No abstract

With the increasing demand for real-time computing services in the Internet ofVehicles(IoV) and the inability to meet the latency requirements of IoVapplications by using cloud computing, multi-access edge computing(MEC) isconsidered a promising paradigm to address the latency-sensitive andcomputation-intensive requirements for meeting IoV applications. We considerthe dependencies in IoV applications and design a hybrid cloud-edge architectureto ensure adaptability and flexibility. We propose a scheduling algorithm thatoffloads tasks to the eNode or cloud for execution after sorting them by urgency.Experimental results show that our algorithm outperforms existing algorithms interms of average completion time, average waiting time, and in-time completionrate of the application, which can provide a better quality of service(QoS) for IoV.

With the skyrocketing need for low-latency services on the Internet of Vehicles (IoV) and elastic cross-layer resource provisioning, multi-access edge computing (MEC) is considered a high-potent solution, which evolves from cloud and grid computing to meet the above needs in IoV scenarios. Instead of considering single-point and monolithic IoV tasks, in this paper, we consider the IoV applications to be with structural properties and the supporting environment to be with a hybrid cloud-edge architecture. We develop a scheduling method that offloads tasks to the eNode or cloud according to their estimations of latest starting time. Simulative results clearly demonstrate that our method beat existing solutions in terms of average completion time, average waiting time, and in-time completion rate.