5G communication framework for smarter autonomous vehicles

• Main Author: Mavromatis, Ioannis
• Other Authors: Piechocki, Robert ; Nix, Andrew
• Published: University of Bristol 2019
• Online Access: https://ethos.bl.uk/OrderDetails.do?uin=uk.bl.ethos.767965

In recent years, the interest around Intelligent Transportation Systems (ITSs) has rapidly increased, and they are currently being deployed in commercial, industrial and residential domains. The Next-Generation Cooperative ITSs (C-ITSs) are expected to bring the paradigm of Mobility-as-a-Service (MaaS) to a whole new level. A crucial role to this is attributed to Connected and Autonomous Vehicles (CAVs), as the focus is steered from being autonomous systems to becoming cooperative entities. The cooperation between them is mainly enabled by sharing sensor data and manoeuvring intentions in a Vehicle-to-Everything (V2X) fashion. Thus, an efficient, reliable and robust communication plane between CAVs and the infrastructure network, able to accommodate the demanding exchange of sensor data, is of paramount importance. This thesis presents the author's work and contributions in the field of connectivity for CAVs within the scope of an ITS. As connectivity is a broad term, it can be examined from different perspectives. This thesis focuses on the design of a novel heterogeneous C-ITS framework that can handle scalable city-wide ITS application data streams. Later, it identifies the various drawbacks and limitations of the current Radio Access Technologies (RATs), and investigates ways to overcome them and improve the system performance. The contributions of this thesis commence with the introduction of a conceptual design of a city-scale C-ITS implementation. Briefly, this proposed system operates as a heterogeneous network, maps different application data streams on to different V2X layers, and employs different RATs to deliver each stream in a V2X manner. Based on this fundamental concept, various drawbacks are primarily identified on existing simulation models and tools for IEEE 802.11p. Later, the existing limitations are addressed by designing an experimental testbed and conducting a sizeable experimental campaign. This not only helps with the accurate calibration of the existing models, but also provides a deep comprehension of how a real-world large-scale implementation works. In the later parts of this thesis, the focus is directed towards Millimetre Waves (mmWaves), as is a competent technology for accommodating significant amounts of exchanged sensor data. At first,the beamforming problems in mobile environments are identified. They are approached through an intelligent MAC-layer solution for V2X beam steering. Furthermore, the resource allocation problem at the highly dynamic network topology of CAVs is investigated, designing a multi-link association scheme for V2V mmWave communications. Finally, the last research effort of this thesis, is related to the city-scale positioning of the mmWave Road-Side Units (RSUs), where an optimal automated procedure within an urban scenario is proposed. The thesis is concluded with the critical review of the research activities mentioned above as well as some ideas for future research.