In this study, we investigate the estimation of the Two-Dimensional (2D) Direction Of Arrival (DOA) in monostatic multiple-input–multiple-output radar with cross array and propose a novel, highly accurate DOA estimation method based on unitary transformation. First, we design a new unitary matrix using the central symmetry of a cross array at transmit and receive sites. Then, the rotational invariance relationships of these arrays with long and short baselines can be transformed into a real-value field via unitary transformation. In addition, non-ambiguous and highly accurate 2D DOA estimations can be obtained using a unitary dual-resolution ESPRIT algorithm. Simulations show that the proposed method can estimate 2D highly accurate spatial angles using automatic pairing without incurring the expense of array aperture and peak searching. Compared with traditional unitary transformation, the steering vectors of transmit and receive arrays can be transformed into real-value fields via the unitary matrix and the transformation method of our scheme, respectively. This effectively overcomes the problem of shift invariance factors in real-value fields that cannot be extracted using traditional algorithms. Therefore, the proposed method can absolutely compute eigenvalue decomposition and estimate parameters in a real-value field, resulting in lower computational complexity compared with traditional methods. Simulation results verify both the correctness of our theoretical analysis and the effectiveness of the proposed algorithm.%该文针对十字型阵列配置下的单基地MIMO雷达2维空间角度估计问题，提出一种新的基于酉变换的高精度估计算法，算法利用收、发阵列中十字型阵列的中心对称性，通过酉矩阵设计及酉变换过程，将发射、接收阵列中对应的长、短基线平移不变关系映射到实数域，并基于双尺度酉ESPRIT算法实现2维参数的无模糊高精度估计。仿真结果表明：该文算法不牺牲阵列孔径，在获得2维空间角度无模糊高精度估计的同时，无需额外的配对算法和谱搜索；与传统酉变换过程相比，该文算法的酉变换过程等效将发射与接收导向矢量分别实数化，有效克服了传统酉变换算法中实数域旋转不变因子无法提取的问题，在实现特征分解与参数求解完全实数化的同时，具有更低的运算复杂度。最后，通过仿真实验验证了该文算法的正确性和算法的有效性。
展开▼
