Dynamic Multistatic Synthetic Aperture Radar Imaging: Resolving Motion via Differential Semblance and k-Space Diversity

Abstract

Moving targets are well known to appear displaced and defocused in synthetic aperture radar (SAR) imagery, complicating both detection and recognition—especially when displaced into cluttered regions, where signal-to-background ratio is reduced. Multistatic SAR offers the potential to extract rich scene information, including enhanced moving target imaging, yet fully exploiting this diversity remains a significant challenge. We present a novel multistatic SAR collection geometry and processing framework that enables moving target indication, approximate signature extraction, motion estimation, and high-fidelity image formation. By minimising differences in refocused signatures across multistatic channels via differential semblance optimisation, we estimate target velocity and use it to form joint multichannel reconstructions which suppress background clutter and enhance target features. The proposed collection geometry allows flexibility to balance spatial resolution and diversity in observation angles with motion sensitivity, and maximises the information content of multistatic data for a desired deployment aim. Our methodology is validated using laboratory-collected SAR data, and is readily applicable to UAV-based systems, offering practical flexibility in deployment for scenarios requiring both fine-resolution imaging and robust moving target detection.