Research On Sparse Target Direction Of Arrival Estimation Under Complex Conditions

With the acceleration of social progress and the evolution of science and technology,the operating electromagnetic environment and scenarios of radar is becoming increasingly complex.Whether in daily life or military applications,the detection and positioning of space radiation sources by radar is facing increasingly severe challenges.Direction-of-Arrival is one of the key technologies,which is mainly used for direction finding and positioning of airspace targets.Traditional DOA estimation techniques,such as subspace methods,will significantly reduce performance in complex environments that are often faced in modern applications such as small signal-to-noise ratio and small number of snapshots.Modern signal processing technology based on sparse representation and compressed sensing provides ideas for breaking through this dilemma,and it is also a hot research topic in recent years.However,the computational time complexity of this type of method is generally relatively high,which limits its application in scenarios with high real-time requirements.The thesis mainly focuses on the sparse Bayesian learning method based on compressed sensing framework,and studies the robust DOA estimation algorithm in complex environments such as low signal-to-noise ratio,small snapshots and high realtime requirements.The main contents of the paper are as follows:1.In view of the problem of high computational time complexity of the algorithm,a generalized approximate message passing technique is used in the sparse Bayesian algorithm to improve the computational efficiency.Then a fast SBL algorithm based on GAMP was proposed.The algorithm constructs a factor graph model under multiple time-dependent measurement vectors,and converts the high-dimensional joint posterior probability density calculation problem into a scalar form of edge posterior probability density calculation.Compared with the traditional SBL method,it significantly reduces calculation amount and has stronger robustness in complex environments.At the same time,for the sparse arrays that are currently studied,the implementation form of the algorithm under the nested array is deduced,which expands the application range of the algorithm.2.Aiming at the error caused by grid division,a new grid refining method based on gradient is proposed.This method adaptively adjusts the grid according to the gradient direction at the target point,which is suitable for any formation and occupies less computing resources.The simulation experiment shows that under the same conditions,the estimation accuracy is the same as the root finding method,but the calculation time is less.3.Aiming at the problem of algorithm performance degradation in complex environments,a DOA estimation method based on the fusion of different radar data is proposed.This method establishes the connection between different radar observation data through the spatial geometric relationship between the system and the target.The effect is equivalent to increasing the number of snapshots.To a certain extent,it alleviates the lack of snapshots in complex environments and improves the robustness of the algorithm.