Modeling and enhancement of wireless access and mesh infrastructure vehicular networks

A major technological breakthrough to improve road safety and traffic efficiency  will be cooperative communications for transport systems.

Through the use of wireless communications, cooperative systems will allow the dynamic exchange of messages between transportation systems such as vehicle-to-vehicle (V2V) and vehicle-to-infrastructure (V2I) communications. Cooperative Vehicular Systems are composed by Road Side Units (RSUs) and On Board Units (OBUs). OBUs provide network connectivity to the users and are inside their cars whereas the RSUs are fixed and deployed equipment on the public roads. The RSUs can be connected to each other and to the Internet using Wireless Mesh Network backbone. We call "service level/access of the vehicular network" the network composed by the RSUs and the OBUs and we call "backbone level/infrastructure of the vehicular network" the network connecting the RSUs to the Internet. In this thesis, we develop several contributions to improve the vehicular network performances at both service level and backbone level. For the service level, 802.11p defines a control frequency channel for control and most critical data packets and one or several service frequency channels for less critical packets.

As a first step, we propose an analytical model for the 802.11p operations in the control frequency. It captures all suggested enhancements important for exchanging data packets generated by road safety applications. The model is a simple tool that is able to reproduce expected results. This is an important step toward the improvement of vehicular networks performances. The model is then combined with optimization criteria for optimal placement of roadside units. For the backbone level, we conduct a rich study to identify what are the significant factors and mechanisms that have the most important impact on the WMNs backbone performances. We concluded that investigating on channel allocation and routing is the best answer to that issue. After that, we focused and classified the most challenging approaches and proposals for channel allocations and routing. As a conclusion, we can say that the most interesting work are those proposing a common-layer design between MAC layer and network layer and are mostly concerning a centralized manner. However, in a wireless environment, presenting a centralized approach is not too realistic. Our contributions in this area concern the development of two interesting approaches to solve the addressed problem in a distributed fashion.