Repository logo
 

Vehicular Dynamic Grouping: Virtualization and Network Re-orchestration

Files

Thesis(5.88 MB)

Date

2023

Authors

Al-Hamid, Duaa Zuhair Abduljabbar

Supervisor

Chong, Peter

Item type

Degree name

Doctor of Philosophy

Journal Title

Journal ISSN

Volume Title

Publisher

Auckland University of Technology

Abstract

The dynamic and heterogeneous features of vehicular networks (VNs), coupled with rapid data flow within the topological structure of VNs, present challenges for sustaining network connectivity on the road and improving the vehicular network system architecture. This calls for an adaptive and flexible network mechanism that can keep up with the dynamics of the network in order to maintain communication between the vehicles and the transport management system and structure an intelligence-enabled network with ever-increasing traffic data demands. For instance, the dynamics of highways encourage the need for a real-time and self-organizing mechanism to handle the rapid events of vehicles entering and exiting the road. Such demands are aligned with advancements in software-defined networking (SDN) and related dynamic network re-orchestration to enable rapid response. This, in turn, promotes the concept development of software-defined vehicular wireless sensor networks (SDVWSN), where the core WSN functions are integrated into the vehicular network through softwarization. This coupled with cloud-based virtualization, which enables specific functional configuration parameters to be generated and provided to physical nodes, can be seen as a highly promising solution that can adapt to the service requirements associated with such dynamic vehicular settings. As the topological network structure requires adaptation to the road traffic and its dynamic challenges, traffic analysis modelling and full network connectivity can be provided using powerful communication techniques such as intelligent clustering (grouping) to provide a flexible network structure and connection maintenance. The objective of this thesis is to provide a flexible vehicular grouping structure through the use of WSN core functions (i.e., ‘IoT gateway’, ‘Router’, and ‘Leaf’). Virtualizing the structured group based on the programmability/softwarization of the nodes to have one or more than one function, using the Contiki-Cooja simulator, can provide the network with the flexibility to react to dynamic events through the re-orchestration with less or no network downtime. One of the other main objectives of this thesis is to identify the size of the group, which can be adaptable to the road structure and traffic scenarios, such as the road capacity of the highway, as well as the communication requirements. This thesis has the following main contributions: • An analytical model for the road traffic scenario (focusing on the highway case) is proposed and presented. This model reflects the capacity of the road and the possible distribution of vehicles on a road with a given number of lanes. Factors such as the road stretch, the number of lanes, the 2-second safety distance are considered in the model. This model can support the size of a vehicular group, which can be structured and re-structured on the road. • The system architecture, which includes virtualization and network functions, is proposed to provide a flexible and controlled grouping structure prior to implementation. The proposed architecture employs softwarization that helps in testing the functions of the network to provide a working model in the virtual platform. • The thesis proposes the self-X phases (i.e., self-formation, self-leaving, self-healing, and self-joining) for structuring and healing a vehicular group through the concept of virtualization. This approach offers a flexible and re-orchestrated group on the go. • Utilizing the physical network for proof of concept provides the vision of a cyber-physical system. The virtual network structure where the model is structured and tested can provide the parameters and performance required to cope with the physical implementation. The main network evaluation and performance metrics focus on testing the ideal size of the vehicular group identified based on a given road capacity. Herein, a group size of 10 vehicles with a message rate of 1 to 5 messages/second delivers 100% received messages. According to our developed analytical latency model, the hop latency obtained by Cooja is associated to the computation of the delay involved in the hop. Herein, the hop latency of 5 vehicles connected to a router node is 1 second. This is less time compared to the communication latency observed during the pre-election process. From the network re-orchestration latency point of view, research findings suggest that testing needs to focus on the ‘Event Trigger’ and ‘Event Response’ stages to reflect the time it takes for the network to go through recovery. The physical network testing and evaluation focuses on group size. Herein, the physical network experienced slightly more packets loss due to real-world factors impacting the quality of the RSSI signal, such as the network deployment environment, obstacles, etc. However, the analysis reflects the importance of the soft trials on the virtual platform to identify the implementation test requirements of the physical network. In a nutshell, this thesis proposes the formulation of a vehicular network dynamic grouping strategy that can adapt to dynamic events and re-orchestration demands by attaining the required response in group re-organization when the vehicular network dynamic changes occur on the road. These events pose the challenge of initiating, forming, and maintaining any structured group (such as on-going healing in response to a rupture caused by the departure or failure of network components within the group) on the road with less or no network downtime. Dealing with a structured group, such as group size, that experiences these dynamics requires analysis based on its adaptability to the road and communication requirements.

Description

Keywords

Source

DOI

Publisher's version

Rights statement

Permanent link

https://hdl.handle.net/10292/16220

Collections

Doctoral Theses