DOA Estimation Of UWB Antenna Array Based On Sparse Bayesian Learning

Direction of arrival(DOA)estimation is one of the important researches in array signal processing.It has a wide range of applications in communications,positioning and radar,etc.Because the common DOA estimation methods can hardly achieve higher estimation accuracy in the environment of lower SNR and lessen snapshots,and the performance of coherent signals is influenced,which limits the further development of DOA estimation.With the development of the theory of compressed sensing and sparse reconstruction,new solutions to the DOA estimation problem has emerged.Under certain conditions,the sparse signal can accurately recover the original signal with the sampled data obtained far below the Nyquist sampling rate.New research has been developed on the research of DOA estimation based on sparse reconstruction.The DOA estimation method based on sparse reconstruction theory mainly includes greedy algorithm,convex optimization algorithm and sparse Bayesian learning algorithm.The former two can be explained from the perspective of Bayesian statistical optimization,using sample data and prior information.Inferring the sparse parameters to achieve sparse reconstruction shows a relatively obvious advantage.Based on the sparse Bayesian learning theory,the main research work is as follows:1.The signal model in array signal processing and the commonly used DOA estimation method are introduced.The sparse signal model in sparse reconstruction theory is introduced.Several estimation methods commonly used in DOA estimation based on sparse reconstruction are introduced,and their respective methods are analyzed.Advantages and limitations,mainly introduces a narrow-band DOA estimation algorithm based on sparse Bayesian learning algorithm.2.For the off-grid problem of signal arrival,the off-grid parameter-oriented vector model is introduced.The Taylor series expansion method is used to introduce additional parameters to represent the quantization error,and the error is used as the estimated parameter.A sparse reconstruction method based on variational Bayesian inference expectation maximization is used to sparsely estimate the established parameter error.The proposed method can improve the outlier error and improve the spatial resolution,and can achieve high precision DOA estimation.3.In order to learn the sparse reconstruction of UWB signals,the normal wideband DOA estimation method is introduced.After analyzing its limitations,an ultra-wide-band DOA estimation method based on multi-bands joint block sparse Bayesian is proposed and the ultra-wideband signal model is established.Considering the frequency characteristics of UWB signals,the signal joint probability density function is constructed.The joint sparse property and the correlation information between different narrow-bands are used to solve the maximum posterior probability density.In order to reduce the amount of the algorithm operation,the fast sparse joint probability density is used to reduce the sampling rate and processing requirements,simplifying the estimation process.