Learning Multi-Agent Coordination for Enhancing Target Coverage in Directional Sensor Networks

This repository is the official implementation of Learning Multi-Agent Coordination for Enhancing Target Coverage in Directional Sensor Networks.

📋 Project website: HiT-MAC

Environment

If you want to try your own algorithm in our environment, here is a pure and non-hierarchical DSN environment for you.

Requirements

To install requirements:

pip install -r requirements.txt

Training

To train the executor in the paper, run this command:

python main.py --env Pose-v0 --model single-att --workers 6

To train the coordinator in the paper, run this command:

python main.py --env Pose-v1 --model multi-att-shap --workers 6

Evaluation

To evaluate my model, run:

python main.py --env Pose-v1 --model multi-att-shap --workers 0 --load-coordinator-dir trainedModel/best_coordinator.pth --load-executor-dir trainedModel/best_executor.pth

You can use trained models directly from the folder "trainedModel".

Citation

If you found this work useful, please consider citing:

@article{xu2020learning,
  title={Learning Multi-Agent Coordination for Enhancing Target Coverage in Directional Sensor Networks},
  author={Xu, Jing and Zhong, Fangwei and Wang, Yizhou},
  journal={Advances in Neural Information Processing Systems},
  volume={33},
  year={2020}
}

Contact

If you have any suggestion/questions, get in touch at jing.xu@pku.edu.cn