Please use this identifier to cite or link to this item: https://hdl.handle.net/10356/157210
Title: Wireless channel modeling and spectrum monitoring for interference mitigation and link reliability insurance for existing and future CBTC systems
Authors: Kalyankar Shravan Kumar
Keywords: Engineering::Electrical and electronic engineering::Wireless communication systems
Issue Date: 2021
Publisher: Nanyang Technological University
Source: Kalyankar Shravan Kumar (2021). Wireless channel modeling and spectrum monitoring for interference mitigation and link reliability insurance for existing and future CBTC systems. Doctoral thesis, Nanyang Technological University, Singapore. https://hdl.handle.net/10356/157210
Abstract: The train signal failure (also known as wireless loss or communication failure) results in the disruption of train services and remains a primary performance bottleneck in the Communication-Base Train Control (CBTC) system. The signal failure in the CBTC system can be caused by the malfunctioning of radio equipment (faulty access points), weak signal (inadequate network coverage), and/or wireless interference. The weak signal problem can be addressed by performing wireless channel characterization, modeling, and optimizing the access point parameters. A simulation and measurement-based study of a wireless channel along the MRT tunnels, known to be the most complex radio propagation environment, is presented in this thesis. This study suggests the existence of two wave propagation mechanisms in the curved tunnel; enhanced waveguiding mechanism induced by rich multipath components from the curved tunnels and degraded waveguiding mechanism due to the blockage from the curved tunnel walls. For efficient radio planning, a new propagation model with a curvature-dependent breakpoint is proposed. The proposed breakpoint indicates the end of the enhanced waveguiding mechanism and the beginning of the degraded waveguiding mechanism. A two-slope radio wave propagation model is proposed for radio communications inside curved tunnels using the determined breakpoint with performance evaluation. The variation in the rich multipath propagation in a curved tunnel compared to the straight tunnel with respect to Fresnel theory and the blockage of radio waves because of tunnel curvature is well explained with the help of ray-tracing simulation and wideband measurements. CBTC operators use a 2.4 GHz ISM frequency band to reduce the operating cost. However, this introduces a significant risk of wireless interference from other ISM band applications to the CBTC signals. Therefore, an extensive spectrum monitoring campaign is performed along the Singapore MRT train route to understand the risk better. The wireless interference is quantified using spectrum occupancy, and the potential sources of interference to the CBTC system are identified as wireless CCTV systems installed in the MRT stations and the WiFi systems along the MRT train route. As a result, it is found that the wireless interference in the MRT environment is higher at the MRT stations compared to in-between MRT stations. In addition, the urban areas have a more significant risk of interference than the industrial area and the tunnel section. The biggest challenge in spectrum monitoring studies is to identify the CBTC signals and the interference signals. A simple hand-crafted spectral signature-based signal identification algorithm is proposed to overcome the signal identification problem. This technique is validated by performing signal identification on extensive measurement data collected from spectrum monitoring. In case of challenges in extracting many spectral signatures from the measured data, we proposed using GAN to generate the synthetic spectral signature. Furthermore, the hand-crafted spectral signatures are replaced with the synthetic spectral signatures, and the identification of signals is performed using a deep learning algorithm. As a continuation to spectrum monitoring campaign, spectrum occupancy analysis, signal identification, a Gaussian Mixture Model (GMM) clustering technique is proposed to evaluate the performance of access points from identified CBTC signals when the train is stopped at the MRT station. The statistics of the clustered data are analyzed to understand the characteristics of radio signals from different access points, and the findings of this technique are presented. The study shows that the proposed approach can be used in real-time monitoring of access points (APs) and help the MRT operators identify the faulty APs without any in-field tests. In summary, we have studied the weak signal problem in tunnel environments and proposed the wireless channel models that can help to optimize the access points and improve coverage. Furthermore, the spectrum monitoring study, proposed signal identification, and source identification algorithms can be the beginning step towards implementing the real-time condition monitoring of radio infrastructure in the CBTC system, which can reduce train signal failures.
URI: https://hdl.handle.net/10356/157210
DOI: 10.32657/10356/157210
Schools: School of Electrical and Electronic Engineering 
Research Centres: SMRT NTU Smart Urban Rail Corporate Lab
Rights: This work is licensed under a Creative Commons Attribution-NonCommercial 4.0 International License (CC BY-NC 4.0).
Fulltext Permission: open
Fulltext Availability: With Fulltext
Appears in Collections:EEE Theses

Files in This Item:
File Description SizeFormat 
Thesis_combined pt1&2.pdf132.22 MBAdobe PDFThumbnail
View/Open

Page view(s)

240
Updated on Mar 28, 2024

Download(s) 50

101
Updated on Mar 28, 2024

Google ScholarTM

Check

Altmetric


Plumx

Items in DR-NTU are protected by copyright, with all rights reserved, unless otherwise indicated.