Projet de thèse en Sciences des réseaux, de l'information et de la communication
Sous la direction de Antoine Berthet et de Antonella Molinaro.
Thèses en préparation à université Paris-Saclay en cotutelle avec Mediterranea University of Reggio Calabria (UNIRC) , dans le cadre de École doctorale Sciences et technologies de l'information et de la communication , en partenariat avec Laboratoire des Signaux et Systèmes (laboratoire) , Télécoms et Réseaux (equipe de recherche) et de Faculté des sciences d'Orsay (référent) depuis le 01-10-2019 .
Voir la présentation détaillée du projet doctoral en anglais.
5G and beyond Vehicle to Everything (V2X) communications
Vehicle-to-Everything (V2X) communication is the key enabler for a safer, greener, connected and automated transport [1]. After more than a decade of research and standardization efforts, the V2X technology is ready today, with the cellular system preparing itself to play a prominent role in the immediate future. The growing industrial interest in V2X is witnessed by manifold initiatives like the creation of the 5G Automotive Association (5GAA), wherein the major car manufacturers and ICT players (Audi AG, BMW Group, Daimler AG, Ericsson, Huawei, China Mobile, ZTE, Intel, Nokia, Qualcomm Inc., etc.) promote interoperable end-to-end solutions based on cellular V2X (C-V2X) [12] and the New Radio (NR) access technology [4], currently under specifications by the Third Generation Partnership Project (3GPP) to provide the ultra-high reliability (99.999%) and ultra-low latency (1 ms) demands of 5G V2X applications (e.g., vehicle platooning, advanced driving, extended sensors, remote driving). Worldwide field-trials demonstrated the feasibility of the IEEE 802.11 and the cellular technology families to support basic vehicular applications (e.g., road hazard warnings) [2], while there is still a lot of work to do for the current radio access technologies to be able to support advanced V2X applications, especially under high-speed and high-density conditions. Lots of challenges raise, from the frequency band allocation, to the way of guaranteeing service continuity across borders, to the enhancement of the radio interface and the resource scheduling, to the design of multihop relaying strategies etc. The NR design will also encompass flexible numerologies and frame structure, and entail the investigation of new waveforms, Non-Orthogonal Multiple Access (NOMA) strategies [5], advanced (nonlinear) receivers capable to cancel interference, new media like millimeter wave (mmWave) [6] to provide broadband communications and massive connectivity. In-band full-duplex techniques [7] are going to revolutionize the wireless communication performance, and represent an attractive technological solution for V2X. Such technologies could play a key role to bring high data rate services seamlessly to vehicles, to enable cooperative sensing and manoeuvring for fully automated vehicles and support high density platooning with ultra-low latency and high reliability. However, the entire end-to-end chain of radio, networks, applications and services needs to be tailored to meet the V2X daunting demands. Besides a competitive radio access segment, V2X communications entail heavy computing capabilities to be provided at the network edge, in order to process in near real time a big amount of data generated by vehicles and contributing to enhance the driver and traveling experience. Multi-Access Edge Computing (MEC) [8] is a key enabler to ensure such a vision, since it provides cloud-like (computing, storage) resources to the edge close to where the data is being generated and also likely consumed, after being processed. V2X applications deployed at the edge can leverage low-latency communication with other vehicles in the covered area and can benefit from additional (context) information via data fusion from multiple available sources and accurate positioning. For the take-off of MEC solutions for V2X applications, main issues concern the placement of virtualized services at the edge to ensure low-latency while the vehicles move, which entails the design of smart migration procedures and lightweight virtualization technologies [9]. Cloudification will affect not only the edge but the entire network: the underlying network procedures (i.e., network functions) need to be virtualized and deployed in cloud-based environments. Such a deployment option will allow for a more optimized usage of the infrastructure resources and to dynamically and flexibly adapt to the traffic demands of different vertical scenarios. User plane (UP) and Control plane (CP) functionalities can be deployed as virtual network functions (VNFs) and be independently scaled and instantiated in convenient locations to flexibly support various services. While distributed UP functions at edge cloud facilities reduce latency, edge-located CP functions benefit from radio-related information and mobility-awareness in a prompt and agile manner, which is crucial for V2X applications. Innovative Radio Access Network (RAN), edge and core components design will be capitalized in the network slicing paradigm, which allows customer-specific network behaviors and configurations, and bundle VNFs on top of the same shared physical infrastructure [10]. Slicing is recognized by the V2X stakeholders as a key element to support innovative applications in the automotive vertical market [11] [12]. The V2X ecosystem entails the simultaneous support of multiple heterogeneous services with diversified requirements, in a multi-tenant, multi-technology, multi-operator, multi-device, multi-RAT scenario. Such unique features make the design of V2X slices much more complicated than for other verticals.