基于信号子空间的多维参数联合估计

Multiple parameters joint estimation via single signal subspace

旋转不变的子空间中特征值的估计与关联是多维谱估计和阵列信号处理的重要问题。采用平行的交叉偶极子均匀线阵,利用各子阵列输出信号的相关矩阵构造多个旋转不变的子空间,并用其中一个子空间的信号子空间实现了频率、二维到达角和极化的自动配对。整个计算过程只需进行一次矩阵特征值的分解,就可完成多维参量估计与配对,且不存在参量兼并现象,理论分析与计算仿真结果证明了方法的有效性。

The eigenvalues estimation and association of subspace rotation invariance is an improtant problem in the multidimensional spectral and array processing. This paper exploits multiple invariance subspace of the second-order statistic of the observations collected from parallel crossed dipoles uniform linear array, and automatically paired frequency, two-dimensional directions and the polarization of incoming narrow-band signals via single signal subspace as long as each signal parameters are unique. The algorithm provides multiple parameter estimation via single eigenvector matrix that neither requires any expensive search procedure nor any other pairing strategy. Moreover it does not break down if several independent signals have common parameters. The Monte Carlo simulations are shown to be in close agreement with theoretically analysis.