| This thesis develops an urban traffic management framework using distributed vehicle-to-vehicle communication and computing. Most of the existing vehicle navigation and route guidance systems rely on GPS-enabled location information, coupled with FM radio or cellular facilitated centralized content services. However, this mode of traffic management is expected to suffer from long term business scalability and reliability issues, due to its reliance on centralized paid services. As a consequence, the evolving technology of wireless vehicular ad hoc networks (VANET) is currently being perceived as a plausible communication platform for this application. Guided by this industry and research trend, this thesis proposes an infrastructure-less framework designed specifically for the purpose of traffic congestion alleviation and avoidance. In specific, the application layer algorithms have been developed, along with the underlying wireless VANET schemes suited for distributing the application layer induced traffic management information. The designed architectural components include mechanisms for congestion detection, broadcast based VANET routing for efficient congestion notification, vehicular transport oriented VANET routing used for connecting network partitions in sparse vehicular scenarios, and a constrained vehicle navigation algorithm for congestion mitigation. Extensive simulation studies have been performed, and the results indicate that the system is able to successfully manage vehicular traffic, while delicately handling the wireless broadcast medium. |
