KLI

REM-Based Indoor Localization with an Extra-Trees Regressor

Metadata Downloads
Alternative Title
REM-Based Indoor Localization with an Extra-Trees Regressor
Abstract
As a widely established and accessible infrastructure, wireless local area networks (WLANs) have emerged as a viable option for indoor localization for both mobile and stationary users. However, WLANs present several challenges that must be fulfilled to achieve localization based on Wi-Fi signals and to obtain proper coverage prediction maps. This paper presents a study based on the application of extra-trees regression (ETR) for indoor localization using coverage prediction maps. The aim of the proposed method is to accurately estimate a user’s position within a radio environment map (REM) area using collected signal strength indicator (RSSI) values collected by a mobile robot. Our methodology consists of utilizing the RSSI collected values to construct the REM, which is then leveraged to create a dataset for indoor localization. This process involves tracking a user’s movements within a specific area of interest while considering a single access point. The proposed scheme explores various machine learning (ML) regression algorithms, with hyperparameter tuning carried out to optimize their performance through 10-fold cross-validation. To assess the REM, we employed metrics, such as the root mean square error, absolute error, and R-squared error. Additionally, we evaluated the indoor localization accuracy using location error metrics. Among the ML techniques assessed, our proposed ETR-based approach demonstrates the highest performance based on these error metrics. The combination of generating coverage maps and utilizing regression techniques for localization presents a potent approach for analyzing the radio frequency environment in indoor spaces.
Author(s)
Toufiq AzizMario R. CamanaCarla E. GarciaTaewoong HwangInsoo Koo
Issued Date
2023
Type
Article
Keyword
machine learningindoor localizationradio environment mapextra-trees regressorcross-validation
DOI
10.3390/electronics12204350
URI
https://oak.ulsan.ac.kr/handle/2021.oak/16997
Publisher
ELECTRONICS
Language
영어
ISSN
2079-9292
Citation Volume
12
Citation Number
20
Citation Start Page
1
Citation End Page
23
Appears in Collections:
Engineering > IT Convergence
공개 및 라이선스
  • 공개 구분공개
파일 목록
  • 관련 파일이 존재하지 않습니다.

Items in Repository are protected by copyright, with all rights reserved, unless otherwise indicated.