The RFRL Gym is intended as a training and research environment for wireless communications applications designed to provide comprehensive functionality, such as custom scenario generation, multiple learning settings, and compatibility with third-party RL packages. Additionally, through a gamified mode of the RF spectrum, this tool can be used to teach novices about the fields of AI/ML and RF.
| Jamming Agent before Learning | Jamming Agent after Learning |
|---|---|
 |
 |
Note: Pardon our mess as this project is under active development. Please let us know of any feature requests or bugs to be squashed!
- Install necessary prerequist software using the terminal:
sudo apt install python3 python3-pip python3-venv python3-wheel
- Set up a Python virtual environment in the root directory of the repository:
python3 -m venv rfrl-gym-venv
- Ensure that venv is fully updated:
python3 -m venv --upgrade rfrl-gym-venv
- Activate the virtual environment (you will need to do this everytime you being working with the repository in a new terminal):
source rfrl-gym-venv/bin/activate
- Install setuptools:
pip3 install pip wheel setuptools --upgrade
- Install the repository:
pip3 install --editable .
python3 scripts/preview_scenario.py -m abstract
python3 scripts/preview_scenario.py -m iq
A terminal output should print out showing the observation space upon successful execution.
pip3 install -e ".[rl_packages]"
python3 scripts/sb3_example.py -m abstract
python3 scripts/sb3_preview_scenario.py -m abstract
@inproceedings{rfrlgym,
Title = {{RFRL Gym: A Reinforcement Learning Testbed for Cognitive Radio Applications}},
Author = {D. Rosen, I. Rochez, C. McIrvin, J. Lee, K. D’Alessandro, M. Wiecek, N. Hoang, R. Saffarini, S. Philips, V. Jones, W. Ivey, Z. Harris-Smart, Z. Harris-Smart, Z. Chin, A. Johnson, A. Jones, W. C. Headley},
Booktitle = {{IEEE International Conference on Machine Learning and Applications (ICMLA)}},
Year = {2023},
Location = {Jacksonville, USA},
Month = {December},
Url = {}