Stage-Wise and Hierarchical Training of Linked Deep Neural Networks for Large-Scale Multi-Building and Multi-Floor Indoor Localization Based on Wi-Fi Fingerprinting

Sihao Li; Kyeong Soo Kim; Zhe Tang; Jeremy S. Smith

doi:10.1109/CANDARW60564.2023.00019

Stage-Wise and Hierarchical Training of Linked Deep Neural Networks for Large-Scale Multi-Building and Multi-Floor Indoor Localization Based on Wi-Fi Fingerprinting

Sihao Li
, Kyeong Soo Kim^*
, Zhe Tang
, Jeremy S. Smith

^*Corresponding author for this work

Department of Communications and Networking

Xi'an Jiaotong-Liverpool University
University of Liverpool

Research output: Chapter in Book or Report/Conference proceeding › Conference Proceeding › peer-review

1 Citation (Scopus)

Abstract

This paper present a new solution to the problem of large-scale multi-building and multi-floor indoor localization based on linked deep neural networks (DNNs) - each of which is dedicated to a sub-estimation problem (i.e., building/floor and floor-level location) - trained under the stage-wise and hierarchical training framework. The proposed hierarchical stage-wise training framework extends the original stage-wise training framework to the case of multiple networks by training the DNN for the estimation of floor-level location based on the prior knowledge gained from the training of the DNN for the estimation of building and floor identifiers. The experimental results, with the publicly-available UJIIndoorLoc multi-building and multi-floor Wi-Fi fingerprint database, demonstrate that the linked DNNs trained under the newly-proposed stage-wise and hierarchical training framework can achieve a three-dimensional localization error of 8.19 m, which, to the best of the authors' knowledge, is the most accurate results obtained for the whole of the UJIIndoorLoc database based on DNN-based models.

Original language	English
Title of host publication	Proceedings - 2023 11th International Symposium on Computing and Networking Workshops, CANDARW 2023
Publisher	Institute of Electrical and Electronics Engineers Inc.
Pages	63-68
Number of pages	6
ISBN (Electronic)	9798350306941
DOIs	https://doi.org/10.1109/CANDARW60564.2023.00019
Publication status	Published - 2023
Event	11th International Symposium on Computing and Networking Workshops, CANDARW 2023 - Matsue, Japan Duration: 28 Nov 2023 → 1 Dec 2023

Publication series

Name	Proceedings - 2023 11th International Symposium on Computing and Networking Workshops, CANDARW 2023

Conference

Conference	11th International Symposium on Computing and Networking Workshops, CANDARW 2023
Country/Territory	Japan
City	Matsue
Period	28/11/23 → 1/12/23

Keywords

deep neural networks
Indoor localization
stage-wise training
Wi-Fi fingerprinting

Access to Document

10.1109/CANDARW60564.2023.00019

Cite this

Li, S., Kim, K. S., Tang, Z., & Smith, J. S. (2023). Stage-Wise and Hierarchical Training of Linked Deep Neural Networks for Large-Scale Multi-Building and Multi-Floor Indoor Localization Based on Wi-Fi Fingerprinting. In Proceedings - 2023 11th International Symposium on Computing and Networking Workshops, CANDARW 2023 (pp. 63-68). (Proceedings - 2023 11th International Symposium on Computing and Networking Workshops, CANDARW 2023). Institute of Electrical and Electronics Engineers Inc.. https://doi.org/10.1109/CANDARW60564.2023.00019

Li, Sihao ; Kim, Kyeong Soo ; Tang, Zhe et al. / Stage-Wise and Hierarchical Training of Linked Deep Neural Networks for Large-Scale Multi-Building and Multi-Floor Indoor Localization Based on Wi-Fi Fingerprinting. Proceedings - 2023 11th International Symposium on Computing and Networking Workshops, CANDARW 2023. Institute of Electrical and Electronics Engineers Inc., 2023. pp. 63-68 (Proceedings - 2023 11th International Symposium on Computing and Networking Workshops, CANDARW 2023).

@inproceedings{347b6369ecee42f79443a5ef1e26676f,

title = "Stage-Wise and Hierarchical Training of Linked Deep Neural Networks for Large-Scale Multi-Building and Multi-Floor Indoor Localization Based on Wi-Fi Fingerprinting",

abstract = "This paper present a new solution to the problem of large-scale multi-building and multi-floor indoor localization based on linked deep neural networks (DNNs) - each of which is dedicated to a sub-estimation problem (i.e., building/floor and floor-level location) - trained under the stage-wise and hierarchical training framework. The proposed hierarchical stage-wise training framework extends the original stage-wise training framework to the case of multiple networks by training the DNN for the estimation of floor-level location based on the prior knowledge gained from the training of the DNN for the estimation of building and floor identifiers. The experimental results, with the publicly-available UJIIndoorLoc multi-building and multi-floor Wi-Fi fingerprint database, demonstrate that the linked DNNs trained under the newly-proposed stage-wise and hierarchical training framework can achieve a three-dimensional localization error of 8.19 m, which, to the best of the authors' knowledge, is the most accurate results obtained for the whole of the UJIIndoorLoc database based on DNN-based models.",

keywords = "deep neural networks, Indoor localization, stage-wise training, Wi-Fi fingerprinting",

author = "Sihao Li and Kim, \{Kyeong Soo\} and Zhe Tang and Smith, \{Jeremy S.\}",

note = "Publisher Copyright: {\textcopyright} 2023 IEEE.; 11th International Symposium on Computing and Networking Workshops, CANDARW 2023 ; Conference date: 28-11-2023 Through 01-12-2023",

year = "2023",

doi = "10.1109/CANDARW60564.2023.00019",

language = "English",

series = "Proceedings - 2023 11th International Symposium on Computing and Networking Workshops, CANDARW 2023",

publisher = "Institute of Electrical and Electronics Engineers Inc.",

pages = "63--68",

booktitle = "Proceedings - 2023 11th International Symposium on Computing and Networking Workshops, CANDARW 2023",

}

Li, S, Kim, KS, Tang, Z & Smith, JS 2023, Stage-Wise and Hierarchical Training of Linked Deep Neural Networks for Large-Scale Multi-Building and Multi-Floor Indoor Localization Based on Wi-Fi Fingerprinting. in Proceedings - 2023 11th International Symposium on Computing and Networking Workshops, CANDARW 2023. Proceedings - 2023 11th International Symposium on Computing and Networking Workshops, CANDARW 2023, Institute of Electrical and Electronics Engineers Inc., pp. 63-68, 11th International Symposium on Computing and Networking Workshops, CANDARW 2023, Matsue, Japan, 28/11/23. https://doi.org/10.1109/CANDARW60564.2023.00019

Stage-Wise and Hierarchical Training of Linked Deep Neural Networks for Large-Scale Multi-Building and Multi-Floor Indoor Localization Based on Wi-Fi Fingerprinting. / Li, Sihao; Kim, Kyeong Soo; Tang, Zhe et al.
Proceedings - 2023 11th International Symposium on Computing and Networking Workshops, CANDARW 2023. Institute of Electrical and Electronics Engineers Inc., 2023. p. 63-68 (Proceedings - 2023 11th International Symposium on Computing and Networking Workshops, CANDARW 2023).

Research output: Chapter in Book or Report/Conference proceeding › Conference Proceeding › peer-review

TY - GEN

T1 - Stage-Wise and Hierarchical Training of Linked Deep Neural Networks for Large-Scale Multi-Building and Multi-Floor Indoor Localization Based on Wi-Fi Fingerprinting

AU - Li, Sihao

AU - Kim, Kyeong Soo

AU - Tang, Zhe

AU - Smith, Jeremy S.

PY - 2023

Y1 - 2023

N2 - This paper present a new solution to the problem of large-scale multi-building and multi-floor indoor localization based on linked deep neural networks (DNNs) - each of which is dedicated to a sub-estimation problem (i.e., building/floor and floor-level location) - trained under the stage-wise and hierarchical training framework. The proposed hierarchical stage-wise training framework extends the original stage-wise training framework to the case of multiple networks by training the DNN for the estimation of floor-level location based on the prior knowledge gained from the training of the DNN for the estimation of building and floor identifiers. The experimental results, with the publicly-available UJIIndoorLoc multi-building and multi-floor Wi-Fi fingerprint database, demonstrate that the linked DNNs trained under the newly-proposed stage-wise and hierarchical training framework can achieve a three-dimensional localization error of 8.19 m, which, to the best of the authors' knowledge, is the most accurate results obtained for the whole of the UJIIndoorLoc database based on DNN-based models.

AB - This paper present a new solution to the problem of large-scale multi-building and multi-floor indoor localization based on linked deep neural networks (DNNs) - each of which is dedicated to a sub-estimation problem (i.e., building/floor and floor-level location) - trained under the stage-wise and hierarchical training framework. The proposed hierarchical stage-wise training framework extends the original stage-wise training framework to the case of multiple networks by training the DNN for the estimation of floor-level location based on the prior knowledge gained from the training of the DNN for the estimation of building and floor identifiers. The experimental results, with the publicly-available UJIIndoorLoc multi-building and multi-floor Wi-Fi fingerprint database, demonstrate that the linked DNNs trained under the newly-proposed stage-wise and hierarchical training framework can achieve a three-dimensional localization error of 8.19 m, which, to the best of the authors' knowledge, is the most accurate results obtained for the whole of the UJIIndoorLoc database based on DNN-based models.

KW - deep neural networks

KW - Indoor localization

KW - stage-wise training

KW - Wi-Fi fingerprinting

UR - https://www.scopus.com/pages/publications/85185718345

U2 - 10.1109/CANDARW60564.2023.00019

DO - 10.1109/CANDARW60564.2023.00019

M3 - Conference Proceeding

AN - SCOPUS:85185718345

T3 - Proceedings - 2023 11th International Symposium on Computing and Networking Workshops, CANDARW 2023

SP - 63

EP - 68

BT - Proceedings - 2023 11th International Symposium on Computing and Networking Workshops, CANDARW 2023

PB - Institute of Electrical and Electronics Engineers Inc.

T2 - 11th International Symposium on Computing and Networking Workshops, CANDARW 2023

Y2 - 28 November 2023 through 1 December 2023

ER -

Li S, Kim KS, Tang Z, Smith JS. Stage-Wise and Hierarchical Training of Linked Deep Neural Networks for Large-Scale Multi-Building and Multi-Floor Indoor Localization Based on Wi-Fi Fingerprinting. In Proceedings - 2023 11th International Symposium on Computing and Networking Workshops, CANDARW 2023. Institute of Electrical and Electronics Engineers Inc. 2023. p. 63-68. (Proceedings - 2023 11th International Symposium on Computing and Networking Workshops, CANDARW 2023). doi: 10.1109/CANDARW60564.2023.00019

Stage-Wise and Hierarchical Training of Linked Deep Neural Networks for Large-Scale Multi-Building and Multi-Floor Indoor Localization Based on Wi-Fi Fingerprinting

Abstract

Publication series

Conference

Keywords

Access to Document

Other files and links

Fingerprint

Hierarchical Stage-Wise Training of Linked Deep Neural Networks for Multi-Building and Multi-Floor Indoor Localization Based on Wi-Fi RSSI Fingerprinting

On the Use and Construction of Wi-Fi Fingerprint Databases for Large-Scale Multi-Building and Multi-Floor Indoor Localization: A Case Study of the UJIIndoorLoc Database

IEEE Sensors Journal (Journal)

Exploiting unlabeled RSSI fingerprints in multi-building and multi-floor indoor localization through deep semi-supervised learning based on mean teacher

Cite this

Stage-Wise and Hierarchical Training of Linked Deep Neural Networks for Large-Scale Multi-Building and Multi-Floor Indoor Localization Based on Wi-Fi Fingerprinting

Abstract

Publication series

Conference

Keywords

Access to Document

Other files and links

Fingerprint

Research output

Hierarchical Stage-Wise Training of Linked Deep Neural Networks for Multi-Building and Multi-Floor Indoor Localization Based on Wi-Fi RSSI Fingerprinting

On the Use and Construction of Wi-Fi Fingerprint Databases for Large-Scale Multi-Building and Multi-Floor Indoor Localization: A Case Study of the UJIIndoorLoc Database

Activities

IEEE Sensors Journal (Journal)

Exploiting unlabeled RSSI fingerprints in multi-building and multi-floor indoor localization through deep semi-supervised learning based on mean teacher

Cite this