Interpretable Physics Informed WiFi Indoor Localization: Learning an Effective Access Point Geometry and Using It to Prune

2026-09-26

AI Summary

This study addresses the limited interpretability and difficulty of feature reuse in deep learning-based indoor localization models by proposing a physics-informed hierarchical deep learning framework. Without requiring ground-truth access point (AP) coordinates, the method jointly predicts user locations and infers AP geometric structures from received signal strength indicator (RSSI) measurements. A physics-informed decoder guides Fisher information matrix-based AP pruning, thereby unifying interpretability with principled feature selection. Experimental results demonstrate that the proposed approach reduces 3D localization error to 7.07 meters, achieving a 26%–36% improvement over baseline methods. Notably, pruning 50% of APs incurs only marginal accuracy degradation, substantially enhancing both the efficiency and transparency of WiFi fingerprinting-based localization systems.

Abstract

Deep learning models can achieve high accuracy for indoor localization, but their black-box nature limits interpretability and the reuse of learned information. We propose a hierarchical deep learning framework for WiFi fingerprint-based indoor localization that jointly predicts user location and learns an effective geometry of the surrounding access points (APs). Physics-informed decoders infer this geometry directly from RSSI measurements and labelled user positions, without requiring the true AP coordinates during training. The learned geometry is then used to rank and prune APs. On the UJIIndoorLoc dataset, the proposed chained model achieves a mean 3D localization error of 7.07 m, reducing error by 26% to 36% compared with baseline models. Previously published methods evaluated on the same official split report errors 10.6% to 31.0% higher. Pruning 35% or 50% of the APs causes only a small loss in localization accuracy. The inferred geometry also enables Fisher-information-based AP ranking even when fingerprint databases do not contain surveyed AP coordinates. Experiments on the Tampere/TUT and UTSIndoorLoc datasets show that geometry-guided AP selection performs comparably to selectors built directly from labelled data. These results show that physics-informed interpretability can improve indoor localization while also supporting effective feature selection.

Problem

• Indoor Localization
• Interpretability
• WiFi Fingerprinting
• Access Point Selection
• Deep Learning

Innovation

• Physics-Informed Deep Learning
• Indoor Localization
• Access Point Geometry
• AP Pruning
• Interpretable AI