A Toeplitz Preprocessing Method for DOA Estimation With Integrated Model Mismatch

In the field of array signal processing, it is crucial to estimate the information of the direction of arrival (DOA). However, the DOA estimation performance will be deteriorated significantly in the scenario with system imperfections. In this letter, the DOA estimation problem is investigated, and a novel system model with integrated model mismatch is formulated by combining channel amplitude-phase inconsistency, sensor position perturbation and mutual coupling. Then, after analyzing the performance of the multiple signal classification (MUSIC) in the scenario with integrated model mismatch, an iterative Toeplitz preprocessing method is proposed for the MUSIC algorithm to maintain the Toeplitz property during obtaining the covariance matrix of the received signals. Finally, simulation results are given to show the estimation performance, and a prototype system is tested in the anechoic chamber. Results show that the proposed method outperforms the existing methods in the scenario with model mismatches.