Abstract: A new two-dimensional （2-D） Direction Of Arrival （DOA） estimation algorithm based on Compressed Sensing （CS） is proposed with L-shaped array. After a space combined angle consists of azimuth and elevation angles is defined, a redundant dictionary is constructed through equally cosine partitioning the space combined angle. And then the output data vector of subarrays can be sparsely represented by the dictionary. In this work, Gauss measuring matrix is constructed by randomly sampling unit column vectors, and the bearings of spatial targets are estimated by improved Orthogonal Matching Pursuit （OMP） algorithm. By contrast with conventional 2-D MUSIC algorithm, simulations are evaluated and the results show that the new algorithm can achieve much better performance under the condition of lower Signal-to-Noise Ratio （SNR）, single snapshot of data, near and coherent sources.

Key words: array signal processing, two Dimensional Direction of Arrival（DOA） estimation, Compressed Sensing（CS）, redundant dictionary, spare representation, Gauss random measuring matrix

摘要： 针对L阵列，提出基于压缩感知的二维波达方向估计新方法。定义方位角和俯仰角余弦乘积为空间合成角，利用等余弦网格划分空间合成角构造超完备的冗余字典，将子阵接收的单快拍数据矢量转化为冗余字典下的稀疏表示问题；采用单位列向量组合矩阵随机抽取的新方法构造高斯随机测量矩阵；通过改进正交匹配追踪算法求解二维波达角。指出所提出的算法比传统二维MUSIC算法在高信噪比、多快拍条件下估计性能更好，并且有一定的阵元节约效能。计算机仿真实验证明了以上结论。

关键词: 阵列信号处理, 二维波达方向估计, 压缩感知, 冗余字典, 稀疏表示, 高斯随机测量矩阵