A Two-Step Weighted Least Squares Self-Calibration Method for Gain-Phase Errors in Uniform Linear Arrays

This paper investigates the self-calibration of gain-phase errors in Uniform Linear Arrays (ULA). Traditional methods that exploit the Toeplitz structure of the signal covariance matrix offer closed-form solutions for gain-phase errors. However, they primarily account for external noise, whereas the internal thermal noise of the receiver typically exerts a more significant influence in radar signal processing. To address this issue, a two-step weighted least squares (TWLS) algorithm is proposed for the signal model which accounts for internal noise. Furthermore, phase prior information is incorporated as a constraint based on practical application requirements. Simulation results demonstrate that the proposed method maintains high computational efficiency while achieving estimation performance close to the Cramér-Rao lower bound.