Mobility-aware dynamic offloading strategy for C-V2X under multi-access edge computing

Abstract

Multi-access edge computing (MEC) technology is envisioned as a promising paradigm to achieve the user needs of low-latency applications. Complex computation tasks are offloaded from resource-constrained devices to resource-rich devices to reduce task completion delays. However, given the high-speed mobility of vehicles, the traditional MEC network architecture is insufficient for the Internet of Vehicle (IoV) computing scenarios. Besides, many resource-rich idle vehicles can be used as mobile servers to perform computing tasks, extending the choices for vehicle users. To better integrate MEC and IoV technologies, we introduce a Cellular-V2X-based MEC offloading scenario, enabling MEC servers and vehicle nodes in the 5G cellular network environment to perform offloading cooperatively. For a typical delay-sensitive task, taking the Augmented Reality (AR) application as an example, we divide the task into multiple subtasks with linear correlation. A mobility-aware dynamic offloading algorithm (MADO) is proposed to minimize the impact of vehicular mobility on offloading. The algorithm is used to find an optimal resource allocation, which reduces the overhead of the entire offloading process. It continuously updates the resource selection and the offloading strategy with the vehicular location changes. Simulation results corroborate that our proposed MADO algorithm can effectively reduce task completion time, improve task completion success rate and adapt to a dynamic environment.