TSN Scheduler Benchmarking: Results v2.0.1

TSN Scheduler Benchmarking Results

This dataset provides test results of multiple Time-Sensitive Networking (TSN) schedulers that were tested and compared in the following PhD thesis (to be published):

E. Schweissguth, "Routing and Scheduling of Time-Triggered Ethernet Networks: ILP Models and Scheduler Benchmarking.". University of Rostock.

An earlier version of this repository contained results from the following publication:

E. Schweissguth, S. Mehner, D. Hellmanns, J. Falk, H. Parzyjegla, P. Danielis, O. Hohlfeld, G. Muehl, and D. Timmermann, "TSN Scheduler Benchmarking." in WFCS 2023. IEEE.

These results are still available in the history (see releases). Compared to v1.0.0, this version contains results for TSN schedulers that were tested on an updated scenario dataset with a general update to scenario parameters, additional multicast scenarios, and heterogeneous cycle times and frame sizes within a single stream set. Moreover, schedulers used are not by different working groups (as in v1.0.0) but model variations of ILP-based schedulers, which are described in detail in the thesis.

Please refer to the thesis for detailed background information. The results are based on the following scenarios (i.e., topologies and stream sets): "TSN Scheduler Benchmarking: Scenarios v2.0.0" (see https://doi.org/10.5281/zenodo.13993512 or https://github.com/EikeSG/TSNBenchScenarios/releases/tag/v2.0.0). When accessing this dataset through Zenodo, also check the corresponding GitHub repository (https://github.com/EikeSG/TSNBenchResults) for updates and for a good in-browser view (especially for markdown files).

This repository contains all runtime/schedulability/latency plots for all benchmarking tests run throughout the thesis, whereas result plots in the thesis are excerpts from the plots in this repository. See README.md for further details on the directory contents.

Note that scheduler result files are available in json format as github release assets and as separate zenodo dataset (https://doi.org/10.5281/zenodo.13995796). They are not part of this dataset because zenodo does not pick up github release assets of linked repositories. For the same version (here: v2.0.1), the dataset on github and the linked zenodo dataset are identical.