C. Caiazza, C. Cicconetti, V. Luconi, and A. Vecchio, “Measurement-driven design and runtime optimization in edge computing: Methodology and tools”, Computer Networks (2021), doi: https://doi.org/10.1016/j.comnet.2021.108140, BibTeX

The paper presents the results obtained in the Fed4Fire+ experiment Estimating the Mobile Edge Computing Infrastructure Performance (MECPerf) where we have:

1. defined an architecture to enable run-time and offline decision making driven by network-layer performance measurements
2. developed a reference implementation of the measurement agents, collectively called MECPerf, available as open source on GitHub
3. collected performance measurements with MECPerf in the NITOS testbed of the Fed4Fire+ infrastructure, available on Zenodo

To facilitate access to the data collected, we have also implemented a Python library, on GitHub, which can be integrated with other simulation/emulation to obtain (more) realistic performance, in terms of the network-layer dynamics at the edge.