AVE: Autonomous Vehicular Edge Computing Framework with ACO-Based Scheduling
AVE: Autonomous Vehicular Edge Computing Framework with ACO-Based Scheduling
复制标题
DOI：
10.1109/tvt.2017.2714704
复制
发表时间：
2017-12-01
影响因子：
6.8
通讯作者：
Ji, Yusheng
中科院分区：
文献类型：
--
作者：
Feng, Jingyun;Liu, Zhi;Ji, Yusheng
随着车载应用的出现，提供所需的计算能力已成为一个关键问题。本文提出了一种用于道路边缘计算的自动车辆边缘（AVE）框架，旨在以分散的方式提高车辆的计算能力。通过管理车辆上的闲置计算资源并有效地利用这些资源，所提出的AVE框架可以在动态车辆环境中提供计算服务，而无需部署特定的基础设施。具体来说，本文介绍了一种支持车辆边缘自治组织的工作流程。为了更好地调度作业，提出了有效的作业缓存方法，该方法基于收集到的邻近车辆信息，包括GPS信息。设计了一种基于蚁群优化的调度算法来解决这一任务分配问题。进行了大量的仿真，仿真结果表明，在典型的城市和公路场景中，该方法优于竞争方案。
With the emergence of in-vehicle applications, providing the required computational capabilities is becoming a crucial problem. This paperproposes a framework named autonomous vehicular edge (AVE) for edge computing on the road, with the aim of increasing the computational capabilities of vehicles in a decentralized manner. By managing the idle computational resources on vehicles and using them efficiently, the proposed AVE framework can provide computation services in dynamic vehicular environments without requiring particular infrastructures to be deployed. Specifically, this paper introduces a workflow to support the autonomous organization of vehicular edges. Efficient job caching is proposed to better schedule jobs based on the information collected on neighboring vehicles, including GPS information. A scheduling algorithm based on ant colony optimization is designed to solve this job assignment problem. Extensive simulations are conducted, and the simulation results demonstrate the superiority of this approach over competing schemes in typical urban and highway scenarios.