Toward Scalable and Robust Indoor Tracking : Design, Implementation, and Evaluation

Research output: Journal Publications and Reviews (RGC: 21, 22, 62)21_Publication in refereed journal

1 Scopus Citations
View graph of relations

Author(s)

  • Feiyu Jin
  • Kai Liu
  • Hao Zhang
  • Joseph Kee-Yin Ng
  • Songtao Guo
  • Sang H. Son

Related Research Unit(s)

Detail(s)

Original languageEnglish
Article number8897653
Pages (from-to)1192-1204
Journal / PublicationIEEE Internet of Things Journal
Volume7
Issue number2
Online published13 Nov 2019
Publication statusPublished - Feb 2020

Abstract

Although indoor localization has been studied over a decade, it is still challenging to enable many IoT applications, such as activity tracking and monitoring in smart home and customer navigation and trajectory mining in smart shopping mall, which typically require meter-level localization accuracy in a highly dynamic and large-scale indoor environment. Therefore, this article aims at designing and implementing an adaptive and scalable indoor tracking system in a cost-effective way. First, we propose a zero site-survey overhead (ZSSO) algorithm to enhance the system scalability. It integrates the step information and map constraints to infer user's positions based on the particle filter and supports the auto labeling of scanned Wi-Fi signal for constructing the fingerprint database without the extra site-survey overhead. Further, we propose an iterative-weight-update (IWU) strategy for ZSSO to enhance system robustness and make it more adaptive to the dynamic changing of environments. Specifically, a two-step clustering mechanism is proposed to delete outliers in the fingerprint database and alleviate the mismatch between the auto-tagged coordinates and the corresponding signal features. Then, an iterative fingerprint update mechanism is designed to continuously evaluate the Wi-Fi fingerprint localization results during online tracking, which will further refine the fingerprint database. Finally, we implement the indoor tracking system in real-world environments and conduct a comprehensive performance evaluation. The field testing results conclusively demonstrate the scalability and effectiveness of the proposed algorithms.

Research Area(s)

  • Algorithm design, indoor localization, performance evaluation, trajectory tracking, Wi-Fi fingerprint

Citation Format(s)

Toward Scalable and Robust Indoor Tracking : Design, Implementation, and Evaluation. / Jin, Feiyu; Liu, Kai; Zhang, Hao; Ng, Joseph Kee-Yin; Guo, Songtao; Lee, Victor C. S.; Son, Sang H.

In: IEEE Internet of Things Journal, Vol. 7, No. 2, 8897653, 02.2020, p. 1192-1204.

Research output: Journal Publications and Reviews (RGC: 21, 22, 62)21_Publication in refereed journal