Multihop cost awareness task migration with networking load balance technology for vehicular edge computing

面向车载边缘计算的多跳成本感知任务迁移与网络负载均衡技术

阅读:1

Abstract

6G technology aims to revolutionize the mobile communication industry by revamping the role of vehicular wireless connections. Its network architecture will evolve towards multi-access edge computing (MEC) distributing cloud applications to support inter-vehicle applications such as cooperative driving. As the number of tasks offloaded to MEC servers increases, local MEC servers associated with vehicles may encounter insufficient computing resources for task offloading. This issue can be mitigated if neighboring servers can collaboratively provide computing capabilities to the local server for task migration. This paper investigates dynamic resource allocation and task migration mechanisms for cooperative vehicular edge computing (VEC) servers to expand computing capabilities of local server. Then, the multihop cost awareness task migration (MCATM) mechanism is proposed in this paper, which ensures that tasks can be migrated to the most suitable VEC server when the local server is overloaded. The MCATM mechanism begins by addressing whether the nearest VEC server can handle the computational tasks. We subsequently address the issue of duplicate selection to choose an appropriate VEC server for task migration among n-hop neighboring servers. Next, we focus on finding efficient transmission paths between the local and destination VEC servers to facilitate seamless task migration. The MCATM includes (i) the weight variable analytic hierarchy process (WVAHP) to select a suitable server among multihop cooperative VEC servers for task migration, and (ii) the pre-allocation with cost balance (PACB) path selection algorithm. The simulation results demonstrate that the MCATM enables the migration of computational tasks to appropriate neighboring VEC servers with the aim of increasing the task migration success rate while balancing network traffic and computing server capabilities.

特别声明

1、本页面内容包含部分的内容是基于公开信息的合理引用;引用内容仅为补充信息,不代表本站立场。

2、若认为本页面引用内容涉及侵权,请及时与本站联系,我们将第一时间处理。

3、其他媒体/个人如需使用本页面原创内容,需注明“来源:[生知库]”并获得授权;使用引用内容的,需自行联系原作者获得许可。

4、投稿及合作请联系:info@biocloudy.com。