Development of an SDR-Based System for UHF RFID Signal Analysis

Abstract

This thesis presents the development of a Software-Defined Radio based system for capturing and analyzing Ultra High Frequency (UHF) RFID signals. While UHF RFID has been explored as a sensing modality, commercial readers lack the transparency required to analyze the underlying signal and any interference taking place. A custom Software-Defined Radio based system was implemented to capture, detect, and decode RFID backscatter while providing access to the raw In-phase/Quadrature data. The system utilizes a derivative-based metric to detect RFID backscatter and support tag decoding. The system was used to analyze RFID signals in scenarios involving multiple nearby tags, including cases where one tag was undergoing controlled motion. The results indicate significant interference between nearby tags. Stationary tags were observed to exhibit "motion-like" periodic variations in phase and RSSI when in close proximity to a moving tag. This often appeared as an inverted reflection of the actual motion, creating ambiguity that makes it difficult to interpret tag behavior on a per-tag basis. This suggests that RFID systems designed for motion sensing may require fundamentally different signal processing chains than commercial systems designed for rapid retrieval of tag information.