Distributed Source Doa Estimation Algorithm

The phenomenon of distributed sources exists widely in practical applications, such as the local scattering sources in the multipaths environment of cellular mobile communications, the sea return echo sources in the low angle radar systems for target tracking, the scattering sources in the troposphere or ionosphere radio transmissions. Furthermore, the direction-of-arrival estimation for distributed sources is also applied in the passive radar systems, the sonar systems and the location systems for sound sources, etc. In the environment of distributed sources, the classical direction-of-arrival estimation algorithms for point sources may result in serious performance degradation, even lead to erroneous results. The distributed sources result from too many multipaths which are induced by reflection, scattering, diffraction and refraction from the objects in complex environment. And the phenomenon that the sources arrive at the array antennas in a continual angular extension will cause the coherent loss of wavefronts arriving at the array antennas. Therefore it's necessary to construct the effective distributed sources models and develop the high resolution direction-of-arrival estimation algorithms for distributed sources.In this dissertation we review the classical models and algorithms of the distributed sources, and research on the relation between the signal subspace and the array number, the DOA of the distributed source, the angle spread, and the angle distribution functions. We also research on the impact on the performances of the parameter estimation algorithms which is caused by the difference of the array number, the SNR, the DOA of the distributed source, and the angle spread. We apply the beamforming technique in the parameter estimation of the distributed sources, and present a robust beamforming algorithm for distributed sources. For suppressing the influences of the steering vector error and the interfered signals that often exist in practical applications, this algorithm utilizes the wide beamforming and the wide null constraint to improve the MVDR performances. It can not only point the mainlobe of the beam at the direction of the desired signal, but also shape the wide null at the direction of the interfered signal to restrain the interference. The result of simulation shows that the algorithm we proposed has improved performances.