DoA Estimation in the Presence of Mutual Coupling Using Root-MUSIC Algorithm

The effect of unknown mutual coupling in receiving array seriously degrades the performance of direction-of-arrival (DoA) estimation algorithms. In order to compensate this effect, this paper develops an auto-calibration method for the uniform linear array (ULA) based on Root-MUltiple SIgnal Classification (MUSIC) algorithm. The proposed method can estimate the DoAs of the received signal and mutual coupling coefficients using the subspace principle without using any calibration source. Moreover, it can reduce computational complexity due to analytically estimating DoA without any spectrum search. In this paper, Monte-Carlo simulation is used to elucidate the errors in DoA and coupling parameter estimation. Cramer-Rao lower bound (CRB) is also presented to support the estimation results. Simulation results illustrate that the proposed Friedlander & Weiss (F&W) Root-MUSIC method efficiently estimates the DoA and mutual coupling coefficients, and also the performance of F&W Root-MUSIC is better than F&W MUSIC algorithm's.