Machine Learning and OAM Communication System Tutorial

This section provides information for using your own machine learning network to train and test our alphabets of Laguerre-Gaussian beams. Below are the links to the MATLAB live files which will walk you through the process, and access to our datasets to try it yourself. Inside each dataset is a "READ_ME.txt" file which provides the relevant collection information for the dataset. All files are at 224x224 pixels, and can be downloaded via a compressed ".zip" file.

TUTORIAL FILES FOLDER

MATLAB Live Files

OAM_Classification_Tutorial.mlx

OAM_Classification_Tutorial_TransferLearning.mlx

Access to Datasets

What's included in the folder:

1. Easiest (RB Turbulence) 16-symbol alphabet w/ 10,000 images of each beam

2. Easy (RB Turbulence) 16-symbol alphabet w/ 1,000 images of each beam

3. Moderate (RB Turbulence) 256-symbol alphabet w/ 1,000 images of each beam

4. Hard (Thermal Point Source) 16-symbol alphabet w/ 1,000 images of each beam

5. Hardest (Thermal Point Source) 256-symbol alphabet w/ 1,000 images of each beam