Accurate Residual Motion Error Estimation for High-Frequency Drone-Borne SAR Interferometry

The accurate flight trajectory knowledge is crucial for airborne and drone-borne repeat-pass synthetic aperture radar (SAR) interferometry. Given that Global Navigation Satellite Systems (GNSS) inaccuracies are comparable with typical radar wavelengths, especially at high-frequency bands (X band and higher), estimating the residual motion error (RME) is essential to ensure precise coregistration of single-look complex (SLC) stacks and to generate high-quality repeat-pass interferograms. In this framework, traditional RME estimation techniques, such as the multisquint algorithm, often suffer performance degradation under large errors and high-decorrelation conditions, as they rely heavily on the quality of uncoregistered interferograms. To address this limitation, this article introduces an enhanced version of the multisquint technique that solves the estimation of RME using a minimization strategy in the complex domain and uses the retrieved RMEs of the entire interferometric dataset to refine the estimates. This approach improves the RME estimation robustness against large errors and decorrelation artifacts, particularly in high-frequency airborne and drone-borne SAR datasets. The proposed method is validated using a real dataset acquired with a Ku-band drone-borne SAR system, demonstrating significant improvements in coregistration accuracy. Quantitative evaluations based on interferometric coherence and phase fringe frequency metrics confirm its effectiveness, establishing the proposed refinement as a robust solution for high-precision interferometric applications.