Channel State Information Based Indoor Fingerprinting Localization

Indoor localization is one of the key techniques for location-based services (LBSs), which play a significant role in applications in confined spaces, such as tunnels and mines. To achieve indoor localization in confined spaces, the channel state information (CSI) of WiFi can be selected as a featur...

Full description

Saved in:
Bibliographic Details
Published inSensors (Basel, Switzerland) Vol. 23; no. 13; p. 5830
Main Authors Che, Rongjie, Chen, Honglong
Format Journal Article
LanguageEnglish
Published Switzerland MDPI AG 22.06.2023
MDPI
Subjects
Accuracy
Amplitudes
channel state information
Comparative analysis
Confined spaces
Deep learning
Fingerprinting
fingerprinting-based
Fingerprints
full-dimensional subcarriers
Global positioning systems
GPS
indoor localization
Interface cards
Internet of Things
IWKNN
Localization
Location based services
MIMO communications
Network interface cards
Signal strength
Subcarriers
Wi-Fi
Wireless communications
IWKNN
channel state information
indoor localization
full-dimensional subcarriers
fingerprinting-based
Online AccessGet full text

Cover

More Information
Summary:Indoor localization is one of the key techniques for location-based services (LBSs), which play a significant role in applications in confined spaces, such as tunnels and mines. To achieve indoor localization in confined spaces, the channel state information (CSI) of WiFi can be selected as a feature to distinguish locations due to its fine-grained characteristics compared with the received signal strength (RSS). In this paper, two indoor localization approaches based on CSI fingerprinting were designed: amplitude-of-CSI-based indoor fingerprinting localization (AmpFi) and full-dimensional CSI-based indoor fingerprinting localization (FuFi). AmpFi adopts the amplitude of the CSI as the localization fingerprint in the offline phase, and in the online phase, the improved weighted K-nearest neighbor (IWKNN) is proposed to estimate the unknown locations. Based on AmpFi, FuFi is proposed, which considers all of the subcarriers in the MIMO system as the independent features and adopts the normalized amplitudes of the full-dimensional subcarriers as the fingerprint. AmpFi and FuFi were implemented on a commercial network interface card (NIC), where FuFi outperformed several other typical fingerprinting-based indoor localization approaches.
Bibliography:ObjectType-Article-1
SourceType-Scholarly Journals-1
ObjectType-Feature-2
content type line 23
ISSN:1424-8220
1424-8220
DOI:10.3390/s23135830