盖氏半径法在超分辨测向算法中的应用研究

Application of Gerschgorin Radii to Super-resolution Direction Estimation Algorithm

确定空间同频窄带信号源数在超分辨测向中至关重要,是能否成功测向的前提条件.本文将盖氏圆盘定理应用于阵列天线超分辨测向算法中,对协方差矩阵进行变换,将噪声圆盘和信号圆盘分开,进行盖氏圆盘估计,此方法能准确估计出被测信号源数,从而可提高超分辨测向的精度.同时在任意形天线阵条件下,进行了计算机仿真,并与利用最小描述准则的方法进行了比较,结果表明本方法对于超分辨测向有很大实际意义,能够提高测向精度.

Gerschgorin Radii is applied to the superresolution direction estimation algorithm of antenna arry to accurately estimate the source number and impove the pricision of superresolution diretion estimation . Covariance matrix transform is conducted to segregate noise disks from signal disks,then estimation is made using Gerschgorin disks.Computer simulation made under circumstance of spatial arbitrary antenna arry, and comparison made with MDL principle show that the methods are useful for superresolution diretion estimation ,and can impove pricision of diretion estimation . 