Efficient 2D DOA Estimation Based on Decoupled Steering Vector for Uniform Circular Array

This paper proposes a novel decoupled twodimensional (2D) DOA estimation method for the uniform circular array (UCA). In the proposed method, the Jacobi-Anger expansion is employed to decompose the array steering matrix into the product of two matrices, each related to azimuth and elevation angles, respectively. Notably, one of these matrices exhibits Vandermonde structure, which enables the application of the root-MUSIC algorithm. Although the performance of conventional 2D MUSIC can be enhanced by decreasing search step size, this typically incurs significant computational costs. To reduce the computational complexity associated with search, the decoupled array steering facilitates the implementation of root-MUSIC for 2D angle estimation, which requires only 1D azimuth search, while elevation angles are determined through root-MUSIC. Moreover, the estimated azimuth angles are automatically paired with the corresponding estimated elevation angles. Simulation results demonstrate the effectiveness of the proposed method, and it has better performance than the other existing methods while reduces computational complexity.