An indoor positioning system, with ultra-high frequency radio frequency identification based on phase difference (PD), is presented in this paper. Several groups of phases are obtained by setting reader’s different transmitting frequency to reduce the phase ambiguity. Two positioning algorithms are proposed with rough PDs, which are pattern matching based on PD correction and machine learning based on without PD correction. The experiments of two methods, in which machine learning gets a better positioning precision of 0.43 m, are also presented and discussed.

中文翻译：

---

基于相位差的UHF RFID室内定位

本文提出了一种基于相位差（PD）的超高频射频识别室内定位系统。通过设置阅读器不同的发射频率来获得多组相位，以减少相位模糊度。针对粗糙PD，提出了两种定位算法，即基于PD校正的模式匹配和基于无PD校正的机器学习。还介绍并讨论了两种方法的实验，其中机器学习获得了0.43 m的更好定位精度。