Skip to main content Skip to main navigation

Publication

Interoperability and AI-Driven Enhancements of BLE-Based Localization in Upcoming 6G Subnetworks

Matthias Rüb; Jan Herbst; Benjamin Rother; Fabian Kummer; Frank Golatowski; Ahmed Eid; Clemens Möllenhoff; Florian Herman; Dennis Salzmann; Thomas Neumuth; Hans Dieter Schotten
In: Proceedings of the 8th International Conference on Information and Communications Technology 2025. International Conference on Information and Communications Technology (ICOIACT-2025), December 4, ISBN 979-8-3315-5408-8, IEEE, 2025.

Abstract

The research of future wireless systems of the sixth generation (6G) envision a "network-of-networks" paradigm, seamlessly interconnecting heterogeneous wireless, sensing, and computing infrastructures as a key enabler for many different technologies, such as future smart hospital ecosystems. Within this framework, two complementary Bluetooth Low Energy (BLE)-based localization modalities, RSSI-driven multilateration and Angle-of-Arrival (AoA), are evaluated under a unified testbed to satisfy the precision and robustness requirements of indoor healthcare environments. Both modalities undergo initial calibration via BLE fingerprinting using a multilayer perceptron (MLP) trained on high-precision Vicon motion-capture ground truth. External calibration reduces the AoA system's mean Euclidean error from 0.58 m to 0.48 m and the RSSI multilateration's mean error from 1.68 m to 0.95 m. To eliminate dependence on costly motion-capture systems, raw AoA measurements (0.46 m median error) subsequently serve as virtual ground truth to retrain the RSSI model, yielding a median error of 1.39 m. This two-stage architecture delivers fail-safe redundancy: in the event of AoA hardware unavailability, the AI-calibrated RSSI system maintains acceptable positioning performance, even with only automated calibration with no additional cost. Finally, a "bouquet" strategy is proposed for 6G subnetworks, combining independently calibrated localization techniques into a robust, scalable, and energy-efficient indoor positioning infrastructure for next-generation healthcare environments.

Projects

More links