Rapid RFID location and orientation recovery

Abstract

This thesis aims to enable virtual and augmented reality (VR/AR) systems to track objects accurately through occlusions via RFID localization. Currently, three major obstacles prevent the use of RFID localization in VR/AR systems: (1) there exists a trade-off between measurement speed and ability to deal with multipath, so systems which can produce accurate results either require a highly constrained environment or several seconds to localize; (2) past RFID localization techniques lack robustness to changes in tag orientation; and (3) current RFID orientation extraction methodologies are largely inaccurate. To overcome these challenges, this thesis presents RF-Reality, a new system that leverages a novel OFDM backscatter technique and differential channel estimation algorithm to perform accurate, rapid RFID position and orientation recovery.