Direction-of-Arrival Estimation Method Based On Sparse Bayesian Learning

Direction-Of-Arrival(DOA) estimation has been an area of great research interest because it plays a very important role in many practical applications. Traditional DOA estimation algorithm requires a lot of sampling data, and higher signal to noise ratio,in addition, poorly in coherent signal processing, Used in some emerging areas has been greatly restricted. Therefore how to use a small amount of measurement data to achieve high resolution DOA estimation algorithms become an emerging research direction. The continuous development of the theory of sparse representation for DOA estimation provides a broad direction. It pointed out that the sampled data which is far lower than the Nyquist sampling theorem needed can be used to recover the original signal exactly when the sparse or compressible signals meet certain conditions.Sparse reconstruction have three categories algorithms:greedy algorithm, convex relaxation algorithms and Bayesian learning algorithms. in which, the Bayesian learning or inference combines the priori knowledge, the distribution of information and sample information of unknown parameters,which from a statistical optimization perspective to achieve sparse reconstruction, easy to understand and showing many advantages, It has become a hot research topic. This paper focuses on the DOA estimation algorithm based on sparse Bayesian learning. The main work and achievements can be summarized as follows:1. Briefly Introduced three sparse reconstruction algorithm and analyzes their characteristics; gives an introduction to narrow-band signal DOA estimation algorithms based on Sparse reconstruction, including L1-SVD algorithm, L1-SRACV algorithm and dimension reduced L1-SRACV algorithm.2.Expounded Bayesian estimation theory and the framework of Bayesian learning algorithm, analyzes the advantages of Bayesian learning algorithms, Finally, proposed a DOA estimation algorithm based on Multi-measurement vector Bayesian framework.By simulation experiments, the results show that the algorithm have a good performance in low snapshot, Also it can be applied to the coherent signal processing.3.Study on the signal model of Distribution signal and Block Sparse Bayesian learning(BSBL) algorithm, And applied BSBL to distributed source DOA estimation. Finally, analyzed the estimated performance by simulation test.