PSI - Issue 84

Mario Costantini et al. / Procedia Structural Integrity 84 (2026) 859–866

860

1. Introduction Very high-resolution VHR spotlight SAR creates "angle-and-time diversity" by illuminating scenes for seconds. Standard SAR processing relies on coherent integration, assuming static targets with known heights. This assumption for long dwell is often violated, in fact moving or vibrating targets generate Doppler spectrum phase residuals and defocusing artifacts. However, these Micro-Doppler (MD) effects can be exploited to extract dynamic information: micro-motions introduce time-varying Doppler signatures, measurable as motion-related spectral components or paired echoes. To recover unknown parameters like target height and motion, several advanced processing families exist: • Time-Frequency Approaches: Techniques like the Short-Time Fourier Transform (STFT) and Wigner-Ville distributions visualize temporal changes (Harmanny et al. 2014, Chen et al. 2006). However, they are limited by resolution trade-offs (time-frequency trade-off) and sensitivity to clutter and noise (Djurovic et al., 2017). • Parametric and Phase Estimation: Methods like Inverse SAR (ISAR) (Raney 1971, Martorella et al. 2014, Zavagli et al. 2024, Eichel et al. 1989), High-Order Ambiguity Function (HAF) (Djurovic et al. 2017, Barbarossa et al. 1998), and Quasi-Maximum Likelihood (QML) estimators (Djurovic et al., 2014) maximize image intensity or contrast to "refocus" the image. While effective given a correct model, they are fragile in cases of model mismatch or low Signal-to-Noise Ratio (SNR). • Sub-Aperture and Tracking Strategies: These segment the aperture to estimate motion via inter-look displacements (Clemente et al. 2024, Biondi et al. 2019). Although effective on strong scatterers, performance degrades in low Signal-to-Clutter Ratio (SCR) environments. An exhaustive recent review of the latest advances in Micro-Doppler from SAR can be found in Rollo et al. (2025). These methodologies apply broadly, from tracking vehicles to monitoring vibrating structures. In civil infrastructure, changes in dynamic characteristics often indicate degradation. Consequently, micro-motion estimates serve as Structural Health Monitoring (SHM) observables (Biondi et al. 2020a, Clemente et al. 2019), providing information on the health of a structure subject to geological or tectonic movements (Buffarini et al. 2023, Clemente 2024). However, infrastructure monitoring is challenging due to dense urban clutter and multipath, which complicate extraction—especially for amplitude-based methods. Although techniques like Singular Value Decomposition (SVD) help mitigate noise (Wang et al. 2012, Clemente et al. 2012), standard estimators often struggle in these complex environments. The increasing exposure of critical infrastructure to natural and anthropogenic threats has prioritized systematic, preventive monitoring. Structures like bridges and strategic buildings face continuous loads and degradation that can escalate into significant risks during extreme events. In this framework, Directive (EU) 2022/2557 (CER Directive) mandates a systematic approach to disaster risk reduction, emphasizing preventive vulnerability analysis and risk proportionate monitoring. Structural vibration characterization serves as a direct indicator of asset health. Advanced Micro-Doppler analysis of VHR satellite SAR data extends this capability to a territorial scale, enabling synoptic, non-invasive monitoring. Relevant to the CER Directive, these techniques provide quantitative tools for establishing vibrational baselines and detecting anomalies, supporting decision-making in both prevention and post-event scenarios. This approach offers a concrete contribution to critical infrastructure resilience strategies, aligning with European disaster risk reduction objectives. This paper presents a unified scientific view of satellite SAR Doppler-variation sensing for infrastructure monitoring, framing structural vibrations as a subset of the broader Doppler estimation problem. To address clutter limitations, we introduce a practical algorithmic framework focused on two estimator families: • Sub-Pixel Offset Tracking (SPOT), which utilizes azimuth sub-apertures. • Doppler Phase Coherence (DPC), a novel, patent-pending approach that improves robustness by prioritizing Doppler-spectrum phase coherence over amplitude. We clarify the physical quantities observable via single SAR acquisitions and summarize validation tests on controlled vibrating target and bridges. By focusing on vibration mode estimation, we demonstrate how Doppler variation products contribute to critical infrastructure resilience in alignment with the CER directive.