Higher-Order Statistics And Their Applications To Array Signal Processing

Space spectrum estimation technology has a comprehensive perspective in the military and civilian applications, The estimation method based on higher-order statistics has better performance than that based on covariance.Firstly, the thesis introduces the definitions and the attributes of the Higher-Order Statistics. Higher-Order Cumulants are blind to additive Gaussian noise (white or colored) and this is the theoretical advantage of Higher-Order Cumulants.Secondly, the thesis introduces the DOA problems and the widely used MUSIC method. To overcome the shortcomings of the traditional MUSIC method, the thesis proposes the FOC-MUSIC method, which has better performance than the traditional MUSIC method. At the same time, the thesis has made an in-depth study in the applicable condition of MMUSIC method, which can .reduce the correlation among the signals and reach better performance.Thirdly, Array errors are inevitable. To overcome the bad influence of Array errors, the thesis introduces two kinds of important algorithms that can get accurate estimation of DOA in the condition of Array errors, the thesis compares the two algorithms through the computer simulations.Lastly, based on formers'work on the Distributed micro-satellites system , the thesis has theoretically proved the feasibility that we will use the self-calibration of the array errors algorithm based on the Maximal Set of Non-redundant cumulants to the Distributed micro-satellites system.