基于幅相误差阵列的远近场混合信号超分辨测向方法

大多数的超分辨测向方法都需要掌握准确的阵列流型，然而在实际应用当中，各阵元通道对信号的幅度增益和延时往往不一致，使得真正的阵列流型和它的理论模型之间存在一定的误差，最终造成测向性能的下降.针对这个问题，论文提出了一种基于幅相误差阵列的远近场混合信号超分辨测向估计方法.首先对信号的空间谱函数进行变换判断出远场信号方向，接着根据远场信号子空间和噪声子空间的正交性估计出阵列误差并对数据进行校正，在此基础上通过矩阵分解判断出近场信号方向，同时还能够实现近场信号的定位.所提方法直接对信号空间谱函数分母的多项式求根得出信号方向和近场信号距离，回避了谱峰搜索的过程，在保证一定精度的前提下大大提高了计算速度.

Super-resolution direction finding method for mixed far-field and near-field signals based on gain-phase error array

Most of super-resolution direction finding methods always need precise array manifold.However, the amplitude gain and time delay of each channel are usually inequable, which results in the calculation performance degradation eventually.Therefore, this paper presents a new method of super-resolution direction finding for mixed far-field and near-field sources based on gain-phase error array.Direction of far-field signals is firstly acquired by transforming the space spectrum function.Then orthogonality of far-field signal subspace and noise subspace is employed to compute the gain and phase uncertainty for calibration.Finally, by removing the array error and part of far-field sources, directions and distances of near-field signals are obtained,meanwhile, their locations are also determined.The method estimates the signals by roots finding directly instead of spectrum searching, so as to speed up the calculation greatly with precision guaranteed as well.