Research On Distributed Beamforming Algorithms Based On Array Networks

Beamforming is used in a wide range of fields,such as radar,wireless communi-cation,sonar,acoustics and other fields.In recent years,with the rapid development of wireless sensor network technology and embedded technology,distributed collaborative beamforming based on array networks,a newly arisen subject,has aroused increasing at-tention from scholars.By applying the distributed adaptive signal processing technology to the array network,compared with the centralized signal processing technology,it has better scalability,diffusion,robustness and other characteristics.This dissertation takes the distributed beamforming algorithms based on the array networks as the research topic,from the directions of adaptive combination coefficients,reduced-rank technology,tensor decomposition,etc.The main research content is divided into the following three parts.(1)Aiming at the problem of the poor robustness of the existing distributed diffusion strategy with fixed combination coefficients to the spatial variation of the signal and noise statistics in the network,a distributed beamforming algorithm with adaptive combination coefficients is derived in this thesis.Simulation experiments have proved that,compared with the existing combination scheme with fixed combination coefficients,the combi-nation scheme with adaptive combination coefficients can improve the robustness of the distributed beamforming algorithm to the spatial variation of signal and noise statistics.(2)In view of the slow convergence speed of traditional distributed beamforming al-gorithms under large-scale array networks and the performance of the traditional reduced-rank beamforming algorithm will deteriorate or even fail when the angle of arrival of the desired signal changes rapidly.Leveraging the reduced-rank methodology and dis-tributed beamforming technology,a diffusion reduced-rank beam coordination algorithm along with an inherently adaptive combination scheme based on the combination matrix is derived,and two adaptive implementation schemes are given in this thesis.The derived distributed reduced-rank beamforming algorithm can effectively improve the convergence speed of the algorithm in a large-scale array network,and are capable of maintaining the approximate unit gain in the DOA of the desired signal even when the DOA of the desired signal changes,which is demonstrated via simulations.(3)Aiming at the problems of slow convergence speed,high computational complex-ity and high communication cost of traditional distributed beamforming algorithms under large-scale array networks.Leveraging the related theories of tensor and distributed signal processing algorithms,and utilizing the structure of the multi-linear translation invariance of the array,a distributed tensor beamforming algorithm based on the array networks is derived in this thesis.Compared with the traditional distributed beamforming algorithm,the distributed tensor beamforming algorithm can effectively accelerate the convergence speed of the algorithm,reduce the computational complexity and signaling overhead be-tween the arrays.