基于降维实值ESPRIT的多输入多输出雷达波达方向估计

In order to improve the performance of angle estimation in Multi-Input Multi-Output (MIMO) radar and reduce the computational burden, this paper proposes a reduced-dimensional real-valued ESPRIT algorithm for Direction-of-Arrive (DOA) estimation in monostatic MIMO radar. First, the reduced-dimensional transformation is used to transform the received data into a lower-dimensional space. Second, the real and imaginary parts of the received data are spliced by taking advantage of the characteristic of the reflected signal to construct a real-valued observation data vector in which the number of elements is doubled. Third, the real-valued rotational invariance relationship of the extended steering matrix is constructed. Finally, the DOAs are obtained by solving the real-valued rotational invariance equation. Compared with the conventional ESPRIT algorithm, this algorithm can obtain better angle estimation performance with greatly reduced computational complexity. Simulation results demonstrate the effectiveness of the proposed algorithm.