Ensembling Multiple Radio Maps with Dynamic Noise in Fingerprint-based Indoor Positioning

Joaquin Torres-Sospedra; Fernando J. Aranda; Fernando J. Alvarez; Darwin Quezada-Gaibor; Ivo Silva; Cristiano Pendao; Adriano Moreira

doi:10.1109/VTC2021-Spring51267.2021.9448947

Ensembling Multiple Radio Maps with Dynamic Noise in Fingerprint-based Indoor Positioning

Joaquin Torres-Sospedra^*
, Fernando J. Aranda
, Fernando J. Alvarez
, Darwin Quezada-Gaibor
, Ivo Silva
, Cristiano Pendao
, Adriano Moreira

^*Corresponding author for this work

Electrical Engineering

Research output: Chapter in Book/Report/Conference proceeding › Conference contribution › Scientific › peer-review

1 Citation (Scopus)

16 Downloads (Pure)

Abstract

Fingerprint-based indoor positioning is widely used in many contexts, including pedestrian and autonomous vehicles navigation. Many approaches have used traditional Machine Learning models to deal with fingerprinting, being k-NN the most common used one. However, the reference data (or radio map) is generally limited, as data collection is a very demanding task, which degrades overall accuracy. In this work, we propose a novel approach to add random noise to the radio map which will be used in combination with an ensemble model. Instead of augmenting the radio map, we create n noisy versions of the same size, i.e. our proposed Indoor Positioning model will combine n estimations obtained by independent estimators built with the n noisy radio maps. The empirical results have shown that our proposed approach improves the baseline method results in around 10% on average.

Original language	English
Title of host publication	2021 IEEE 93rd Vehicular Technology Conference, VTC 2021-Spring - Proceedings
Publisher	IEEE
Number of pages	5
ISBN (Electronic)	9781728189642
DOIs	https://doi.org/10.1109/VTC2021-Spring51267.2021.9448947
Publication status	Published - 2021
Publication type	A4 Article in conference proceedings
Event	IEEE Vehicular Technology Conference - Helsinki, Finland Duration: 25 Apr 2021 → 28 Apr 2021

Publication series

Name	IEEE Vehicular Technology Conference
Volume	2021-April
ISSN (Electronic)	2577-2465

Conference

Conference	IEEE Vehicular Technology Conference
Country/Territory	Finland
City	Helsinki
Period	25/04/21 → 28/04/21

Funding

Corresponding Author: J. Torres-Sospedra ([email protected]) The authors gratefully acknowledge funding from Ministerio de Cien-cia, Innovación y Universidades (INSIGNIA, PTQ2018-009981 and MI-CROCEBUS, RTI2018-095168-B-C54); Ministerio de Economía, Industria y Competitividad (REPNIN+, TEC2017-90808-REDT); European Union’s H2020 Research and Innovation programme under the Marie Skłodowska-Curie grant agreement No.813278 (A-WEAR, http://www.a-wear.eu/); FCT – Fundac¸ão para a Ciência e Tecnologia within the R&D Units Project Scope: UIDB/00319/2020 and the PhD fellowship PD/BD/137401/2018

Keywords

Ensemble
Fingerprinting
Indoor Positioning
Noisy samples
Radio Map

Publication forum classification

Publication forum level 1

ASJC Scopus subject areas

Computer Science Applications
Electrical and Electronic Engineering
Applied Mathematics

Access to Document

10.1109/VTC2021-Spring51267.2021.9448947

Ensembling Multiple Radio Maps
This publication is copyrighted. You may download, display and print it for Your own personal use. Commercial use is prohibited.
Accepted author manuscript, 2.05 MBLicence: Unspecified

https://urn.fi/URN:NBN:fi:tuni-202112179349Licence: Unspecified

Cite this

Torres-Sospedra, J., Aranda, F. J., Alvarez, F. J., Quezada-Gaibor, D., Silva, I., Pendao, C., & Moreira, A. (2021). Ensembling Multiple Radio Maps with Dynamic Noise in Fingerprint-based Indoor Positioning. In 2021 IEEE 93rd Vehicular Technology Conference, VTC 2021-Spring - Proceedings (IEEE Vehicular Technology Conference; Vol. 2021-April). IEEE. https://doi.org/10.1109/VTC2021-Spring51267.2021.9448947

@inproceedings{405411119ba9457796738b6c2dddc788,

title = "Ensembling Multiple Radio Maps with Dynamic Noise in Fingerprint-based Indoor Positioning",

abstract = "Fingerprint-based indoor positioning is widely used in many contexts, including pedestrian and autonomous vehicles navigation. Many approaches have used traditional Machine Learning models to deal with fingerprinting, being k-NN the most common used one. However, the reference data (or radio map) is generally limited, as data collection is a very demanding task, which degrades overall accuracy. In this work, we propose a novel approach to add random noise to the radio map which will be used in combination with an ensemble model. Instead of augmenting the radio map, we create n noisy versions of the same size, i.e. our proposed Indoor Positioning model will combine n estimations obtained by independent estimators built with the n noisy radio maps. The empirical results have shown that our proposed approach improves the baseline method results in around 10\% on average.",

keywords = "Ensemble, Fingerprinting, Indoor Positioning, Noisy samples, Radio Map",

author = "Joaquin Torres-Sospedra and Aranda, \{Fernando J.\} and Alvarez, \{Fernando J.\} and Darwin Quezada-Gaibor and Ivo Silva and Cristiano Pendao and Adriano Moreira",

note = "Funding Information: Corresponding Author: J. Torres-Sospedra ([email protected]) The authors gratefully acknowledge funding from Ministerio de Cien-cia, Innovaci{\'o}n y Universidades (INSIGNIA, PTQ2018-009981 and MI-CROCEBUS, RTI2018-095168-B-C54); Ministerio de Econom{\'i}a, Industria y Competitividad (REPNIN+, TEC2017-90808-REDT); European Union{\textquoteright}s H2020 Research and Innovation programme under the Marie Sk{\l}odowska-Curie grant agreement No.813278 (A-WEAR, http://www.a-wear.eu/); FCT {\textendash} Fundac¸{\~a}o para a Ci{\^e}ncia e Tecnologia within the R\&D Units Project Scope: UIDB/00319/2020 and the PhD fellowship PD/BD/137401/2018 Publisher Copyright: {\textcopyright} 2021 IEEE. jufoid=57589; IEEE Vehicular Technology Conference ; Conference date: 25-04-2021 Through 28-04-2021",

year = "2021",

doi = "10.1109/VTC2021-Spring51267.2021.9448947",

language = "English",

series = "IEEE Vehicular Technology Conference",

publisher = "IEEE",

booktitle = "2021 IEEE 93rd Vehicular Technology Conference, VTC 2021-Spring - Proceedings",

address = "United States",

}

Torres-Sospedra, J, Aranda, FJ, Alvarez, FJ, Quezada-Gaibor, D, Silva, I, Pendao, C & Moreira, A 2021, Ensembling Multiple Radio Maps with Dynamic Noise in Fingerprint-based Indoor Positioning. in 2021 IEEE 93rd Vehicular Technology Conference, VTC 2021-Spring - Proceedings. IEEE Vehicular Technology Conference, vol. 2021-April, IEEE, IEEE Vehicular Technology Conference, Helsinki, Finland, 25/04/21. https://doi.org/10.1109/VTC2021-Spring51267.2021.9448947

Ensembling Multiple Radio Maps with Dynamic Noise in Fingerprint-based Indoor Positioning. / Torres-Sospedra, Joaquin; Aranda, Fernando J.; Alvarez, Fernando J. et al.
2021 IEEE 93rd Vehicular Technology Conference, VTC 2021-Spring - Proceedings. IEEE, 2021. (IEEE Vehicular Technology Conference; Vol. 2021-April).

Research output: Chapter in Book/Report/Conference proceeding › Conference contribution › Scientific › peer-review

TY - GEN

T1 - Ensembling Multiple Radio Maps with Dynamic Noise in Fingerprint-based Indoor Positioning

AU - Torres-Sospedra, Joaquin

AU - Aranda, Fernando J.

AU - Alvarez, Fernando J.

AU - Quezada-Gaibor, Darwin

AU - Silva, Ivo

AU - Pendao, Cristiano

AU - Moreira, Adriano

N1 - Funding Information: Corresponding Author: J. Torres-Sospedra ([email protected]) The authors gratefully acknowledge funding from Ministerio de Cien-cia, Innovación y Universidades (INSIGNIA, PTQ2018-009981 and MI-CROCEBUS, RTI2018-095168-B-C54); Ministerio de Economía, Industria y Competitividad (REPNIN+, TEC2017-90808-REDT); European Union’s H2020 Research and Innovation programme under the Marie Skłodowska-Curie grant agreement No.813278 (A-WEAR, http://www.a-wear.eu/); FCT – Fundac¸ão para a Ciência e Tecnologia within the R&D Units Project Scope: UIDB/00319/2020 and the PhD fellowship PD/BD/137401/2018 Publisher Copyright: © 2021 IEEE. jufoid=57589

PY - 2021

Y1 - 2021

N2 - Fingerprint-based indoor positioning is widely used in many contexts, including pedestrian and autonomous vehicles navigation. Many approaches have used traditional Machine Learning models to deal with fingerprinting, being k-NN the most common used one. However, the reference data (or radio map) is generally limited, as data collection is a very demanding task, which degrades overall accuracy. In this work, we propose a novel approach to add random noise to the radio map which will be used in combination with an ensemble model. Instead of augmenting the radio map, we create n noisy versions of the same size, i.e. our proposed Indoor Positioning model will combine n estimations obtained by independent estimators built with the n noisy radio maps. The empirical results have shown that our proposed approach improves the baseline method results in around 10% on average.

AB - Fingerprint-based indoor positioning is widely used in many contexts, including pedestrian and autonomous vehicles navigation. Many approaches have used traditional Machine Learning models to deal with fingerprinting, being k-NN the most common used one. However, the reference data (or radio map) is generally limited, as data collection is a very demanding task, which degrades overall accuracy. In this work, we propose a novel approach to add random noise to the radio map which will be used in combination with an ensemble model. Instead of augmenting the radio map, we create n noisy versions of the same size, i.e. our proposed Indoor Positioning model will combine n estimations obtained by independent estimators built with the n noisy radio maps. The empirical results have shown that our proposed approach improves the baseline method results in around 10% on average.

KW - Ensemble

KW - Fingerprinting

KW - Indoor Positioning

KW - Noisy samples

KW - Radio Map

U2 - 10.1109/VTC2021-Spring51267.2021.9448947

DO - 10.1109/VTC2021-Spring51267.2021.9448947

M3 - Conference contribution

AN - SCOPUS:85112422818

T3 - IEEE Vehicular Technology Conference

BT - 2021 IEEE 93rd Vehicular Technology Conference, VTC 2021-Spring - Proceedings

PB - IEEE

T2 - IEEE Vehicular Technology Conference

Y2 - 25 April 2021 through 28 April 2021

ER -