FSO x ML

Free-Space Optics (FSO) meets Machine Learning (ML) — a project exploring the impact of turbulence on Optical Vortex (OAM) modes and leveraging computational techniques to analyze and mitigate distortions in FSO communication systems.

Overview

Free-Space Optical (FSO) communication is a wireless technology that uses laser beams to transmit data through the atmosphere. However, atmospheric turbulence introduces phase distortions, affecting signal integrity — especially for Orbital Angular Momentum (OAM) modes. This project simulates turbulence effects, applies attenuation models, and visualizes OAM fields to study their resilience in turbulent environments.

Features

• OAM Field Simulation – Generates helical wavefronts with topological charge l.
• Kolmogorov Turbulence Modeling – Simulates phase distortions due to atmospheric turbulence.
• Attenuation Effects – Applies a Gaussian radial attenuation to model beam propagation losses.
• Field Visualization – Plots phase and intensity distributions for analysis.
• Customizable Parameters – Allows user-defined OAM charge, grid size, turbulence strength, and attenuation factors.

Installation

Ensure you have Python installed, then install the dependencies:

pip install numpy matplotlib

Installing JAX (for Apple Silicon or GPU acceleration)

If you’re using an Apple Mac with an M1/M2/M3 chip and want to utilize the Metal GPU backend:

pip install jax jaxlib –no-cache-dir

For CUDA (NVIDIA GPU) acceleration, install JAX with CUDA support (ensure CUDA and cuDNN are properly set up):

pip install jax jaxlib -f https://storage.googleapis.com/jax-releases/jax_cuda_releases.html

For CPU-only usage:

pip install jax jaxlib

Installing CuPy (for NVIDIA GPU Acceleration)

If you have an NVIDIA GPU with CUDA installed, install CuPy for GPU-accelerated computations:

For CUDA 12 (Latest)

pip install cupy-cuda12x

For CUDA 11

pip install cupy-cuda11x

Clone the repository:

Run this code to clone the repository

git clone https://github.com/Aravind-Sathesh/FSO-x-ML.git
cd FSO-x-ML

Usage

Run the main script and enter the required parameters when prompted:

python main.py

The script will:

1. Generate an OAM field with a given topological charge.
2. Apply Gaussian attenuation based on the chosen factor.
3. Introduce Kolmogorov or Von Karmen turbulence to simulate atmospheric distortions.
4. Display phase and intensity distributions of the field.

Plots will be saved automatically.

Contributors

• Sruti Guduru – Developer
• Aravind Sathesh – Developer