Research On Algorithms Of Narrow-band Signal DOA Estimation Based On Compressed Sensing Theory

DOA（Direction of arrival）estimation of array signal is a key technology of signalprocessing, it is widely used in military and civil field. The parameter of array signalestimation based on the theory of sparse reconstruction has the features of high resolution.The sparse reconstruction algorithm has natural decorrelation ability with its specialreconstruction approach.The narrowband signal DOA estimation based on Compressed Sensing (CS) is studiedin this paper. The basic theory of compress sensing and array signal DOA estimation isintroduced in this paper too.It is proved feasible to use the Compressed Sensing theoryestimate DOA of array signal.The paper studies three main aspects: space meshing methoh,the DOA estimation based on sparse reconstruction algorithm and the joint sparsereconstruction algorithm based on subspace theory.Firstly, when use compressed sensing for DOA estimation, array manifold matrix actsas measurement matrix of CS model. Considering the MIP criterion，the quality of differentspace grid division is analyzed. Then, three different kinds of sparse reconstructionalgorithm based on Compressed Sensing is used to array signal DOA estimation.Secondly, based on SOMP algorithm DOA estimation is analysed and studied in thispaper. A new approach based on high order cumulant is presented in order to reduce theinfluence of Gaussian color noise.The simulation results show that, the new algorithm canefficient suppress additional Gaussian color noise.Thirdly, with the number of snapshots increasing, the computational will be increasedtoo. In this paper proposes a new SOMP algorithm based on SVD decomposition, calledSSOMP. The simulation results show that, the new algorithm helps to improve the stabilityproperty of the estimator and reduces calculation amount. The sparse reconstructionalgorithm is not sensitive to “coherent source”, and the MUSIC algorithm can gets very high estimation precision when it obtaines complete subspace information.The SA-MSUICalgorithm is the bridge between MUSIC algorithm and sparse reconstruction algorithm, itapplied to DOAestimation is studied in this paper which is combined the advantage of twokind’s algorithm. The simulation results show that, the SA-MUSIC algorithm is apply to“coherent source” problem within the sensor array signal processing context, TheSA-MUSIC algorithm promotes the SOMP algorithm stability as well.