Intelligent architectures and tools for Internet of Vehicles

点击翻译按钮获取中文摘要

The Internet of Vehicles (IoV) is defined as an integration of three networks: an inter-vehicle network, an intra-vehicle network and a vehicular mobile Internet. Based on this concept of three networks integrated into a single one, we can define an Internet of Vehicles as a large-scale distributed system for wireless communication and information exchange between vehicles, roadside units and Internet, and humans, according to agreed communication protocols and data interaction standards. Vehicular networks rely on gathering information from spatially distributed heterogeneous sources for applications such as traffic monitoring, infotainment, etc. Although these resources can provide rich information, a vast and overwhelming amount of data are usually generated. In IoV, the rapidly growing number of connected components is creating data at an exponential rate. A connected car, for example, can create ten megabytes of data per second. Sending all the data to the cloud will require prohibitively high network bandwidth and encounters growing challenges in meeting many new quality of service (QoS) requirements. The main objective of this research program is to design and evaluate intelligent models, protocols and architectures for achieving real-time and high accuracy situational awareness for cooperating vehicles in context-aware environments. More specifically, this program research aims at:1. Defining a set of requirements for an intelligent IoV architecture supporting seamless integration of heterogeneous devices used in different smart applications and in intelligent transportation system domains;2. Designing an architecture that is not only capable of supporting dynamic topology changes, scalability, real time and availability but which also enables efficient and reliable data sharing between devices and services of the IoV system;3. Prototyping and evaluating the performance of the proposed architecture using simulations and comparisons with the predefined set of requirements;4. Designing and evaluating the performance of a model that distributes the resources and services of computation, communication, control and storage closer to the users in order to enhance overall system efficiency, reliability and performance;5. Proposing and evaluating a new cognitive information gathering mechanism in a context-aware environment for urban traffic management; 6. Designing and validating an intelligent microscopic traffic mobility model dedicated to connected and automated vehicles, capable of taking into account drivers' behaviors and vehicles trajectories.We will conduct simulations to assess performance and accuracy. To evaluate objective 5, Probabilistic Nave Bayesian, hidden Markov models, as well as spatio-temporal data mining will be used. Finally, this research program will directly support three masters and two doctoral students.