Dataset and MATLAB Implementation for an Artificial Neural Network Surrogate Model for Q-Factor Estimation in Free-Space Optical Links Under Multi-Parameter Variation

This repository contains the dataset and MATLAB implementation supporting the manuscript "A Fast Artificial Neural Network Surrogate Model for Q-Factor Estimation in Free-Space Optical Links Under Multi-Parameter Variation." The dataset consists over 1,400 OptiSystem-generated simulation samples covering multiple combinations of transmission bit rate, transmitter aperture diameter, receiver aperture diameter, transmission range, and optical filter type. Each sample includes the corresponding simulated Q-factor. The repository also includes the exported MATLAB neural network prediction function, trained model, and supporting documentation required to reproduce the surrogate-model predictions described in the manuscript. The dataset is intended to support reproducibility, verification of reported results, and future research in machine-learning-assisted design of free-space optical communication systems.