Sustainable broadcasting in Blockchain Network with Reinforcement Learning

Working paper

Valko, D., & Kudenko, D. (2024). Sustainable broadcasting in Blockchain Network with Reinforcement Learning. arXiv. https://doi.org/10.48550/arXiv.2407.15616

@misc{ValkoKudenko2022,
title={Sustainable broadcasting in Blockchain Network with Reinforcement Learning}, 
author={Danila Valko and Daniel Kudenko},
year={2024},
publisher={arXiv},
howpublished={arXiv},
doi = {https://doi.org/10.48550/arXiv.2407.15616},
}

Setup

BlockSim requires some virtual environment with certain dependencies, see requirements.txt.

conda create -n blocksim python=3.9
conda activate blocksim
pip install -r requirements.txt 

Run

• test simulator

python -m blocksim.blocksim

• run training

source activate blocksim && python train.py --n_envs 4

• run experiments

python exp.py --k 1000

References

• We used BlockSim - a framework for modeling and simulating blockchain protocols (Faria & Correia, 2019), see GitHub.