Unsupervised Super Spatial Resolution Brillouin Frequency Shift Extraction Based on Physical Enhanced Neural Network
Journal of Lightwave Technology · 15.04.2026 · 44 (8) · 3280-3290
DOI: 10.1109/jlt.2026.3658374 ↗Kurz gefasst
Fachartikel in der Zeitschrift Journal of Lightwave Technology, erschienen am 15.04.2026. Verfasst von Zhao Ge und 5 weiteren Autor:innen. Die Arbeit wurde bisher 2-mal zitiert (Crossref) und nennt 28 Literaturangaben. Gefördert von: National Natural Science Foundation of China, Major Program (JD) of Hubei Province, Hubei Provincial Natural Science Foundation of China.
Abstract (Originalsprache)
Spatial resolution (SR), a core parameter of Brillouin optical time-domain analysis (BOTDA) sensors, determines the minimum fiber length over which physical perturbations can be accurately detected. However, the phonon lifetime in the fiber imposesan inherent limit on the SR, making sub-meter-level SR challenging to achieve. Conventional SR enhancement approaches, constrained by hardware limitations, often involve complex systems, or increased measurement times. Although deconvolution methods can mitigate hardware constraints, they suffer from distortion due to the nonlinear nature of the BOTDA response. Supervised deep learning approaches have recently emerged as an alternative, offering faster and more accurate post-processing through data-driven models. However, the need for extensive labeled data and the lack of physical priors lead to high computational costs and limited generalization. To overcome these challenges, we propose an unsupervised deep learning deconvolution framework, Physics-enhanced SR deep neural network (PSRN) guided by an approximate convolution model of the Brillouin gain spectrum (BGS). By embedding domain-specific physical knowledge directly into the architecture, PSRN enhances its interpretability and generalization across diverse BOTDA configurations. Besides, total variation regularization is introduced and proven to be effective in reducing artifacts. Simulated and experimental results demonstrate that PSRN can accurately retrieve sub-meter SR Brillouin frequency shift from any low-SR BGS input in a plug-and-play fashion, leveraging the interplay between neural network inference and embedded physical priors. This physics-neural network fusion establishes a generalizable paradigm for diverse fiber-optic sensing challenges.
Förderung
- National Natural Science Foundation of China
- Major Program (JD) of Hubei Province
- Hubei Provincial Natural Science Foundation of China
Zitieren
APA
Ge, Z., Wu, H., Zhao, Z., Zou, X., Wei, W., & Tang, M. (2026). Unsupervised Super Spatial Resolution Brillouin Frequency Shift Extraction Based on Physical Enhanced Neural Network. Journal of Lightwave Technology, 44(8), 3280-3290. https://doi.org/10.1109/jlt.2026.3658374
Vancouver
Ge Z, Wu H, Zhao Z, Zou X, Wei W, Tang M. Unsupervised Super Spatial Resolution Brillouin Frequency Shift Extraction Based on Physical Enhanced Neural Network. Journal of Lightwave Technology. 2026;44(8):3280-3290. doi:10.1109/jlt.2026.3658374
BibTeX
@article{Ge2026,
author = {Ge, Zhao and Wu, Hao and Zhao, Zhiyong and Zou, Xuan and Wei, Weilun and Tang, Ming},
title = {Unsupervised Super Spatial Resolution Brillouin Frequency Shift Extraction Based on Physical Enhanced Neural Network},
journal = {Journal of Lightwave Technology},
year = {2026},
volume = {44},
number = {8},
pages = {3280--3290},
doi = {10.1109/jlt.2026.3658374}
}
EUVIMED