Study Of The Improved MUSIC Algorithm In DOA Estimation

The DOA(Direction-of-Arrival) estimation is a research hotspot and important research subject of array signal processing, it is also one of the basic task of radar, sonar, communication areas and so on. The main studied topic of spatial spectrum estimation is the DOA of spatial signals. In this paper, it has been introduced that research background, development and status of spatial spectrum estimation. It introduces two traditional algorithms at the same time, and has compared with two algorithms simulations results which point out that latter has a better resolution compared to former. The focal point of this thesis was focus on the improvement of MUSIC algorithm, which extracts the mutual covariance matrix of two matrices,and carries on the corresponding transformation, then the average of its and respective covariance matrix, finally to its characteristic decomposition. Simulation result indicates the improved MUSIC algorithm in this paper can estimate effectively the DOA of not only independent signals but also coherent signals. The simulation has been on other algorithms of DOA estimation which include the MUSIC algorithm based on the subspace decomposition, ESPRIT and farword/back spatial smoothing algorithms, at the same time has carried on the detailed analysis to estimation performance of these algorithms. The result indicates that, the MUSIC spectrum peak enhances correspondingly along with SNR increasing, and the neighboring signal MUSIC spectrum resolution reduces correspondingly along with increasing of the angle;ESPRIT algorithm simulation indicates that the estimation precision of the signal parameter enhances correspondingly along with increasing of SNR and the quick racket number. The incident angle is smaller, and the DOA estimation performance has better;Farword/back spatial smoothing algorithm can estimate effectively the DOA of not only the independent signals but also coherent signals.