Evaluation of VANET protocols for driving assistance of connected vehicles: a strategy based on a multi-anticipative car-following model

Abstract

In recent years, Vehicular Ad-hoc Networks (VANETs) have become a remarkable research area due to their potential to improve communication among people and to become an alternative offload channel for traditional networks. The main idea behind a VANET is to harness the spatial distribution of vehicles and advanced technical resources in communication networks to establish a self-managed and reliable channel to communicate with vehicles and any device able to access the network. Due to their independent nature, VANETs have also been envisioned for vehicular mobility solutions, more precisely with the só-called Cooperative Advanced Driver Assistance Systems (CADAS). In such systems, each vehicle is expected to gather its closest neighbor’s kinematic information by transmitting short and periodic messages, known as beacons, whose content may contribute not only to decision-making on mobility but also to improving the own vehicular network’s stability. Traditional research approaches address communication with VANETs and mobility assistance systems independently, só little attention has been given to studying the coexistence of CADAS and end-user communication. This thesis presents a comprehensive study on the practical impacts that controlling the mobility of a connected vehicle through periodic messages in VANETs has on the network performance. In addition, a network-based and traffic-aware strategy for driving assistance of electric and connected vehicles through a car-following model is proposed. The results show that a traditional protocol featuring beacons, like the Greedy Perimeter Stateless Routing (GPSR), can perform better than a set of approaches using the Delay Tolerant Network (DTN) protocol. However, a traffic-aware beacon strategy is necessary to reduce the negative impact a short time interval of beacons has on the network.