Range and Bearing Estimation of an UHF-RFID Tag Using the Phase of the Backscattered Signal

The narrow bandwidth of UHF RFID signals does not allow the direct measurement of the range between the reader and the tag. A Multi-Hypothesis Extended Kalman Filter is proposed in this paper to solve this issue, by fusing the phase measurements with the odometry readings of a mobile robot. While estimating the tag-reader distance, the proposed approach provides also an on-line estimation of the bearing of the tag. The knowledge of range and bearing is beneficial in positioning-based applications as in robotics, to develop algorithms for reaching and grasping a tagged object or to localize a robot using tags as anchors, or for solving a Simultaneous Localization and Mapping (SLAM) problem if positions of both tags and robot are unknown. A few steps are usually enough to obtain a univocal effective estimate of these quantities, as long as the robot is moving over a non-straight path. A comprehensive numerical analysis supports and assesses the theory.