Unitary-ESPRIT Algorithm for 2D DOA Estimation in Planar Arrays

This article introduces a MATLAB implementation of the Unitary-ESPRIT algorithm for two-dimensional angle estimation in planar arrays. Compared to traditional MUSIC algorithms, this approach demonstrates superior computational efficiency while directly estimating two-dimensional directions. The implementation has been verified on MATLAB 2017b, confirming successful operation. Furthermore, this article provides an in-depth exploration of the Unitary-ESPRIT algorithm's working principles, accompanied by detailed code implementation processes to facilitate better understanding and practical application. The MATLAB code implements key algorithmic steps including: - Real-valued transformation through unitary matrix operations to reduce computational complexity - Signal subspace estimation via eigenvalue decomposition of the covariance matrix - Closed-form solution for automatic pairing of elevation and azimuth angles - Implementation of the invariance principle for structured arrays Key functions include data covariance matrix calculation, real-valued subspace estimation, and automatic angle pairing without requiring additional search procedures. The algorithm efficiently handles the estimation of both azimuth and elevation angles simultaneously through inherent pairing mechanisms, eliminating the need for computationally expensive spectral searches characteristic of MUSIC-based approaches.