Novel Antenna Array Calibration Metrics and Methods for Radar Imaging Systems

Accurate antenna array calibration is crucial for achieving high image quality in multichannel radar systems, as it compensates for unknown deterministic perturbations in individual channels.However, existing calibration methods either require costly and complex measurement setups or offer limited accuracy, thereby restricting their practical applicability. This doctoral thesis introduces a set of novel methods for antenna array calibration and calibration quality assessment, which exploit the linear relationship between channel imbalances and characteristic artifacts in the respective reconstructed synthetic aperture radar images. Building on this relationship, a new online monitoring procedure was developed that enables continuous assessment of calibration quality during regular radar operation over extended periods of time. Furthermore, a new calibration method was proposed and subsequently refined, achieving high calibration accuracy with minimal measurement effort and computational cost. The developed methods were validated through simulations and measurements using a multichannel radar system. The results demonstrate that the proposed monitoring procedure – tested in outdoor environments similar to those encountered by automotive radars – can reliably detect even minor degradations in calibration quality. Moreover, the proposed calibration approach demonstrates that accurate calibration can be achieved using only linear aperture motion, thereby eliminating the need for costly and complex setups such as anechoic chambers, as confirmed under challenging measurement conditions in highly reflective indoor environments. These contributions enable multichannel radar systems to continuously evaluate their functional integrity during their regular operation, thereby enhancing safety in critical applications such as autonomous driving or space exploration. In addition, the proposed calibration method makes high-quality radar calibration accessible to a much broader range of users due to its low complexity and cost. The simplicity and practicality of both methods, combined with their low measurement requirements, pave the way for fully automated online calibration procedures in future radar systems.