基于分数低阶矩的空时最大似然DOA估计

为了提高现有基于分数低阶空时矩阵类算法的估计精度和其在低信噪比时的解相干性能,在对称α稳定分布噪声条件下,提出改进分数低阶空时矩阵用于独立信源DOA估计,基于分数低阶矩的空时最大似然算法用于相干信源DOA估计.仿真结果表明:针对独立信源,改进的分数低阶空时矩阵进行DOA估计的成功概率高于分数低阶空时矩方法;针对相干信源,分数低阶矩的空时最大似然算法的估计成功概率比SS-FLOM算法高,均方根误差比SS-FLOM算法小;新方法克服了空间平滑类算法严重损失阵列孔径的弊端,节省阵元.

DOA estimation of spatial-temporal and maximum-likelihood based on fractional lower-order statistics

In order to improve the estimation accuracy of existing fractional lower order spatial-temporal matrix algorithms, and their estimation performance for coherent sources at low SNR, within the conditions of impulsive noise fields modeled as symmetric α-stable distribution, fractional lower order spatial-temporal matrix is improved to be used in DOA estimation for independent sources, and a new spatial-temporal and maximum likelihood method based on fractional lower order statistics is studied to be used in DOA estimation for coherent sources. The results of simulation show the efficient and high-performance of new method. First, according to the independent source, the success probability of improved method based on improved fractional lower order spatial-temporal matrix is much more than fractional lower order spatial-temporal matrix. Second, according to coherent sources, the success probability of the new spatial-temporal and maximum likelihood method is much more than SS-FLOM, and RMS error is lower. Third, the new method avoids the drawbacks of sensor loss of space-smoothing algorithm.