Cited 3 time in webofscience Cited 3 time in scopus

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

Title
Toward Scalable and Robust Indoor Tracking: Design, Implementation, and Evaluation
Authors
Jin, FeiyuLiu, KaiZhang, HaoNg, Joseph Kee-YinGuo, SongtaoLee, Victor C. S.Son, Sang Hyuk
DGIST Authors
Jin, Feiyu; Liu, Kai; Zhang, Hao; Ng, Joseph Kee-Yin; Guo, Songtao; Lee, Victor C. S.; Son, Sang Hyuk
Issue Date
2020-02
Citation
IEEE Internet of Things Journal, 7(2), 1192-1204
Type
Article
Article Type
Article
Author Keywords
Algorithm designindoor localizationperformance evaluationtrajectory trackingWi-Fi fingerprint
Keywords
POSITIONING SYSTEMSLOCALIZATION
ISSN
2327-4662
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.
URI
http://hdl.handle.net/20.500.11750/11674
DOI
10.1109/JIOT.2019.2953376
Publisher
Institute of Electrical and Electronics Engineers Inc.
Files:
There are no files associated with this item.
Collection:


qrcode mendeley

Items in DSpace are protected by copyright, with all rights reserved, unless otherwise indicated.

BROWSE