CALNNLS: A Cost-Effective Approach for UWB Anchor Calibration
IEEE Sensors Journal · 2026-08 · 26 (16) · 24991-25005
DOI: 10.1109/jsen.2026.3710635 ↗Kurz gefasst
Fachartikel in der Zeitschrift IEEE Sensors Journal, 2026-08. Verfasst von Sancho Amaral Simoes und 2 weiteren Autor:innen. Die Arbeit nennt 31 Literaturangaben. Der Volltext ist frei zugänglich (Open Access, CC BY 4.0). Gefördert von: “Sustainable Stone by Portugal” funded by the PRR—Recovery and Resilience Plan and the Next Generation EU European, Component 5---Capitalization and Business Innovation---Mobilizing Agendas for Business Innovation, FCT—Funda˜cc˜ao para a Ciˆencia e a, Instituto Público.
Abstract (Originalsprache)
Ultra-wideband (UWB) indoor localization systems are increasingly employed in robotics, industry, and asset tracking due to their high accuracy and robustness in multipath-prone environments. Like GPS, UWB systems require precise positions of fixed reference points, called anchors, for accurate object localization. The process of determining anchor positions, known as anchor calibration, directly impacts location accuracy through trilateration, as any misplacement propagates into significant errors. Localization inaccuracies stem not only from measurement noise and non-line-of-sight (NLOS) conditions but also fundamentally from these initial calibration errors. This article presents a novel anchor calibration approach, CALNNLS, based on a feedforward neural network followed by nonlinear least squares (NLS), both using a loss function that incorporates multiple geometrical constraints. Evaluated in real scenarios with 3–11 DECAWAVE UWB anchors, CALNNLS outperforms the manufacturer’s built-in calibration and most state-of-the-art methods, achieving root-mean-square error (RMSE) improvements ranging from 60.9% to 80.0% for calibration and 75.0% to 82.5% for localization. In addition, CALNNLS is cost-effective, scalable, and simple, requiring no additional hardware and minimal user intervention, making it suitable for UWB deployments. We also propose a theoretical extension of CALNNLS to handle more complex spatial configurations.
Förderung
- “Sustainable Stone by Portugal” funded by the PRR—Recovery and Resilience Plan and the Next Generation EU European
- Component 5---Capitalization and Business Innovation---Mobilizing Agendas for Business Innovation
- FCT—Funda˜cc˜ao para a Ciˆencia e a, Instituto Público
Zitieren
APA
Simoes, S. A., Araujo, H., & Abreu, P. H. (2026). CALNNLS: A Cost-Effective Approach for UWB Anchor Calibration. IEEE Sensors Journal, 26(16), 24991-25005. https://doi.org/10.1109/jsen.2026.3710635
Vancouver
Simoes SA, Araujo H, Abreu PH. CALNNLS: A Cost-Effective Approach for UWB Anchor Calibration. IEEE Sensors Journal. 2026;26(16):24991-25005. doi:10.1109/jsen.2026.3710635
BibTeX
@article{Simoes2026,
author = {Simoes, Sancho Amaral and Araujo, Helder and Abreu, Pedro Henriques},
title = {CALNNLS: A Cost-Effective Approach for UWB Anchor Calibration},
journal = {IEEE Sensors Journal},
year = {2026},
volume = {26},
number = {16},
pages = {24991--25005},
doi = {10.1109/jsen.2026.3710635}
}
EUVIMED