基于频域数据矩阵奇异值分解的宽带测向算法

Wide-Band Direction Finding Based on Frequency-Domain Data Matrix Singular Value Decomposition

文中提出了一种新的宽带信号测向算法。利用宽带阵列数据时间-带宽乘积大的特性,构造一个频域数据矩阵,对其进行奇异值分解,应用子空间的方法进行一维谱峰搜索,就得到来波方向的估计值。仿真实验结果证明了算法的有效性,并且给出了方向估计精度、分辨率与信噪比、快拍数、相对带宽之间的关系。

This paper proposes a novel wide-band direction finding algorithm. A frequency-domain data matrix is constructed making use of the large time-bandwidth product of wide band array data. Direction of arrival is estimated using approach based on subspace after the singular value decomposition of the matrix. Extensive simulation results prove the efficacy of the algorithm. At the same time the relation between accuracy, resolution and signal to noise ratio, snapshots, relative bandwidth is got.