Delay-Aware Task Scheduling for Multi-Access Edge Computing on the Internet of Vehicles

The demand for computing and networking in cars has grown with the advancement of the Internet of vehicles (IoV). Using multi-access edge computing (MEC) can deal with the issue of high latency in cloud computing. However, fast movement of cars and limited resources of MEC servers brings challenges. As a car moves into a cell (i.e., handoff), its MEC server may have no enough resources to serve the car's task. Therefore, the paper proposes a delay-aware task scheduling (DTS) scheme. When an MEC server has resources in short supply, some tasks are selected to be offloaded to nearby MEC servers. If a car is about to leave a cell, its task is offloaded to the MEC server in the car's target cell. Otherwise, we choose MEC servers for offloading based on their resources, bandwidth, and serving tasks. Simulation results reveal that the DTS scheme can improve the service ratio while lowering the response latency.