Machine Learning Models to Identify Promising Nested Antiresonance Nodeless Fiber Designs

TL;DR

This paper introduces a machine learning framework that efficiently predicts high-performance nested antiresonance nodeless fiber designs, significantly reducing computational costs and enabling large-scale exploration of fiber geometries.

Contribution

A novel two-stage machine learning approach that accurately predicts fiber confinement loss using minimal training data, facilitating rapid design optimization of complex fiber structures.

Findings

Successfully identified fiber designs with CL of 0.25 dB/km

Demonstrated stable predictions with small datasets of 1,819 samples

Enabled exploration of 14 million design cases efficiently

Abstract

Hollow-core fibers offer superior loss and latency characteristics compared to solid-core alternatives, yet the geometric complexity of nested antiresonance nodeless fibers (NANFs) makes traditional optimization computationally prohibitive. We propose a high-efficiency, two-stage machine learning framework designed to identify high-performance NANF designs using minimal training data. The model employs a neural network (NN) classifier to filter for single-mode designs (suppression ratio $\ge$ 50 dB), followed by a regressor that predicts confinement loss (CL). By training on the common logarithm of the loss, the regressor overcomes the challenges of high dynamic range. Using a sparse data set of only 1,819 designs, all with CL greater or equal to 1 dB/km, the model successfully identified optimized designs with a confirmed CL of 0.25 dB/km. {This demonstrates the NN has captured underlying physical behavior and is able to extrapolate to regions of lower CL. We show that small data sets are sufficient for stable, high-accuracy performance prediction, enabling the exploration of design spaces as large as $14e6$ cases at a negligible computational cost compared to finite element methods.