On leveraging Named Data Networking for vehicular and edge computing applications.

Abstract

Modern vehicles are increasingly equipped with advanced processing, storage, and wireless communication capabilities, making them more intelligent and interconnected. In this context, Vehicular Ad Hoc Networks (VANETs) are crucial for enabling communication models among vehicles, distributed infrastructures, and monitoring devices. Despite these benefits, integrating VANET solutions into Smart Cities for Edge Computing scenarios requires a network architecture capable of handling heterogeneous communication requirements across distributed applications in multi-access environments. Vehicular mobility, dynamic communication models, and security aspects pose fundamental challenges to the development of scalable and reliable applications. Constant topology changes, caused by mobility, lead to intermittent connections that further complicate the design of distributed applications. Under these conditions, the TCP/IP architecture exhibits significant limitations. In contrast, emerging Information-Centric Networking models—and, in particular, the Named Data Networking (NDN) architecture—have arisen as promising alternatives, providing an optimized communication model with additional network-layer services like security support, content-centric delivery, and independence from the physical location of data and service providers. This thesis investigates the development of different classes of vehicular applications in realistic Smart City and Edge Computing scenarios, supported by the NDN architecture. First, it presents a study of vehicular networks, addressing their main characteristics, applications, and critical aspects. Then, it provides a detailed analysis of the data-centric communication model, highlighting its intrinsic properties and advantages over the TCP/IP model in VANETs. Finally, proof-of-concept implementations are proposed for representative distributed applications, showcasing practical design aspects and employing NDN as the underlying communication architecture between network entities. The main contributions of this thesis are as follows: (i) the development of an environment for simulating realistic applications in Vehicular Named Data Networking, the NDN4IVC simulator; (ii) the design of an Intelligent Transportation System, named NDN-Waze, for monitoring and optimizing vehicular traffic; (iii) the proposal of a service-oriented architecture for data offloading in Vehicular Edge Computing scenarios, called iETR (intelligent Edge-Traffic Routing), designed to efficiently transport large data volumes with the support of the NDN architecture and the orchestration of mobile agents (i.e., vehicles) integrated into the edge computing environment. The proposed solutions were evaluated through simulations, which demonstrated that the intrinsic properties of the NDN architecture favor the development of new classes of services in emerging networks.