Citation

Bakhuraisa, Yaser Awadh Karama (2026) Vehicle localisation for Intelligent Transportation Systems. PhD thesis, Multimedia University.

Full text not available from this repository.

Abstract

Aa accurate and a reliable localisation system is critical for the successful integration of advanced technologies such as Connected Vehicles (CV) and Autonomous Vehicles (AV) into modern Intelligent Transportation Systems (ITS). However, in complex environments, common localisation methods based on the Global Navigation Satellite System (GNSS) are limited in their ability to meet the stringent localisation requirements of such technologies. As a result, geometric localisation methods based on channel propagation characteristics have been suggested as promising alternative methods. Among these methods, received signal strength (RSS)-based localisation is particularly attractive due to its low implementation complexity and suitability for real-time operation without requiring additional infrastructure or precise synchronization. A comprehensive analysis of the state-of-the-art in this area reveals three research gaps that remain essentially unaddressed. First, RSS-based geometric methods have received limited investigation in 5G NR-V2X at mmWave bands. As mmWave emerges as a key enabler of vehicular communications in modern ITS, studying the performance of these methods in realistic 5G New Radio Vehicle to Evrything (5G NR-V2X) scenarios becomes necessary. Second, most RSS-based geometric methods rely on explicit mathematical models, which fail to capture the variability of mmWave channels. This highlights the need for an efficient ranging model capable of capturing the complex and nonlinear characteristics of mmWave. Third, existing research lacks comprehensive localisation systems that consider multiple factors simultaneously. To address these research gaps, the methodology of this thesis is structured as follows. First, a framework is proposed to investigate and analyse the performance of the recent RSS-based geometric methods under realistic 5G NR-V2X scenarios, with mmWave propagation modelled using ray tracing techniques. Next, a machine learning-based ranging model is developed utilising high-resolution channel impulse responses (CIR) that obtained from raytracing techniques to capture the complex relationship between mmWave propagation characteristics and distance. Finally, a robust machine learning-assisted localisation methods are proposed by integration of machine learning-based ranging model with geometric localisation methods. Results demonstrate that the recent RSS-based geometric methods approach the Cramér-Rao lower bound (CRLB) under small measurement errors; however, their performance degrades steadily as the measurement error variance increases. In contrast, the proposed machine learning-assisted localisation methods have achieved significantly improved performance, maintaining near-CRLB accuracy even at an error variance of 8 dB. The RMSE is reduced from approximately 24 m to 3.58 m, yielding localisation errors close to the requirements for ITS technologies.

Item Type:	Thesis (PhD)
Additional Information:	Call No.: TE228.3 .B35 2026
Uncontrolled Keywords:	Intelligent transportation systems
Subjects:	T Technology > TE Highway engineering. Roads and pavements > TE210-228.3 Construction details Including foundations, maintenance, equipment
Divisions:	Faculty of Engineering and Technology (FET)
Depositing User:	Ms Nurul Iqtiani Ahmad
Date Deposited:	05 Nov 2026 11:21
Last Modified:	05 Nov 2026 11:21
URII:	http://shdl.mmu.edu.my/id/eprint/17069

Downloads

Edit (login required)