A Paradigm Shift from Vehicular Ad Hoc Networks to VANET-Based Clouds

Abstract

Vehicular ad hoc network (VANET) is expected to improve our driving experience through enhanced safety, security, robustness, and infotainment. Nevertheless, despite considerable amount of research, VANET did not make it, at least not on a full scale, to the deployment stage because of many issues including security and privacy. However it is speculated that in the future high-end vehicles, on-board computation, communication, and storage resources will be under-utilized. Therefore, recently a new paradigm shift from conventional VANET to vehicular cloud computing was envisioned. This paradigm shift was realized through merging VANET with revolutionary cloud computing. Clearly cloud computing is one of today's tempting technology areas due, at least partially, to its virtualization and cost-effectiveness. However, to date the potential architectural framework for VANET-based cloud computing has not been defined so far. To fill this gap, in this paper, first we put forth the taxonomy of VANET based cloud computing and then define a communication paradigm stack for VANET clouds. Additionally we divide VANET clouds into three architectural frameworks namely vehicular clouds (VC), vehicles using clouds (VuC), and hybrid vehicular clouds (HVC). Each proposed framework provides particular set of services depending upon the underlying communication paradigm. To understand our proposed framework well, we also propose a novel use-case service of the VANET-based cloud namely traffic information dissemination through clouds. In the proposed scheme, vehicles moving on the road are provided with fine-grained traffic information by the cloud as a result of their cooperation with the cloud infrastructure. Vehicles share their frequent mobility dynamics with the cloud and cloud in turn provides them with long range traffic information based on their current and near-future locations. Our simulation results show that the traffic information dissemination through cloud is feasible and the vehicles receive above 83 % of the traffic information from clouds through gateways in worst-case scenarios (i.e. extensive dense traffic) and above 90 % traffic information in average case scenarios. Finally we also outline the unique security and privacy issues and research challenges in VANET clouds.