提出了一种新的适用于α稳定噪声环境的基于改进传播算子的DOA估计算法(或称为FLOM-PM算法)。由于在α稳定分布噪声的环境下不存在有限的二阶矩,因此采用最小二乘(LS)方法来求解传播算子的算法将不再适用。在原传播算子算法(PM)的基础上,通过引入分数低阶矩建立新的误差代价函数,并通过对该代价函数的最小化,来估计传播算子,进而完成DOA估计。由于该算法可直接通过对阵列接收的数据向量进行分块来求解噪声子空间,不需要估计和计算共变矩阵(或分数低阶矩)及其特征向量,不但可减小估计的误差,而且可实现对小快拍数据的高分辨处理。仿真实验验证了该算法的有效性和正确性。
A new direction of arrival (DOA) estimation algorithm (or FLOM-PM algorithm) based on an improved propagator method (PM) in alpha stable noise environments is presented in this paper. As there is not finite second order moments in alpha stable noise,the least square (LS) propagator cannot be used. In this paper, the propagator is estimated by minimizing a new cost function,which is defined by introducing the fractional lower order moments (FLOMs) ,so as to carry out the DOA estimation. This algorithm can estimate the noise subspace by partitioning the data received by the matrix. It can reduce the error of the estimation without computing the covariation matrix or FLOM matrix as well the eigenvector, and implement the high resolution processing of a very few snapshots. Simulation proves correctness of this algorithm.