Institutional Repository
Technical University of Crete
EN  |  EL

Search

Browse

My Space

Real-time, robot-based, 3D localization of RFID tags, by transforming phase measurements to a linear optimization problem

Tzitzis Anastasios, Malama Andreana, Drakaki Vasiliki, Bletsas Aggelos, Yioultsis Traianos, Dimitriou Antonis G.

Full record


URI: http://purl.tuc.gr/dl/dias/3C0BD185-B764-4B3E-B80E-B7A2E39A26AF
Year 2022
Type of Item Peer-Reviewed Journal Publication
License
Details
Bibliographic Citation A. Tzitzis, A. Malama, V. Drakaki, A. Bletsas, T. V. Yioultsis and A. G. Dimitriou, "Real-time, robot-based, 3D localization of RFID tags, by transforming phase measurements to a linear optimization problem," IEEE J. Radio Freq. Identif. (RFID), vol. 6, pp. 439-455, 2022, doi: 10.1109/JRFID.2021.3103393. https://doi.org/10.1109/JRFID.2021.3103393
Appears in Collections

Summary

In this paper, we propose a prototype method for fast and accurate 3D localization of RFID-tagged items by a mobile robot. The robot performs Simultaneous Localization of its own pose and Mapping of the surrounding environment (SLAM). It is equipped with RFID readers and antennas placed at different heights, collecting phase-measurements by all tags. Thanks to the self-localization property of the robot, a synthetic aperture is created for each tag. In this paper, we have manipulated the set of phase-measurements, combined with the known poses of the robot to craft an overdetermined system of linear equations which pinpoints the location of the tag. The linearity of the system preserves that localization is achieved rapidly. The problem is solved for any arbitrary movement of the robot and extended in three dimensions. The proposed method is experimentally compared against the state-of-the-art in SAR-based RFID localization. It is the fastest and the most accurate, when the robot moves along straight-paths. Its accuracy slightly worsens, when the robot moves along non-straight paths, while preserving its exceptionally small running-time. It can be applied in real-time 3D localization problems, demonstrating mean 3D accuracy below 20 cm.

Services

Statistics