This work presents an indoor Local Positioning System (LPS) based in the combination of active RFID technology and Bayesian techniques for positioning estimation, from the signal strength of the received RF signals. The complexity of indoor propagation of RF waves causes large fluctuations in the signal strength levels, which can be handled better by statistical Bayesian techniques, than by more common methods like multilateration, quadratic minimization, or fingerprinting. In the empirical validation of our RFID-LPS system we achieved an estimate of the user's location with an average error of 2.10 m, median value of 1.84 m, and 3.89 m for 90% of the cases, in a displacement area of 475 m2 (with 29 RFID tags), and with velocities up to 0.5 m/s; this performance is similar or improves the state of the art of this kind of positioning systems. Even though there exists a background in the field of Robotic Navigation, the combination of Bayesian methods and active RFID technology presented in this work is original within the framework of location systems for people, whose movements are less predictable than those of robots. Other novel aspects investigated are the possibility of joint estimation of the position and the orientation of the user, with two different techniques (use of directive antennas and employing the attenuation of RF signals by the human body), the system scalability, and the capacity of position estimation by mere detection of RFID tags, without signal strength measurements. © 2013 CEA.
Seco, F., Koutsou, K., Ramos, F., & Jimenez, A. R. (2013). Localizacion personal en entornos interiores con tecnologia RFID. RIAI - Revista Iberoamericana de Automatica e Informatica Industrial, 10(3), 313–324. https://doi.org/10.1016/j.riai.2013.05.004