Developing a Novel Real-Time Indoor Positioning System Based on BLE Beacons and Smartphone Sensors

In this work, we study the problem of fusing one Pedestrian-Dead-Reckoning-based (PDR-based) position measurement and one instant Received-Signal-Strength-based (RSS-based) position measurement. This situation can arise in a smartphone-based indoor positioning system when we want to locate a moving...

Full description

Saved in:
Bibliographic Details
Published in:IEEE sensors journal 2021-10, Vol.21 (20), p.23055-23068
Main Authors: Dinh, Thai-Mai Thi, Duong, Ngoc-Son, Nguyen, Quoc-Tuan
Format: Article
Language:English
Subjects:
Bayes fusion
BLE beacon
Bluetooth Low Energy
Dead reckoning
Estimation
Heterogeneity
indoor localization
indoor positioning system
Kalman filters
least square estimation
Location awareness
Moving targets
Normal distribution
pedestrian dead reckoning
Position measurement
Real time
Real-time systems
Sampling
Sensor systems
Sensors
Smartphones
Uncertainty
Wireless fidelity
Citations: Items that this one cites
Items that cite this one
Online Access:Get full text
Tags: Add Tag
No Tags, Be the first to tag this record!
Description
Summary:In this work, we study the problem of fusing one Pedestrian-Dead-Reckoning-based (PDR-based) position measurement and one instant Received-Signal-Strength-based (RSS-based) position measurement. This situation can arise in a smartphone-based indoor positioning system when we want to locate a moving user in real-time with sustainable accuracy, but the RSS sampling ability of smartphones is limited; for example, one RSS sample per second. Firstly, by investigating RSS's heterogeneity, we offer a solution for RSS-based continuous positioning problems under a low RSS sampling rate that satisfies real-time requirements. Secondly, we propose a method to improve accuracy for the RSS-based position estimation method, i.e., multilateration using Least Square Estimation. We consider PDR-based and improved RSS-based positions both have Gaussian uncertainty due to initial position plus drifting and RSS-to-distance conversion, respectively. Then, the Kalman filter will fuse two kinds of Gaussian distribution to produce more precise positions. The method is intended to design a real-time system for locating a moving target. Experiments are conducted in real indoor space with a commodity device. Its results show that our proposed solution is highly accurate and feasible in actual deployment.
ISSN:1530-437X
1558-1748
DOI:10.1109/JSEN.2021.3106019