A Subspace Spectrum Estimation Method for MIMO Radar Models

This method implements a subspace spectrum estimation technique for MIMO radar models using an over-estimation approach to avoid source number estimation issues. It directly transforms the data covariance matrix to construct signal subspace projection matrices and noise subspace projection matrices, eliminating the need for eigenvalue decomposition required in classical MUSIC algorithms. The implementation avoids the complexity of distinguishing between small and large eigenvalues in non-ideal scenarios, reduces computational overhead, and maintains performance regardless of snapshot numbers. The algorithm effectively handles coherent sources by accurately estimating target incidence angles without generating false peaks.