Date of Award
Doctor of Philosophy (PhD)
Electrical & Computer Engineering
Modeling and Simulation
Dimitrie C. Popescu
With the advancement of intelligent transportation systems, autonomous driving and connected driving are seen as potential solutions to alleviate traffic congestion, enhance traffic safety, and improve efficiency. Extensive testing and validation of autonomous vehicles (AVs) and connected vehicles (CVs) including connected autonomous vehicles are crucial to ensure their safety and reliability. However, testing and validating AVs and CVs on public roads faces challenges such as high costs, inadequate support from transportation infrastructure with communication technologies, and safety concerns, among others. Simulations have become essential tools for testing autonomous driving and connected driving. As mixed traffic involves multiple domains including traffic flow, vehicle dynamics, and communication, a simulation system encompassing these domains needs to be developed to support the testing of AVs and CVs in a mixed traffic environment. With this objective in mind, this dissertation aims to develop a comprehensive simulation system that serves as a testing platform for AVs and CVs.
This dissertation began by utilizing a high-level architecture to integrate an extensible simulation system. Specifically, this dissertation integrated CARLA, a state-of-the-art autonomous driving simulator, with Eclipse MOSAIC, a Vehicle-to-Everything (V2X) co-simulation framework. Two approaches, namely the TraCI-Relay approach and the MessagePack approach, were employed in developing the CARLA integration. The TraCI-Relay approach utilized the Traffic Control Interface (TraCI) protocol for the communication protocol in the CARLA integration while the MessagePack approach employed a customized protocol developed using MessagePack binary serializer. The integrated simulation system was demonstrated, and the performance of the proposed simulation system was analyzed. The use cases illustrated the utility of the integrated simulation system for testing connected driving.
Furthermore, this dissertation utilized CARLA in conjunction with virtual reality (VR) technologies to introduce two human-in-the-loop autonomous driving simulation frameworks, i.e., a single-user and a multiuser human-in-the-loop frameworks to allow humans to actively participate in the autonomous driving simulation in an immersive manner. The prototypes of two simulation framework were developed in Unreal Engine. The multiuser human-in-the-loop autonomous driving simulation framework enables multiple users to participate in the autonomous driving simulation simultaneously.
Additionally, this dissertation proposed and developed a user-friendly graphical user interface (GUI) within CARLA. This GUI aims to facilitate setting up simulation scenarios for testing and validation of AVs and CVs in CARLA. The GUI includes a comprehensive configuration menu. The prototype of GUI was developed using user widgets and blueprints in Unreal Engine. The configuration menu enables users to configure various aspects of simulation environments such as simulation maps, weather conditions, road conditions, and traffic signal phase times, without extensive Python programming currently required by CARLA.
In Copyright. URI: http://rightsstatements.org/vocab/InC/1.0/ This Item is protected by copyright and/or related rights. You are free to use this Item in any way that is permitted by the copyright and related rights legislation that applies to your use. For other uses you need to obtain permission from the rights-holder(s).
"An Advanced Simulation Architecture for Testing Autonomous and Connected Vehicles Enabled by Virtual Reality"
(2023). Doctor of Philosophy (PhD), Dissertation, Electrical & Computer Engineering, Old Dominion University, DOI: 10.25777/xpj4-je79
Available for download on Friday, October 03, 2025