A Streamlined Two-Antenna Architecture for Passive RFID Phase Localization via Accelerometer-Based Yaw Disambiguation

Recent advances in energy harvesting have enabled low-cost passive RFID tags to power simple sensors such as humidity meters and tri-axis accelerometers, offering observables beyond RSSI and carrier phase. However, most existing methods either rely solely on RSSI or on phase, or fuse active radios with power-hungry IMUs, which undermines RFID’s core advantages of low cost and easy deployment. To explore the potential of passive tags, this paper designs a measurement setup and an algorithm tailored for an embedded, battery-free 3-axis accelerometer RFID tag. The approach exploits (i) changes in the gravity projection induced by target pose, and (ii) an approximately linear relation between tag orientation and backscatter phase. Fitting the two-antenna phase model to the measured phases via a Gauss–Newton nonlinear leastsquares optimizer, a centimeter-level, step-wise position tracking under planar motion constraints is acieved, while maintaining a simple hardware configuration and reduced algorithmic complexity. The approach is targeted at mechanically guided scenarios (e.g., doors, rails, panels) where coplanarity can be enforced by design. Under similar geometric constraints, the method achieves comparable accuracy while reducing hardware complexity and deployment requirements.