Towards robust micro-motion estimation in bistatic and distributed SAR using extended Kalman filtering

Tools

Rollo, Finlay Gardiner and Clemente, Carmine; (2025) Towards robust micro-motion estimation in bistatic and distributed SAR using extended Kalman filtering. In: SPIE sensors and Imaging 2025. SPIE, ESP. (In Press)

[thumbnail of Rollo-and-Clemente-2025-Towards-robust-micro-motion-estimation-in-bistatic-and-distributed-SAR]
Preview
 Text. Filename: Rollo-and-Clemente-2025-Towards-robust-micro-motion-estimation-in-bistatic-and-distributed-SAR.pdf
Final Published Version
License: All rights reserved
Download (2MB)| Preview

Abstract

Recent advances in commercial synthetic aperture radar (SAR) technologies and the growing, open availability of high-resolution SAR data have significantly expanded the potential for micro-motion analysis within SAR imagery. Building upon our previously introduced modified backprojection algorithm - in which pulses are concatenated rather than coherently summed - more complex imaging geometries are now investigated for micromotion analysis. Among these include bistatic geometry with one active and one passive radar as well as instances where two monostatic satellites capture the same scene simultaneously. These distributed collections not only reflect emerging capabilities of real SAR constellations but also motivate fusion strategies to exploit these collection geometries more effectively. To address the challenges in combining these signals, a phase tracking extended Kalman filter (EKF) pipeline is presented. The EKF strategy is repurposed from previous implementations where it was shown to be highly effective in estimating polynomial phase signals. Working under similar assumptions, this paper shows that it can be an effective pipeline for processing micro-Doppler signals too. This is developed further by expanding the EKF to fuse phase information from multiple sensors, to further improve micro-motion estimates. Results showcasing this approach are shown on a simulated dataset.

ORCID iDs

Rollo, Finlay Gardiner and Clemente, Carmine ORCID logoORCID: https://orcid.org/0000-0002-6665-693X;
CORE (COnnecting REpositories)