Source Code for DOA Estimation Using Rotation Invariant Subspace ESPRIT Algorithm

This document explores the application of the ESPRIT algorithm in source code for Direction of Arrival (DOA) estimation. The ESPRIT algorithm operates through rotation invariant subspace principles, where a rotation invariant subspace refers to a vector space that remains unchanged under rotational transformations. The algorithm's implementation typically involves signal subspace decomposition through eigenvalue decomposition of the covariance matrix, followed by exploiting the rotational invariance property between signal subspaces from different subarrays. Key advantages include high-resolution signal parameter estimation, elimination of the need for explicit signal number and power estimation, and strong noise resistance. The code implementation generally involves constructing sensor array data covariance matrices, performing singular value decomposition (SVD) or eigenvalue decomposition (EVD) to extract signal subspaces, and solving rotational invariance equations to obtain DOA estimates. This content focuses on how the algorithm is applied to DOA estimation, detailing both its advantages and limitations. We also examine the algorithm's development history and potential future research directions, including computational efficiency improvements and extensions to more complex array geometries.