Recent years have witnessed the emergence of Vehicular Ad Hoc Networks, also known as VANETs, fueled largely by growing interest in Intelligent Transportation Systems. They are specialized from well-known Mobile Ad Hoc Networks (MANETs) to include vehicle-to-vehicle and vehicle-to-infrastructure communications. Vehicular networks are typically characterized by their decentralized and autonomous setup, highly dynamic network topology, varying vehicle density, and intermittent route connectivity. Enabled by short-range communication subsystems, vehicular networks are targeted at safety applications, real-time emergency warning systems, mobile infotainment, traffic congestion management, localized traffic updates, and automatic highway toll collection. Several of these applications impose strict quality of service requirements for message transmission. For example, traffic safety and emergency warning systems require messages to be delivered in a reliable and timely manner. However, the specific characteristics of VANETs make it extremely challenging to design high-throughput, delay-efficient, robust, and scalable methods for data delivery solutions.
In this dissertation, the efficacy of cross-layer design that enables the join optimization of different layers is investigated and demonstrated. Such a cross-layer design is further complemented with location-aware and delay-aware techniques so that the issues related to dynamic network topology and application-level delay constraints are effectively handled.
First, an overview of design challenges and implementation issues in developing cross-layer protocols and packet forwarding methods for vehicular networks are introduced. An extensive study on the design architecture of existing solutions and their limitations are discussed. Second, a delay-aware and location-aware communication protocol called PROMPT for multi-hop data delivery service from vehicles to base stations is proposed. PROMPT employs a cross-layer design, jointly optimizing the routing and Media Access Control (MAC) functions, to discover stable and delay-efficient routes in a highly dynamic communication network. PROMPT relies on a novel delay estimation model, which makes use of various packet traffic statistics collected by the MAC layer from different locations in the network, to determine routes with small end-to-end delay.
Third, a mobility management scheme called MMDD in which the base stations are kept updated with the locations of different vehicles in the network is proposed. Such a scheme is used in reliably transmitting the data packets from base stations to vehicles. MMDD explores the trade-off between location information accuracy and control packets overhead. Fourth, a detailed empirical simulation analysis of both PROMPT and MMDD protocols is performed, and it is demonstrated that they perform significantly better than state-of-the-art approaches. Finally, this dissertation is concluded with the exploration of several open research problems in developing efficient cross-layer protocols for next generation transportation systems.