Benchmark suite for dissertation "Map matching methods for assigning vehicle positions to road networks"
Details
Publikationsdetails
- Einrichtung
- ErstellerIn
- Verlagsangabe
- Otto-Friedrich-Universität Bamberg
- Erscheinungsjahr
- 2026
- Erstellungsdatum
- 22.09.2026
- Sprache(n)
- Englisch
- DOI
- ist Zusatz zu
- Abstract
Map matching methods for assigning vehicle positions to road networks
Complete benchmark suite for the thesis
The benchmark was initially run on a dedicated server with 2x 64 cores AMD Epyc 7742 with 1024 GB of DDR4 system memoryon local PCIe 3.0 NVMe storage.
This benchmark suite contains three major scenarios:
- The data_set_benchmarks for an extensive comparison benchmark of our Map Matching 2 solution against multiple third-party solutions.
- The memory_profiling suite for analysing the memory consumption and memory types of our Map Matching 2 solution over time.
- An analysis set of tools for preparing some data for figures in our thesis.
- The simulation folder contains a copy subset of the SUMO simulation from the data_set_benchmarks prepared and documented for individual usage.
For creating a virtual python environment on the host system, use:
python3 -m venv venv
source venv/bin/activate
pip install -r benchmarks/data_set_benchmarks/requirements.txt
Any *.sh scripts should be run from this python venv and within the respective directories they are located in.
This benchmark suite is licensed under AGPL 3.0.
- Dokumenttyp
- software
- DDC-Klassifikation
- 004 Computer science
- Schlagwort(e)
- benchmarkmap matchinggeographical information scienceopen sourceground truthdata setsimulation
- Umfang
- 86.2 MB
- Format
- application/zip