Abstract: A new 2-D Unitary ESPRIT algorithm is proposed based on the characteristics of two parallel Uniform Linear Arrays（ULAs）. The method constructs a real data style based on the data arrays received, and gains the signal subspace by performing real eigenvalue decomposition. The shift invariance property of the array geometry is employed. A new estimator that contains the azimuth and elevation angles of the incidents signals is reconstructed. The algorithm can utilize both the eigenvalues and eigenvectors to obtain the azimuth and elevation angles. The proposed method can not only hold the advantages of DOA matrix about automatical parameter alignment and no need of peak searching, but also can deal with the data in real number field to obtain the subspace, which can decrease the amount of calculation. This algorithm can estimate more signal than array number. Simulation experiments verify the effectiveness of the algorithm.

Key words: two parallel Uniform Linear Arrays（ULAs）, 2-D Direction Of Arrival（DOA） estimation, Unitary Estimating Signal Parameter via Rotational Invariance Techniques（ESPRIT） algorithm, automatical parameter alignment

摘要： 提出了一种基于双均匀线阵的特性新颖的Unitary ESPRIT算法。算法根据阵列接收数据，构造了一种实值数据形式，对其作实值奇异值分解，获得信号子空间。算法利用阵元几何的平移不变特性，构造了包含入射信号方位角和俯仰角的估计量。利用构造的估计量的特征值估计俯仰角信息，利用其对应的特征向量估计方位角信息，自动完成角度的配对。所提方法不仅保持了矩阵DOA估计算法和二维ESPRIT算法的优点，如自动的参数配对、不需要二维谱峰搜索等，而且能够在实数域中直接获得信号子空间，能够有效降低运算量。算法可以增加估计信号个数。仿真实验验证了算法的有效性。

关键词: 双平行阵, 二维到达方向（DOA）估计, Unitary旋转不变子空间（ESPRIT）算法, 参数自动配对