The Study On Array Calibrations Algorithms For Spatial Spectrum Estimation

The merits of spatial spectrum estimation algorithms are quite obvious,including the high precision of DOA, the high resolution performance and the ability to measure a lot of signals simultaneously. However, in order to obtain these merits, this kind of method requires quite accurate array manifold. The array perturbations would lead to the serious deviations between the real array manifold and the ideal one, which make the performance of algorithms sharply reduced, and in some conditions, the algorithms even become useless. Thus, array calibration methods have attracted much attention and play a significant role in the area of array signal processing.Calibration algorithms have been divided into two major parts. On one hand, some algorithms calibrate the array perturbations through finding the best estimations of the array perturbations. There were two major branches in this field: self-calibration method (auto calibration method) and calibration with a set of calibration sources. On the other hand, some algorithms could obtain well performance of finding the DOA without estimating the perturbations which are named robust algorithm.In this thesis, the author studied on array calibration methods for spatial spectrum estimation. The main works in this paper as follows, at the beginning, the author has introduced the basic information on array calibration methods and the classical MUSIC algorithm. Secondly, the analysis has been made in order to build the models of array perturbations. In this part, the author pays much attention on the influence from gain/phase errors and mutual coupling errors. Thirdly, several specific algorithms have been discussed. At last, some simulations and graphs have been made in order to illustrate the influence from perturbations and verify the performance of calibration algorithms.