The DOA Estimation Of Dense Targets And Robust Beamforming Technique

As the primary aspects of array signal processing, direction of arrival(DOA) estimation and adaptive beamforming technique are widely used through radar, sonar and communication, etc. However, the strong signal will cover the weak signal in the vicinity of strong and weak signals which makes it difficult to give the weak signal an accurate estimation. In addition, array response errors, element position errors and the fact that the desired signals may exit in the sample data will degrade the performance of adaptive beamforming algorithm obviously. This dissertation aims at the problems mentioned above. The main work can be summarized as follows:(1) Take the multiple signal classification(Music) for example, we analyze the typical DOA estimation algorithms for strong and weak signals, involving the RELAX algorithm, the jamming jam method(JJM) and SSMusic algorithm, and discuss their requirements, advantages and disadvantages. Then, reconstruct the pseudo-covariance matrix by arranging the order of eigen-values afresh and propose a new DOA estimation algorithm for strong and weak signals based on pseudo-covariance matrix. The method is simply realizable and has low demands for hardware. Finally, we propose a DOA estimation algorithm based on maximum likelihood(ML) for strong and weak signals which avoids the loss of weak signals induced by the suppression of strong signals. Theoretical analysis and simulation verify the accuracy and effectiveness of the proposed methods.(2) Aim to avoid the signal cancellation caused by presence of desired signal in sample data and resolve the problem of sensitivity to array errors for adaptive beamforming, we study on robust adaptive beamforming algorithms, such as signal eigen-subspace method, diagonal loading algorithm, worse-case based method, sequential quadratic programming method, interference-plus-noise matrix reconstruction method and so on. The dissertation describes the basic ideas and the ways of realization as well as the application conditions for the typical algorithms. Finally, we analyze the effectiveness, advantages and disadvantages through the simulations.