IoT erişimli akıllı şehirlerde radyo frekansı parmak izi tabanlı yayıcı konumlandırma
No Thumbnail Available
Date
2023
Authors
Journal Title
Journal ISSN
Volume Title
Publisher
Open Access Color
OpenAIRE Downloads
OpenAIRE Views
Abstract
Kablosuz teknolojinin hızlı gelişimi, Nesnelerin İnterneti'nin (IoT) önemini artırdı. IoT uygulamaları, çeşitli sektörlerde maliyetleri azaltmak ve performansı yükseltmek için kullanılıyor. Akıllı şehirlerde bu tür uygulamalardan yararlanılarak konumlandırma tabanlı hizmetler de sunulmaktadır. Coğrafi bölgelerde konumlandırma talebi nedeniyle uzun yıllardır çeşitli konumlandırma prosedürleri kullanılmaktadır. Radyo frekansı parmak izi (RFF) konumlandırması, makine öğrenimi (ML) yöntemlerindeki son gelişmelerin sağladığı avantajlar dikkate alındığında en etkili yöntemlerden biri haline geldi. Makul fiyatlı ve yüksek performanslı bir IoT kablosuz teknolojisini uygulamak, konumlandırmada zorlu bir konudur. Bu bağlamda, IQRF teknolojisi yeni fırsatlar sunmaktadır. Bu nedenle, 868 MHz bandında çalışan IQRF sensör düğümlerini içeren bir sistemde bu tez, makine öğreniminde denetimli sınıflandırma yöntemlerini uygulayan bir alınan sinyal gücü göstergesi (RSSI) parmak izi tabanlı konumlandırma yöntemi önerir. Bu amaçla, Görüş Hattı (LoS) bağlantıları için yerel bir dış ortamda ölçümler yürütüldü. Elde edilen sonuçlar, 'Torbalı Ağaçlar', 'Ağırlıklı k-NN' ve 'Orta Gaussian SVM' yöntemlerinin son derece güçlü tahmin doğruluğunu gösterir. Tezin sonuçları, akıllı şehirlerde radyo frekansı parmak izine dayalı konumlandırma sistemlerinin ilerlemesine destek olma potansiyeline sahiptir.
The rapid advancement of wireless technology has grown the significance of the Internet of Things (IoT). IoT applications are being used to decrease costs and improve performance across various industries. In smart cities, such applications are also utilized to offer localization-based services. Several localization procedures have been used for long years due to the demand for localization in geographic regions. Radio frequency fingerprinting (RFF) localization has become one of the most effective methods when considering the advantages provided by recent advancements in machine learning (ML) methods. Implementing a reasonable-priced and high-performance IoT wireless technology is a challenging issue in localization. In this regard, IQRF technology presents novel opportunities. Thus, in a system comprising IQRF sensor nodes operating in the 868 MHz band, this thesis proposes a received signal strength indicator (RSSI) fingerprint-based localization method implementing supervised classification methods in ML. To this end, measurements for Line-of-Sight (LoS) links were conducted in a local outdoor environment. The achieved results show the exceptionally strong prediction accuracy of the 'Bagged Trees', 'Weighted k-NN', and 'Medium Gaussian SVM' methods. The results of the thesis have the potential to assist in the advancement of localization systems based on RFF in smart cities.
The rapid advancement of wireless technology has grown the significance of the Internet of Things (IoT). IoT applications are being used to decrease costs and improve performance across various industries. In smart cities, such applications are also utilized to offer localization-based services. Several localization procedures have been used for long years due to the demand for localization in geographic regions. Radio frequency fingerprinting (RFF) localization has become one of the most effective methods when considering the advantages provided by recent advancements in machine learning (ML) methods. Implementing a reasonable-priced and high-performance IoT wireless technology is a challenging issue in localization. In this regard, IQRF technology presents novel opportunities. Thus, in a system comprising IQRF sensor nodes operating in the 868 MHz band, this thesis proposes a received signal strength indicator (RSSI) fingerprint-based localization method implementing supervised classification methods in ML. To this end, measurements for Line-of-Sight (LoS) links were conducted in a local outdoor environment. The achieved results show the exceptionally strong prediction accuracy of the 'Bagged Trees', 'Weighted k-NN', and 'Medium Gaussian SVM' methods. The results of the thesis have the potential to assist in the advancement of localization systems based on RFF in smart cities.
Description
Keywords
Elektrik ve Elektronik Mühendisliği, Electrical and Electronics Engineering
Turkish CoHE Thesis Center URL
Fields of Science
Citation
WoS Q
Scopus Q
Source
Volume
Issue
Start Page
0
End Page
72