QoS-Aware Resource Placement for the LEO Edge

Placement algorithms, simulation environment, and analysis tools for QoS-aware resource placement on LEO satellite constellations.

If you use this software in a publication, please cite it as:

Text

T. Pfandzelter and D. Bermbach, QoS-Aware Resource Placement for LEO Satellite Edge Computing, 6th IEEE International Conference on Fog and Edge Computing (ICFEC 2022), Taormina, Italy, 2022

BibTeX

@inproceedings{pfandzelter_optimal:_2022,
    title = "QoS-Aware Resource Placement for LEO Satellite Edge Computing",
    booktitle = "6th IEEE International Conference on Fog and Edge Computing (ICFEC 2022)",
    author = "Pfandzelter, Tobias and Bermbach, David",
    year = 2022,
    publisher = "IEEE",
}

A preprint is available on arXiv. For a full list of publications, please see our website.

License

The code in this repository is licensed under the terms of the MIT license.

All code in the simulation folder is based on the SILLEO-SCNS project and licensed under the GNU General Public License Version 3.

Usage

This project requires Python 3.9 or later. Install the required packages with pip install -r requirements.txt, a virtual environment is recommended. Alternatively, use the provided Dockerfile and run all subsequent commands inside the container:

docker build -t py .
docker run --rm -it -v "$(pwd)":/run py
cd /run

Please bear in mind that the code in this repository is not optimized for performance. Memory, disk, and CPU usage can be high, especially for larger constellations and with increased simulation granularity. The simulations in our paper were performed on an r6g.16xlarge EC2 instance with a 200GB SSD in the AWS Frankfurt region.

Configuration

Configure simulation parameters in config.py. This includes parameters for the shells as well as service level objectives (SLO).

Placement

The placement algorithm can be found in placement.py. To run the algorithm for the configured shells and SLOs, run python3 shells.py.

Constellation Simulation

The constellation simulation yields ISL distances for the configured shells. Run python3 distances.py to generate these distances.

SLO Analysis

Analyze results with python3 placement-distances.py.

Summarize

To summarize the results, run python3 summarize.py. Results are then available in results.csv (depending on your configuration in config.py).

Visualize

To generate graphs for this information, run python3 graphs.py.