README.md

Welcome to the CLOUD-D RF Codebase!

This codebase exists as an implementation of the results of our paper, CLOUD-D RF: Cloud-based Distributed Radio Frequency Heterogeneous Spectrum Sensing.

Installation and Usage Instructions

To install the codebase, we recommend using pip:

pip install .

Generating plots from the paper

1. Generate training, validation, and testing data:

   python scripts/gen.py

2. Train individual team models:

   python scripts/train.py

3. Train baseline fusion model:

   python scripts/baseline_fusion.py

4. Train RL and RFE fusion models:

   python scripts/rl_rfe_fusion_and_analysis.py

5. Generate plots:

   python feature_contribution_analysis.py

   python generate_confusion_matrices.py

   python plot_accuracy_vs_nuisance_params.py

Running in Amazon SageMaker

Please refer to the sagemaker branch (https://github.com/vtnsi/cloudd-rf/tree/sagemaker) to run in Amazon SageMaker.

Contributors

Name	Role	Title	Email
Caleb McIrvin	Developer	PhD Student, Spectrum Dominance Division, Virginia Tech National Security Institute	[email protected]
Dylan Green	Developer	Masters Student, Spectrum Dominance Division, Virginia Tech National Security Institute	[email protected]
Alyse M. Jones	Developer	Research Associate, Spectrum Dominance Division, Virginia Tech National Security Institute	[email protected]
Maymoonah Toubeh	Developer	Research Assistant Professor, Spectrum Dominance Division, Virginia Tech National Security Institute	[email protected]
William 'Chris' Headley	Developer	Associate Director, Spectrum Dominance Division, Virginia Tech National Security Institute	[email protected]
Joseph Risi	Developer	AI/ML Specialist Solutions Architect, Amazon Web Services	[email protected]

License

MIT