Generalizable Indoor Path Loss Prediction
This paper is presented in the context of the First Indoor Pathloss Radio Map Prediction Challenge at IEEE ICASSP 2025. We propose a deep learning approach using a customized ResUNet architecture with physics-informed features for predicting radio maps in indoor environments. Our architecture progressively adapts to handle increasing task complexity, incorporating dual-stream processing for frequency generalization and specialized antenna gain processing with dilated convolutions. Experimental results demonstrate effective generalization across unknown indoor scenes, frequencies, and antenna patterns while maintaining computational efficiency.
Authors
- Oumaya Baala (ORCID: https://orcid.org/0000-0002-7247-7874)
- François Spies (ORCID: https://orcid.org/0000-0002-9964-2745)
- Cheick Tidiani Cissé (ORCID: https://orcid.org/0009-0001-9984-8065)
- Alexandre Caminada (ORCID: https://orcid.org/0000-0002-4088-7406)
- Valéry Guillet
Institutions
- Centre National de la Recherche Scientifique (FR)
- Institut national de recherche en sciences et technologies du numérique (FR)
- Orange (France) (FR)
- Franche-Comté Électronique Mécanique Thermique et Optique - Sciences et Technologies (FR)
- Université de technologie de belfort-montbéliard (FR)
Publication Details
- Published
- 2025-03-12
- DOI
- https://doi.org/10.1109/icassp49660.2025.10887696
- Citations
- 1
- Primary Topic
- Indoor and Outdoor Localization Technologies
- Type
- article
- Field-Weighted Citation Impact
- 0.73