Multi-Scenario mmWave Localization Using KNN Graphs and Graph Neural Networks
Graph-based wireless localization framework for multi-scenario mmWave transmitter localization by KNN graphs and GraphSAGE style Graph Neural Network. The model achieves scalable and memory-efficient training on the synthetic DeepMIMO Campus3p5 dataset (MAE 0.73, RMSE 1.18), and high level of generalisation for both synthetic and measured mmWave environments on a real-world Effective Channel dataset (R² 0.9917).
Authors
- Aysha Saddiqa
- Asad Ali (ORCID: https://orcid.org/0009-0001-1493-7684)
Institutions
- University of Greenwich (GB)
Publication Details
- Journal
- Zenodo (CERN European Organization for Nuclear Research)
- Published
- 2026-05-18
- DOI
- https://doi.org/10.5281/zenodo.20264594
- Primary Topic
- Indoor and Outdoor Localization Technologies
- Type
- article
- Field-Weighted Citation Impact
- 0.00