| Smart antenna has become a research focus in the field of mobile communication,Direction Of Arrival（DOA）estimation and beamforming are its core technology.The algorithm of the two aspects is mainly researched in this paper.As to DOA estimation, on the base of the research of two traditional DOAestimation algorithms, which are multiple signal classification (MUSIC) algorithmand the modified multiple signal classification (MMUSIC) algorithm, an improvedMMUSIC algorithm is proposed. The algorithm, which combine covariancedifferencing approach with MMUSIC algorithm, can overcome the shortcoming thatcovariance differencing approach used in unknown noise environment can’teffectively estimate highly relevant or coherent signals and simultaneously can solvethe problem that MMUSIC algorithm is only applicable to the ideal white Gaussiannoise environment and is invalid in unknown noise environment. Simulation resultsshow that the improved MMUSIC algorithm can effectively estimate the signals’DOA in the presence of unknown correlation between signals and unknown noisefield, and gets very high estimation accuracy and angular resolution.As to beamforming, the beamforming algorithm based on the reference signal andthe beamforming algorithm based on DOA estimation are detailedly researched in thispaper. First, the least mean square (LMS) algorithm, two kinds of variable step sizeLMS algorithm and recursive least squares (RLS) algorithm are introduced andtheir convergence performance are analyzed and compared. Then we focus on thebeamforming algorithm based on DOA estimation with the existence of coherentinterference. Combining multi-beam constrained minimum variance (MBCMV)algorithm with eigenspace-based algorithm, the eigenspace-based multi-beamconstrained minimum variance (EMBCMV) algorithm is proposed, simulation resultsshow that EMBCMV algorithm can obtain very high output SINR, simultaneouslyhave fast convergence rate and a certain degree of robustness against signal pointingerror. |
PDF Full Text Request