RFID Localization Based on Multiple Feature Fusion

As one of the enabling technologies for Internet of Things (IoTs), radio frequency identification (RFID) has been widely adopted in many applications. Among others, RFID-based localization has attracted much research attention in recent years. Existing RFID localization algorithms are usually based on single signal feature, e.g., received signal strength (RSS) or phase information. These algorithms cannot achieve high localization accuracy and low delay simultaneously: Algorithms based on RSS usually suffer from low localization accuracy, while algorithms based on phase suffer from large localization delay because they need to collect a large number of phase readings to resolve phase ambiguity. In this paper, we propose an RFID localization algorithm that can achieve both high accuracy and low delay by fusing multiple types of signal features. We first use the RSS measurements to quickly shrink the possible region of the target tag, and then use phase measurements to refine the position estimation. Experiment results demonstrate the effectiveness of this novel design.