EUVIMEDEUVIMED entdecken
Open Access · CC BY 4.0

Road-Constrained UE Trajectory Reconstruction Without GNSS Using Multi-RAT Cellular Measurements

Alexandre Rodrigues, João Martins, Carlos Porrais, David Santos, Sameer Ibraimo, Gabriela Soares, Pedro Sebastião

Future Internet · 15.11.2026 · 18 (11) · 628

DOI: 10.3390/fi18110628 ↗

Kurz gefasst

Fachartikel in der Zeitschrift Future Internet, online erschienen am 15.11.2026. Verfasst von Alexandre Rodrigues und 6 weiteren Autor:innen. Die Arbeit nennt 42 Literaturangaben. Der Volltext ist frei zugänglich (Open Access, CC BY 4.0). Gefördert von: Multivision, European Union.

AbstractBibliografische AngabenZitierenVerwandte Studien

Abstract (Originalsprache)

Reliable cellular trajectory estimates support mobile network optimisation, coverage analysis, and Virtual Drive Testing. Drive Tests provide georeferenced radio measurements but are costly to repeat at scale. Minimization of Drive Tests (MDT) offers a broader source of operational measurements, although accurate coordinates may be unavailable. The proposed road-aware method reconstructs user equipment trajectories without Global Navigation Satellite System (GNSS) input. Global System for Mobile Communications (GSM), Wideband Code Division Multiple Access (WCDMA), and Long-Term Evolution (LTE) measurements are grouped into temporal windows and represented by radio access technology (RAT)-specific features together with mobility and road-network evidence. The candidate road locations form the states of a hidden Markov model (HMM)-style sequence model solved by single-pass Viterbi-style inference. The principal evaluation analyses 348 Drive Test files; all have evaluable predictions. The initial Drive Test data also informed parameter refinement. Evaluation uses dedicated Drive Tests; applicability to commercial MDT reporting remains untested. Device reference coordinate fields are withheld from predictor inputs, while GNSS references are used for parameter development and evaluation. The median symmetric route error is 214.48 m, and the 90th-percentile error is 590.32 m. On average, 71.57% of the reference trajectory lies within 100 m of the reconstruction. A controlled comparison against independent selection uses identical candidates, final emissions and route-assembly rules. These results characterise the available Drive Test dataset.

Förderung

  • Multivision
  • European Union

Zitieren

APA

Rodrigues, A., Martins, J., Porrais, C., Santos, D., Ibraimo, S., Soares, G., & Sebastião, P. (2026). Road-Constrained UE Trajectory Reconstruction Without GNSS Using Multi-RAT Cellular Measurements. Future Internet, 18(11), 628. https://doi.org/10.3390/fi18110628

Vancouver

Rodrigues A, Martins J, Porrais C, Santos D, Ibraimo S, Soares G, et al. Road-Constrained UE Trajectory Reconstruction Without GNSS Using Multi-RAT Cellular Measurements. Future Internet. 2026;18(11):628. doi:10.3390/fi18110628

BibTeX
@article{Rodrigues2026,
  author = {Rodrigues, Alexandre and Martins, João and Porrais, Carlos and Santos, David and Ibraimo, Sameer and Soares, Gabriela and Sebastião, Pedro},
  title = {Road-Constrained UE Trajectory Reconstruction Without GNSS Using Multi-RAT Cellular Measurements},
  journal = {Future Internet},
  year = {2026},
  volume = {18},
  number = {11},
  pages = {628},
  doi = {10.3390/fi18110628}
}
Kostenlos starten →