Skip to content
Open access

Microvibration Target Mapping in High-Resolution SAR Using Pixelwise Doppler Centroid Estimation From Subaperture Decomposition

2026 · IEEE Transactions on Aerospace and Electronic Systems · Vol 62, pp. 17647-17665 · 0 citations · 37 references

Abstract

Micro-Doppler (m-D) signatures in synthetic aperture radar (SAR) enable the characterization of subresolution harmonic vibrations. However, most existing approaches rely on short-time/time–frequency decomposition or polynomial phase estimators that introduce resolution tradeoffs and limited robustness. This article proposes a spectral-domain framework that exploits the inner harmonic structure of the azimuth spectrum through subaperture-based Doppler centroid (DC) analysis. The full synthetic aperture is partitioned into relatively large azimuth subbands (long subdwell intervals), and a correlation-based DC estimator is evaluated for each subaperture and image pixel. Although for such extended intervals, the DC estimate no longer preserves strict proportionality to instantaneous radial velocity and therefore does not act as a straightforward proxy for amplitude estimation due to intrinsic ambiguities in SAR DC theory, the sequence of DC values along the subaperture dimension retains a deterministic harmonic modulation induced by sinusoidal target motion. In the proposed approach, this DC trajectory is exploited exclusively for vibration frequency estimation, while amplitude retrieval is deliberately decoupled. The harmonic content of the DC sequence is analyzed via sparsity-driven spectral measures, enabling the generation of a vibration detection map over the entire area of interest, where each pixel is characterized by the concentration of its DC spectrum. This permits large-scale monitoring and automatic highlighting of dynamically active areas prior to any localized refinement or high-resolution parameter estimation. Owing to the use of large subapertures, hence wide bandwidth and long dwell-time integration, the method achieves superior robustness to noise compared with approaches based on pixel-tracking over small subapertures, which typically suffer from reduced signal-to-noise ratio and decorrelation effects. Simulated and experimental SAR results demonstrate accurate vibration frequency retrieval and reliable wide-area detection capability. The framework validation on real-world data was performed using a vibrating corner reflector and its digital twin SAR simulator.

Read PDF

We use cookies to run the site and, with your consent, for analytics and to show ads. See our Cookie Policy.