RFID Indoor Positioning and Navigation Using a Regularized Particle Filter Integrated with a Probability Model

Hui Tang and Don Kim

Abstract:	For indoor mobile robot applications, we have developed active radio frequency identification (RFID) navigation system integrated with speed sensors and inertial sensors. Due to multipath under indoor environments, signal strength variation for a certain range can be very large. Therefore, it may not be an efficient way to build an analytic signal-path-loss observation model. In this paper, we proposed a non-parametric probabilistic observation model to handle uncertainties and errors in signal strength measurements. To deal with arbitrary measurement noise, a regularized particle filter (RPF) was investigated for indoor positioning. This approach has been validated through field experiments. We found that the noise characteristics of signal strength measurements in a typical indoor environment show a dual-mode feature. Also, we validated the advantages of the RPF employed for position estimation, including: 1) it is able to handle arbitrary noise distribution, 2) it can do positioning without initialization, and 3) positioning solution converges quickly when measurements are available. The accuracy of positioning solutions in kinematic tests was 1.64 m ± 1.03 m (1).
Published in:	Proceedings of the 24th International Technical Meeting of the Satellite Division of The Institute of Navigation (ION GNSS 2011) September 20 - 23, 2011 Oregon Convention Center, Portland, Oregon Portland, OR
Pages:	3470 - 3479
Cite this article:	Tang, Hui, Kim, Don, "RFID Indoor Positioning and Navigation Using a Regularized Particle Filter Integrated with a Probability Model," Proceedings of the 24th International Technical Meeting of the Satellite Division of The Institute of Navigation (ION GNSS 2011), Portland, OR, September 2011, pp. 3470-3479.
Full Paper:	ION Members/Non-Members: 1 Download Credit Sign In