基于累积量稀疏表示的二维波达方向估计方法

摘要

A sparse representation based on fourth-order cumulats method was proposed in this paper,which was used for two dimensional DOA (Direction Of Arrival)estimation in Gaussian colored noise.Firstly,a fea-ture of one dimension space angel containing the information about the azimuth and elevation was constructed, then,the fourth-order cumulats matrix using the feature was built for the two sub-arrays.After that,the DOA estimation was obtained by sparse representation using over-complete dictionary which was composed by the first column of the fourth-order cumculats matrix.By using amplitude information of the space angels got by sparse representation,the azimuth and elevation were obtained by computing the matched angels.Theoretical analysis and experimental results showed this method had higher estimation accuracy and Gaussian colored noise suppres-sion.%针对高斯色噪声下的二维 DOA 估计问题，提出了一种基于四阶累积量稀疏表示的估计方法。该方法首先构建包含方位角和俯仰角信息的一维空间锥角特征，然后分别利用该特征构造两个子阵的四阶累积量矩阵；通过四阶累积量矩阵中的第一列构建最小冗余向量完备字典实现了稀疏表示的 DOA估计，利用稀疏分解获得的幅度信息进行角度配对求解，最终得到方位角和俯仰角的二维 DOA 估计。理论分析及仿真实验表明，该方法具有较高的估计精度和对高斯色噪声较好的抑制能力。