Re-fingerprinting RFIDs with augmented unclonability

Abstract

Radio Frequency IDentification (RFID) technology has become increasingly prevalent in modern society, with its applications ranging from logistics and warehouse management to electronic passports. This dissertation focuses on the application of fingerprint technology for UHF passive RFID tags, which represent the most widely used type of RFID technology. The key challenges and opportunities addressed by this dissertation relate to two different applications for RFID fingerprinting systems, namely authentication and localization. The primary aim of this research is to bridge the gap between research and real-life usage by developing new algorithms and designs that can enhance the performance of RFID fingerprinting systems. This dissertation presents three fundamental contributions: two for the authentication fingerprinting application and one for the localization fingerprinting application. Previous authentication fingerprinting approaches have commonly suffered from poor performance or a lack of evaluation with a relatively large dataset. To address this issue, we propose two new approaches. The first approach, TagDNA, takes advantage of the frequency-agnostic phenomenon of commercial RFID tags to acquire ten times more features than previous fingerprints. The second approach, BFD, is defined as the offset between the frequency of the backscatter signal received from a tag and the requested backscatter link frequency (BLF). Both fingerprinting approaches are evaluated using an automatic fingerprint collection system, which collects fingerprints from over 16,000 and 7,000 tags with ten different models, respectively. In the context of fingerprinting systems for localization, we focus on the aspect that relatively lacks attention, namely the arrangement of the antenna array. To optimize the commonly used uniform planar array, we propose a novel sparse antenna array, LSAB, which deploys antennas on a log-spiral-shaped belt in a non-linear manner, following the theory of minimum resolution redundancy that we newly discovered in this work. To deliver these contributions, we model the RFID communication system, and propose new algorithms and designs, from the hardware placements to software algorithms. We implement prototypes to evaluate these algorithms and designs, and how they perform in different real-world scenarios, including large-scale RFID authentication and library RFID localization. To achieve these contributions, we develop a comprehensive model of the RFID communication system and propose novel algorithms and designs that encompass both hardware placements and software algorithms. Subsequently, we create prototypes to evaluate the performance of these solutions in various real-world scenarios, such as large-scale RFID authentication and library RFID localization. Our experimental results demonstrate the effectiveness and practicality of the proposed fingerprinting techniques and antenna array design, as well as their ability to address the challenges and opportunities associated with RFID fingerprinting systems. Overall, this dissertation aims to bridge the gap between research and practical applications of RFID fingerprinting, and provides valuable insights into the development of efficient and reliable RFID-based systems.