Abstract: A low-complexity method for estimating the parameters of an incoherently distributed source is proposed. The proposed method shows that the elements of the noise-free covariance matrix can be separated into two parts denoting, respectively, the amplitude information and phase information under small spread. Based on the above property, the nominal Direction-of-Arrival（DOA） is firstly estimated by using the phase information of the secondary diagonal elements of the sample covariance matrix. And then a cost function is formulated by exploiting the approximate orthogonality property between the column vectors of the noise-free covariance matrix and the sample pseudonoise subspace. The angular spread is estimated by one-dimensional searching. The?proposed?method avoids the eigendecomposition of the sample covariance matrix, and thus does not require the knowledge of the effective dimension of the pseudosignal subspace, which is a main different in the existing subspace methods. In addition, it estimates two parameters by only using once one-dimensional spectrum searching, and thus provides low computational complexity. Simulation results indicate the efficiency of the method.

Key words: incoherently distributed source, nominal DOA, angular spread, noise-free covariance matrix, sample pseudonoise subspace

摘要： 提出了一种低复杂度的非相干分布源参数估计方法，证明了在小角度扩展下，无噪声协方差矩阵的元素能够被分离成幅度信息和相位信息两部分。基于此性质，利用样本协方差矩阵次对角线上元素的相位信息首先估计出分布源的中心波达方向。根据无噪声协方差矩阵列矢量和采样伪噪声子空间之间的近似正交性构造代价函数，并通过一维搜索估计出分布源的角度扩展参数。该方法避免了对样本协方差矩阵进行特征分解，不需要确定伪信号子空间的维数，而这是现有子空间方法的主要困难。此外，它估计分布源的两个角度参数时仅需要完成一次一维谱搜索，计算复杂度较低。仿真结果证实了所提方法的有效性。

关键词: 非相干分布源, 中心波达方向, 角度扩展, 无噪声协方差矩阵列, 采样伪噪声子空间