Second-order Circulation Statistics-based Array Signal Processing

This paper mainly study the second-order cyclostationary signal, at the same time carries out detailed study on LMS algorithm of non- blind adaptive beamforming algorithm and ESPRIT algorithm of blind adaptive beamforming algorithm which are time-domain based reference signals, then proposed improved algorithm combined with cyclostationary features of the second-order circulation statistics.The main results of this work are summarized as follows:First,mastered the mathematical modeling method of adaptive array antenna;Then,master the basic principles of the typical LMS algorithm , and aiming at the shortcomings that the performance of adaptive beamforming which based on a typical LMS algorithm will be reduced with a cyclostationary signal input , proposed to construct period average squared error indicator function by using periodic of second-order cyclostationary statistics. Then formulates the improved circulation least mean square (LMS) algorithm by adaptively optimizing the indicator function using LMS algorithm, and analysis the convergence this algorithm, and finally conducts computer simulation respectively on improved algorithm, LMS algorithm and BLMS algorithms, and at last analyzed and compared the results . Finally,mastered the basic principles of typical ESPRIT algorithm, and promoting the choice frequency vector-oriented method applied in improved narrow-band MUSIC algorithm to the ESPRIT algorithm, forming an improved narrow loop ESPRIT algorithm. In order to achieve the joint estimation of azimuth and elevation angle, the algorithm is applied to a two-dimensional array consisting of two orthogonal linear arrays. Simulation results of improved algorithms prove that improved algorithm has a good signal selectivity and estimation accuracy .