projection method,a DOA algorithm based on gradient descent is proposed.The algorithm can accurately estimate the DOA of the target signal in the case where the interference signal is adjacent to the direction of the target signal.The main work in this dissertation includes:1.For the problem of blind source separation under the linear mixed model,the system model of BSS is first studied,and the disadvantages of the traditional blind source separation algorithm based on independent component analysis are discussed.Then the prior information of the target signal is used to construct the constraints to improve the existing algorithm.The feasibility and effectiveness of the proposed method is also analyzed.The performance of the algorithm is studied through simulation experiments.2.In the problem of DOA estimation for the signal under strong adjacent interference,the system model in this scene is first given and the performance of the traditional MUSIC algorithm is analyzed.Then an oblique projection operator is constructed to separate the interference signal and the target signal.The power difference between the interference signal and the target signal,and the correlation characteristics of the target signal are utilized to design a gradient descent algorithm which could jointly achieve the DOA estimations of both the interference signal and the target signal iteratively.The simulation of the proposed algorithm is performed in a uniform linear array scenario to compare with the existing MUSIC algorithm and RELAX algorithm in the accuracy and computational complexity.The gradient descent algorithm based on oblique projection method for the signal under strong adjacent interference has the advantages of high estimation accuracy and low computational cost.The proposed algorithm outperforms the existing algorithms in similar scenarios.