Journal of Systems Engineering and Electronics ›› 2013, Vol. 35 ›› Issue (3): 504-510.doi: 10.3969/j.issn.1001-506X.2013.03.09

• 传感器与信号处理 • 上一篇    下一篇


郑志东, 张剑云, 周青松, 刘春生   

  1. 解放军电子工程学院, 安徽 合肥 230037
  • 出版日期:2013-03-20 发布日期:2010-01-03

Coherent multitarget localization for bistatic MIMO radar based on vector reconstruction

ZHENG Zhi-dong, ZHANG Jian-yun, ZHOU Qing-song, LIU Chun-sheng   

  1. Electronic Engineering Institute, Hefei 230037, China
  • Online:2013-03-20 Published:2010-01-03


提出了一种基于目标信息矢量重构的双基地多输入多输出(multiple-input multiple-output, MIMO)雷达相干信源角度估计方法。利用目标信息矢量中的元素,构造出解相干处理的通用块Hankel矩阵。证明了该矩阵的秩为总目标数时矩阵行数和列数所应满足的条件,并基于奇异值分解求解出信号和噪声子空间,然后利用ESPRIT算法获得角度的估计值。同时,给出了直接数据提取法和特征矢量提取法来获得目标的信息矢量。仿真实验表明:本文算法能够有效地估计出相干信源的收发角度,且实现自动配对;当兼顾角度估计精度和算法的复杂度时,应满足块Hankel矩阵的行数不大于列数;在低信噪比下,本文算法的估计精度优于空间平滑算法,且特征矢量提取法的估计精度优于直接数据提取法。


The new algorithm based on the information vector reconstruction is presented to estimate the angle for coherent signals in bistatic multiple-input multiple-output (MIMO) radar. The general block Hankel matrix is constructed to achieve the decorrelation by utilizing the elements of the target information vector. The conditions on the numbers of row and column for the general block Hankel matrix are derived under which the rank of the matrix is equal to the number of targets. And the signal and noise subspaces are exploited via the singular value decomposition (SVD) of the general block Hankel matrix. Then the estimation of signal parameters via ratational invariance technique (ESPRIT) algorithm is directly used to estimate the coherent angle with automatic pairing. Furthermore, two methods, receive data extraction and eigenvector extraction, are proposed to obtain the target information vector. The simulation results show that the angle of targets can be estimated efficiently without extra pairing; when striking balance between the angle estimation accuracy and the computation complexity, the column number of the general block Hankel matrix should be no less than the row number; the proposed method has a better estimation performance than the spatial smoothing algorithm in the low signal-to-noise ratio (SNR), and the estimation performance of the receive data extraction method outperforms that of the eigenvector extraction method.