基于EMEMP的雷达二维信号融合成像新方法

雷达信号融合成像是一种能显著提高成像分辨率的参数化新方法.基于改进的Root-Music的传统融合方法对噪声敏感, 且存在模型极点失配的问题.本文通过将MEMP(Matrix Enhancement and Matrix Pencil)的二维频率估计方法推广到稀疏数据域,提出了一种基于扩展矩阵增强矩阵束(EMEMP)的融合新方法.此方法首先构造每一维联合增强矩阵,使其满足MEMP算法的配对要求,然后利用MEMP方法估计模型极点,进行极点配对,然后估计模型系数,最后内插频谱以达到融合的目的.实验结果表明相对于传统融合方法,该方法解决了极点失配的问题,提高了模型参数估计的稳健性.

A New Method for Radar Two-Dimensional Signal Fusion Based on EMEMP

Radar signal fusion imaging is a new parameterized method.It can remarkably improve image resolution.Traditional fusion method based on modified Root-Music is sensitive to noise,and can not pair model poles exactly.Thus this paper extended MEMP(Matrix Enhancement and Matrix Pencil) to the two-dimensional frequencies estimation of spare data,and proposed a novel fusion technique based on Extended Matrix Enhancement and Matrix Pencil(EMEMP) method in order to apply to sparse data fusion.First,combinative enha...