Amplitude and Phase Errors Calibration for MIMO-SAR Based on PAST and Joint Optimization via Image Entropy Minimization

Multiple-Input Multiple-Output Synthetic Aperture Radar (MIMO-SAR) enhances imaging resolution by extending the virtual aperture without increasing the physical size of the antenna. However, its multichannel architecture inevitably introduces amplitude and phase inconsistencies, which significantly degrade image quality. Existing calibration methods typically rely on high-SNR reference targets, but the bistatic geometry inherent to MIMO-SAR systems, combined with the non-ideal scattering characteristics of reference targets, introduces significant estimation bias in channel amplitude and phase calibration. To address this issue, a minimum deployment distance threshold for trihedral corner reflector-based calibration is first determined to mitigate bistatic-induced disturbances. Then, a two-step estimation strategy is proposed to effectively suppress residual errors and improve the accuracy of calibration. Specifically, the initial estimation of amplitude and phase errors is obtained using the Projection Approximation Subspace Tracking algorithm. Subsequently, fine correction is achieved by minimizing image entropy, employing a joint optimization scheme based on Adamax and the Conjugate Gradient method, which enhances convergence stability and avoids local minima. Simulation and experimental results confirm the effectiveness of the proposed method in accurately estimating channel mismatches and substantially enhancing imaging quality. Furthermore, measured data demonstrate its robustness and applicability in high-precision imaging under complex observation conditions.