Research On Key Techniques Of Bayesian Compression Sensing Theory In DOA Estimation

Signal DOA(Direction of Arrival)estimation is one of the important research contents of Array Signal Processing.It is based on the theory of spatial spectrum estimation theory and has been widely used in military and civil fields.The classical DOA estimation algorithm requires a priori information of the number of signal sources and a large number of acquisition data for multiple snapshots,which are very difficult to achieve in practical engineering.Bayesian Compressive Sensing provides a new solution to DOA estimation.The theory breaks the limitation of Nyquist sampling theorem and realizes signal reconstruction with mathematical statistics theory.This paper mainly studies the DOA estimation problem based on BCS theory,which can make up for the shortcomings of traditional classical DOA estimation algorithm.Firstly,selecte the narrow-band far-field array signal to estimate its DOA and two DOA estimation models are established.The first one is that the actual incident direction of the signal is on the sampling grid point,and the second one is that the actual incident direction of the signal is not on the sampling grid point.In order to reduce the computational complexity and improve the performance of the algorithm,this paper makes an iterative accelerated improvement on the basis of the two existing DOA estimation model algorithms.And the scenario that the actual direction isn’t on she sampling grid point can be regarded as the expansion of the scenario that the actual direction in on the sampling grid point.In this paper,the two improved DOA estimation algorithms are simulated,and its results was compared are and analyzed with OMP(Orthogonal Matching Pursuit)algorithm,IRLS(Iterative Reweighed Least Squares)algorithm,SBL(Sparse Bayesian Learning)algorithm and SA_SBL(Support Knowlegde-aided Sparse Bayesian Learning)algorithm in terms of observation-array size,number of sources,signal-to-noise ratio and dictionary density.Finally,a DOA estimation model is proposed for the DOA estimation of some sources-known,and the DOA estimation model is established by bring support information in the BCS theory.The G-SBL algorithm is designed by using traditional Bayesian compressed sensing and Gibbs sampling theory,and the principle and derivation of G-SBL(Gibbs Sparse Bayesian Learning)algorithm are discussed in detail.The results show that the G-SBL algorithm has better performance on the DOA estimation problem.