Topology synthesis is a computationally-intractable task for large-scale ad-hoc robot networks due to various compelxities, such as Hamiltonian analyses of their graphs, whose solutions do not efficiently scale as those graphs expand. So, the idea is the prediction of the optimal topology of ad-hoc robot networks given a particular configuration using a data-driven model. This model, i.e., OpTopNET [1], is a stacked enseamble based on the ground-truth topology information computed by this repository's code. For more information about OpTopNET and the mathematical grounds of this code, one may refer to [1].
If you would like to, totally or partially, use this code, please cite the following paper:
@article{macktoobian2023learning,
title={Learning optimal topology for ad-hoc robot networks},
author={Macktoobian, Matin and Shu, Zhan and Zhao, Qing},
journal={IEEE Robotics and Automation Letters},
year={2023},
publisher={IEEE}
}