An extreme value based algorithm for improving the accuracy of WiFi localization

More mobile devices can obtain WiFi Received Signal Strength (RSS) for indoor positioning, so selecting reasonable Access Points (APs) for rapid positioning is important. Furthermore, RSS is vulnerable to diverse interference, making the positioning accuracy decrease. To solve the above mentioned is...

Full description

Saved in:
Bibliographic Details
Published inAd hoc networks Vol. 143; p. 103131
Main Authors Tao, Ye, Yan, Rong’en, Zhao, Long
Format Journal Article
LanguageEnglish
Published Elsevier B.V 15.04.2023
Subjects
AP selection
Extreme values
Indoor positioning
WiFi fingerprint
AP selection
WiFi fingerprint
Indoor positioning
Extreme values
Online AccessGet full text

Cover

More Information
Summary:More mobile devices can obtain WiFi Received Signal Strength (RSS) for indoor positioning, so selecting reasonable Access Points (APs) for rapid positioning is important. Furthermore, RSS is vulnerable to diverse interference, making the positioning accuracy decrease. To solve the above mentioned issues, two novel algorithms, AP selection and positioning algorithms, are proposed, in this paper, which both use the location distance between RPs to construct a circle and treat the RSS from each AP individually. RPs in the same circle are seen as having similar RSS. 1) For each received signal, the proposed selection algorithm leverages the RSS extreme values (maximum and minimum values) of circles to find out RPs which may surround Test Point (TP), then it uses the degree of RSS difference and location concentration to evaluate the importance of each AP. 2) For the positioning algorithm, it uses the extreme values to find the useful APs and determines the weight of RPs. Extensive experiments are conducted on simulation and two real-world sites, results show the effectiveness of the proposed algorithms.
ISSN:1570-8705
1570-8713
DOI:10.1016/j.adhoc.2023.103131