An Advanced Method for Precise ULA SIMO Radar Calibration Utilizing Synthetic Aperture Radar Imaging Artifacts

Abstract

Radar calibration has always been essential for compensating unavoidable manufacturing inaccuracies, component variations, or aging effects in multichannel radar systems. The demand for high-resolution radars, particularly in automotive applications, necessitates increasing carrier frequencies, bandwidths, synthetic apertures, and channel counts, imposing exceptionally high calibration requirements. Conventional calibration methods often rely on expensive positioning systems and large-scale anechoic chambers, yet offer only limited calibration accuracy. This publication presents a novel calibration method that achieves exceptionally high accuracy for single-input multiple-output (SIMO) radars with uniform linear arrays (ULAs) by estimating amplitude and phase deviations as well as mutual coupling between the channels. The proposed method leverages the established theory of accumulating channel imbalances in synthetic aperture radar (SAR) processing, enabling the isolation of error power from the desired signal and noise within the image. The applied optimization minimizes only at deterministic artifact locations, which enhances the optimization sensitivity and improves calibration precision, while reducing the computational complexity. The proposed approach demonstrates high performance by calibrating an antenna array with eight elements in a 77 GHz frequency-modulated continuous wave SIMO radar. The resulting calibration quality is validated in a test scene, demonstrating significantly reduced artifacts within the generated image compared to the uncalibrated array.