Experimental Studies of High-Accuracy RFID Localization with Channel Impairments

Radio frequency identification (RFID) systems present an incredibly cost-effective and easy-to-implement solution to close-range localization. One of the important applications of a passive RFID system is to determine the reader position through multilateration based on the estimated distances between the reader and multiple distributed reference tags obtained from, e.g., the received signal strength indicator (RSSI) readings. In practice, the achievable accuracy of passive RFID reader localization suffers from many factors, such as the distorted RSSI reading due to channel impairments in terms of the susceptibility to reader antenna patterns and multipath propagation. Previous studies have shown that the accuracy of passive RFID localization can be significantly improved by properly modeling and compensating for such channel impairments. The objective of this paper is to report experimental study results that validate the effectiveness of such approaches for high-accuracy RFID localization. We also examine a number of practical issues arising in the underlying problem that limit the accuracy of reader-tag distance measurements and, therefore, the estimated reader localization. These issues include the variations in tag radiation characteristics for similar tags, effects of tag orientations, and reader RSS quantization and measurement errors. As such, this paper reveals valuable insights of the issues and solutions toward achieving high-accuracy passive RFID localization.