Document Type

Journal Article

Publication Date

2026

Keywords

Optical fibers, Distance measurement, Scattering, Textile fibers, Spatial resolution, Fingerprint recognition, Matrices, Windows, Frequency, Temperature

DOI

10.1109/JLT.2026.3722484

Abstract

We propose an image-assisted optical fiber identification method that exploits the unclonable Rayleigh backscattering signatures measured by optical frequency-domain reflectometry (OFDR). A two-dimensional binary optical fiber identifier (OFID), analogous to a digital fingerprint image, is derived from the Rayleigh backscattering spectrum (RBS) matrix, enabling identification when combined with structure similarity index measure (SSIM). Experiments show that the proposed scheme offers 109% improvement in robustness over conventional one-dimensional local-RBS matching, while reducing data storage from 0.98 MB to 39.6 kB per measurement. A systematic analysis of OFDR parameters reveals that identification accuracy is maximized at a spatial resolution of approximately 3 cm. Environmental tests further demonstrate stable performance under global temperature variations from 30 to 70 °C and local strain perturbations up to 240 με, with intra-fiber SSIM consistently remaining in the 0.54-0.75 range. The results confirm that the proposed two-dimensional Rayleigh fingerprint provides a compact, unclonable, and highly robust physical-layer identifier suitable for practical optical fiber authentication.

Source Publication

Journal of Lightwave Technology

Volume Number

44

Issue Number

19

Download
Link to Publisher’s Version

Share

COinS