Rayleigh scattering-based distributed sensing in multicore optical fibers for shape reconstruction in multiplanar disturbance

In recent years, advancements in smart cities and multifunctional monitoring have driven new demands for shape reconstruction devices. Multicore fibers (MCFs) are increasingly used due to their parallel data transmission capabilities and multiparameter sensing potential. This paper presents a Rayleigh-scattering-based distributed sensing approach with MCFs for shape reconstruction under multiplanar disturbances. Optical Frequency-Domain Reflectometry (OFDR) is utilized to analyze four cores in a seven-core fiber, enabling shape reconstruction through cross-correlation of spectral responses relative to an unstrained reference. Two configurations are compared: Configuration 1, a spatially separated core arrangement enabling independent strain independent strain measurements to be analyzed with a Frenet-Serret frame modeling and Random Forest (RF) algorithms, achieving a high accuracy with a maximum error of 8.72 × 10⁻³ cm; and Configuration 2, a simplified approach analyzing all cores in the same OFDR channel, yielding a higher error of 0.27 cm when a RF algorithm was fed with features derived from the Rayleigh backscattered signal (loss amplitude, strain, spectral shift, and s- and p-polarization) measured by an optical backscatter reflectometer. Three protocols—pure bending, pure torsion, and combined bending/torsion—validate these configurations. The results emphasize this multifunctional MCF-based sensor system’s potential for flexible, integrated shape reconstruction solutions.