This is the code of our paper "Self-play learning strategies for resource assignment in Open-RAN networks".
Please use different branches for training, testing (validation), and comparison with heuristic methods.
bin_packing_various_height: training.
validate_bin_packing: validation of trained models.
vanilla-MCTS: vanilla MCTS method. This branch also contains a file for generating .gif files of bin packing results (for demo).
HVRAA-heuristics: heuristic method.
lego-heuristic: heuristic method.
----Please ignore----
master: not in use.
bin_packing_single_height: training using single height samples. This is not what we did in the paper.
@article{wang2022self,
title={Self-play learning strategies for resource assignment in Open-RAN networks},
author={Wang, Xiaoyang and Thomas, Jonathan D and Piechocki, Robert J and Kapoor, Shipra and Santos-Rodr{\'\i}guez, Ra{\'u}l and Parekh, Arjun},
journal={Computer Networks},
pages={108682},
year={2022},
publisher={Elsevier}
}