A Novel Path Planning-Aware Vehicular Task Offloading Strategy

Abstract

In intelligent transportation system (ITS), the combination of vehicle network (VN) and mobile edge computing (MEC) can provide low delay networking services to vehicles. Since the quality of vehicle access network is mainly determined by location, the service quality of vehicle computation task unloading (CTO) should be considered in the process of vehicle path planning, so as to further improve the performance of a single vehicle in unloading throughput, delay, and cost. In this paper, with employing the SDN-based system architecture, we propose a novel path planning-aware vehicular task offloading strategy for connected vehicles in real-time by considering CTO to provide drivers a good driving experience. To achieve the efficient path planning, an effective road performance evaluation method is proposed, which combines the road capacity and the throughput of computation tasks into the driving index to evaluate the priority selected by vehicle drivers. The simulation results show that the proposed method can guarantee the driving time of the vehicle, while it can also give the driver a better quality of experience.