NeTS: Small: VC-VANET: A Sustainable Vehicle-Crowd Based Vehicular Ad Hoc Network Supporting Mobile Cloudlet Computing

University Of Alabama Tuscaloosa, Tuscaloosa AL

Investigators

Abstract

The development of new and innovative methods for smart transportation is a critical need of the nation as vehicles continue to evolve. Not long from now, we will see more autonomous vehicles and connected vehicles on the road systems in each state. It is predicted that by 2020 there will be approximately 250 million connected vehicles on the road. Such a vehicular environment, with communication capabilities from the moving vehicles and the transportation systems, presents an extremely large and complex mobile networking scenario where multiple communication technologies are available. The goal of this research is to investigate a vehicle-crowd centered, networked system that has the capability to support high-demand mobile edge computing applications. The benefits could range from more prompt and accurate data inputs and result outputs, to being an alternative for infrastructure-based systems. The applications could include: intelligent transportation system applications, transportation related compute-hungry, time-critical situation-aware applications, and visualizations. Results of this effort will be disseminated to multiple disciplines, and be used to develop education and training materials for both graduate and undergraduate students as well as provide research opportunity for traditionally under-represented groups and undergraduate students. The proposed research will advance the knowledge about the interdependent relationships between the physical vehicle network and the transportation operation for supporting high-demand applications. This project envisions the capability to support these applications through harvesting the computing and storage resources by collaborations among grouped vehicles which are knitted together via the vehicular networking protocols. Such a network architecture is hierarchical as these groups of connected vehicles are an additional layer on top of vehicle ad hoc networks and vehicle delay tolerant networks of individual vehicles. The unique emphasis of this research will be finding the vehicle-crowd system properties relating to signal control among other factors, and developing adaptive vehicle-crowd schemes in reacting to adverse conditions using data collected from local city transportation systems. This project focuses on three main research tasks: (1) Investigate the properties of an individual vehicle-crowd, namely, the dynamic patterns of its occurrence at a given location in the time domain, and the latency when recovery is desired. (2) Investigate the spatial and temporal properties relating to multiple vehicle-crowds, and how the occurrences coordinate. (3) Investigate prediction methods to help sustain the vehicle-crowds.