The Optimization for Hyperbolic Positioning of UHF Passive RFID Tags

This paper presents a fine-grained positioning method for radio frequency identification (RFID). The proposed method applies hyperbolic positioning to locate ultrahigh frequency passive RFID tags. In our design, finding the tagged object’s location is formulated as an optimization problem. Phase values, collected by the moving antenna, are exploited to achieve the optimal solution. The intuition of hyperbolic positioning lies in that the difference of distances from a target tag to two antennas can be inferred from phases. When integrating hyperbola curves together, optimization method can be performed to achieve the object’s location. Particle swarm optimization is then applied to enhance computational ability. For random phases, polynomial regression is employed to model the relationship between phase values and distances. We implement a prototype of hyperbolic positioning optimization to pinpoint the RFID tag’s location and evaluate its performance in our laboratory environment. Compared with other RFID localization methods, i.e., received signal strength indicator based and phase-based approaches, our design shows greater tolerance to tag diversity and tag orientation.Note to Practitioners—Radio frequency identification (RFID), as a crucial component of the Internet of things, has been widely deployed in many industrial applications ranging from logistics to retail. There are increasing needs for RFID to not only detect whether the tagged object is within the radio range, but also locate the object with high accuracy. Current RFID systems addressing these problems are either expensive or complex to implement. The RFID-based localization method presented in this paper is cost-efficient and computationally inexpensive. We consider using commercial off-the-shelf RFID devices without deployment of the reference tags to locate the target object. The antenna is mounted on a mobile vehicle to emulate the antenna array. Due to the fact that the target tag lies on the hyperbolic curve with two antennas’ positions as the foci, this paper suggests constructing multiple hyperbolas to determine the target position based on the virtual antenna array. The experimental results demonstrate that the proposed approach is feasible and achieves a median accuracy of 12 cm.