Exploiting the location information for adaptive beamforming in transport systems

Kaur, Jaspreet (2024) Exploiting the location information for adaptive beamforming in transport systems. PhD thesis, University of Glasgow.

Full text available as:
[thumbnail of 2024KaurPhD.pdf] PDF
Download (33MB)


As mobile communication systems evolve, the demand for enhanced network efficiency and pinpoint accuracy in user localization grows, particularly in the context of dynamic environments such as transport systems. This thesis is motivated by the critical challenge of adapting beamforming techniques to the rapidly changing positions of users, a task analogous to hitting a moving target with precision. The aim is to significantly improve cellular network performance by leveraging advanced beamforming and machine learning (ML) for precise user localization. A novel dataset, crucial to this endeavor, has been developed from simulations in open spaces and a digital twin of the University of Glasgow campus, incorporating vital parameters such as direction of arrival (DoA), time of arrival (ToA), and received signal strength indicators (RSSI). Our investigation commences with the deployment of Maximum Ratio Transmission (MRT) and Zero Forcing (ZF) beamforming techniques to evaluate their effectiveness in enhancing network efficiency through both real and simulated user locations. The application of an adaptive MRT algorithm in our beamforming strategy resulted in a remarkable 53% increase in Signal-to-Noise Ratio (SNR), showcasing the potential of contextual beamforming (Cont-BF) using location information. Furthermore, to refine localization accuracy, deep neural networks were employed, achieving a localization error of less than 1 meter surpassing conventional methods in accuracy.

This research also introduces technique for user-assisted beam alignment in high-speed scenarios, addressing the challenges in dynamic transport systems. Venturing beyond traditional approaches, it explores the integration of user locations into beamforming decisions via a P4 switch, crafting a dynamic system responsive to user mobility. This is complemented by extensive data collection from 5G base stations (BS) using a TSMA 6 scanner, which enriches our analysis with detailed parameters for precision localization. Moreover, the study evaluates various MIMO beamforming techniques in 5G networks, demonstrating an average throughput increase from 9 Mbps to 14 Mbps, thereby underscoring the effectiveness of our proposed solutions. The potential of low-cost Software Defined Radios (SDR) forDoA estimation and the design of a beam steering setup was also assessed, aiming to evaluate their utility in highfrequency beamforming. Despite uncovering limitations in sub-6GHz environments, this exploration led to the successful development of a DoA estimation setup using USRPs and antennas, alongside a beam steering system crafted through the design of phase shifters and antennas. By integrating precise location information into adaptive beamforming techniques, especially within the dynamic context of transport systems, this thesis underscores the imperative role of such integration in significantly enhancing communication efficiency. Our findings, which include significant improvements in signal-to-interference-to-noise ratio (SINR) (up to 50%) and received power (up to 40%) through advanced beamforming methods, are pivotal for advancing high-demand applications, including smart vehicles and immersive virtual reality. This marks a crucial advancement towards the realization of next-generation cellular networks, paving the way for more efficient and reliable performance in an evolving technological landscape.

Item Type: Thesis (PhD)
Qualification Level: Doctoral
Subjects: T Technology > T Technology (General)
Colleges/Schools: College of Science and Engineering > School of Engineering > Electronics and Nanoscale Engineering
Supervisor's Name: Abbas, Dr. Hasan Tahir, Abbasi, Professor Qammer and Imran, Professor Muhammad
Date of Award: 2024
Depositing User: Theses Team
Unique ID: glathesis:2024-84223
Copyright: Copyright of this thesis is held by the author.
Date Deposited: 10 Apr 2024 14:25
Last Modified: 10 Apr 2024 14:26
Thesis DOI: 10.5525/gla.thesis.84223
URI: https://theses.gla.ac.uk/id/eprint/84223
Related URLs:

Actions (login required)

View Item View Item


Downloads per month over past year