多维阵列张量模型的鲁棒波束成形方法  

Robust beamforming method for multi-dimensional array tensor models

作　　者：毕权杨 李旦[1] 张建秋[1] BI Quanyang;LI Dan;ZHANG Jianqiu(School of Information Science and Engineering,Fudan University,Shanghai 200433,China)

机构地区：[1]复旦大学信息科学与工程学院,上海200433

出　　处：《系统工程与电子技术》2024年第7期2175-2183,共9页Systems Engineering and Electronics

基　　金：国家自然科学基金(11974082)资助课题。

摘　　要：传统波束成形算法在应对维数较高的多维阵列时,训练快拍数难以满足远大于信号维数这一要求,从而导致波束成形器性能急剧下降。针对这一问题,本文给出了多维阵列输出数据的张量模型,基于多维阵列子维度的可分离性,引入张量波束成形方法,分析了其在训练快拍数需求上具有的优势。然后,基于张量波束成形中子维度干扰协方差矩阵的模型,通过直接估计干扰协方差矩阵实现了一种鲁棒张量波束成形方法。分析表明:该方法可更好地应对非均匀杂波环境以及相干干扰,并得到更高的输出信干噪比,以给出的二维极化敏感阵列为例,所提方法的训练快拍数需求为传统方法的1/3,相干干扰下输出信干噪比提高了约2.5 dB,仿真验证了分析的有效性。Traditional beamforming performance will significantly degrade with insufficient training samples on multi-dimensional arrays with much higher dimensions.In this paper,a tensor beamforming method that has advantages in the number of training snapshots required is introduced based on the proposed tensor model and the separability of sub-dimensions for multi-dimensional arrays.Then,for coherent interference,based on the model of the sub-dimension’s interference covariance matrix in tensor beamforming,a robust tensor beamforming method is proposed by directly estimating the interference covariance matrix.Analysis shows that the proposed method could overcome the non-homogeneous clutter environment and coherent interference and obtain a higher output signal-to-interference-noise ratio.Taking the given two-dimensional polarization-sensitive array as an example,the number of training snaps required by the proposed method is 1/3 of the traditional method,and the output signal-to-interference-to-noise ratio increases by about 2.5 dB under coherent interference scenario.The simulation results verify the effectiveness of the analysis.

关 键 词：鲁棒波束成形 张量波束成形 多维阵列 相干干扰 

分 类 号：V243.2[航空宇航科学与技术—飞行器设计]