Investigation of Vehicle-to-Everything (V2X) Communication for Autonomous Control of Connected Vehicles

Dhawankar, Piyush Shamroa (2021) Investigation of Vehicle-to-Everything (V2X) Communication for Autonomous Control of Connected Vehicles. Doctoral thesis, Northumbria University.

Preview

Text (Doctoral Thesis) dhawankar.piyush_phd_12037572.pdf - Submitted Version Download (4MB) | Preview

Abstract

Autonomous Driving Vehicles (ADVs) has received considerable attention in recent years by academia and industry, bringing about a paradigm shift in Intelligent Transportation Systems (ITS), where vehicles operate in close proximity through wireless communication. It is envisioned as a promising technology for realising efficient and intelligent transportation systems, with potential applications for civilian and military purposes. Vehicular network management for ADVs is challenging as it demands mobility, location awareness, high reliability, and low latency data traffic.

This research aims to develop and implement vehicular communication in conjunction with a driving algorithm for ADVs feedback control system with a specific focus on the safe displacement of vehicle platoon while sensing the surrounding environment, such as detecting road signs and communicate with other road users such as pedestrian, motorbikes, non-motorised vehicles and infrastructure. However, in order to do so, one must investigate crucial aspects related to the available technology, such as driving behaviour, low latency communication requirement, communication standards, and the reliability of such a mechanism to decrease the number of traffic accidents and casualties significantly. To understand the behaviour of wireless communication compared to the theoretical data rates, throughput, and roaming behaviour in a congested indoor line-of-sight heterogeneous environment, we first carried out an experimental study for IEEE 802.11a, 802.11n and 802.11ac standards in a 5 GHz frequency spectrum. We validated the results with an analytical path loss model as it is essential to understand how the client device roams or decides to roam from one Access Point to another and vice-versa. We observed seamless roaming between the tested protocols irrespective of their operational environment (indoor or outdoor); their throughput efficiency and data rate were also improved by 8-12% when configured with Short Guard Interval (SGI) of 400ns compared to the theoretical specification of the tested protocols.

Moreover, we also investigated the Software-Defined Networking (SDN) for vehicular communication and compared it with the traditional network, which is generally incorporated vertically where control and data planes are bundled collectively. The SDN helped gain more flexibility to support multiple core networks for vehicular communication and tackle the potential challenges of network scalability for vehicular applications raised by the ADVs. In particular, we demonstrate that the SDN improves throughput efficiency by 4% compared to the traditional network while ensuring efficient bandwidth and resource management. Finally, we proposed a novel data-driven coordination model which incorporates Vehicle-to-Everything (V2X) communication and Intelligent Driver Model (IDM), together called V2X Enabled Intelligent Driver Model (VX-IDM). Our model incorporates a Car-Following Model (CFM), i.e., IDM, to model a vehicle platoon in an urban and highway traffic scenario while ensuring the vehicle platoon's safety with the integration of IEEE 802.11p Vehicle-to-Infrastructure (V2I) communication scheme. The model integrates the 802.11p V2I communication channel with the IDM in MATLAB using ODE‐45 and utilises the 802.11p simulation toolbox for configuring vehicular channels. To demonstrate model functionality in urban and highway traffic environments, we developed six case studies. We also addressed the heterogeneity issue of wireless networks to improve the overall network reliability and efficiency by estimating the Signal-to-Noise Ratio (SNR) parameters for the platoon vehicle's displacement and location on the road from Road-Side-Units (RSUs). The simulation results showed that inter-vehicle spacing could be steadily maintained at a minimum safe value at all the time. Moreover, the model has a fault-tolerant mechanism that works even when communication with infrastructure is interrupted or unavailable, making the VX-IDM model collision-free.

Actions (login required)

View Item

Downloads