基于车辆成组的V2V协同计算卸载策略研究  

Research on Offloading Strategy of V2V Collaborative Computing Based onVehicle Group

作　　者：吕昌达 李曦[1] 纪红[1] 张鹤立[1] LüChangda;Li Xi;Ji Hong;Zhang Heli(Beijing University of Posts and Telecommunications,Beijing 100876,China)

机构地区：[1]北京邮电大学,北京100876

出　　处：《邮电设计技术》2021年第1期36-42,共7页Designing Techniques of Posts and Telecommunications

基　　金：国家自然科学基金(61771070)。

摘　　要：在车联网中引入V2V计算卸载技术可以缓解当前车载计算卸载热点地区路边单元(RSU)计算资源不足的问题。然而,在计算卸载过程中,服务车辆可能因故障离组或自主选择离开车组。如何返回任务结果并高效地分配计算任务是需要进一步研究的关键问题。提出了一个车组内计算任务分配算法,考虑了可能导致车辆离开车组的因素影响,以及组中每辆车能提供不同计算资源的任务可分割性,以最小化总完成时延为目标,提出了一个任务分配问题并使用模拟退火算法进行求解。The introduction of V2V computation offloading technology in the vehicular networks can alleviate the problem of insufficient computing resources of roadside unit(RSU)in the hot spot.However,in the process of computation offloading,the service vehicle may leave the group due to breakdown or choose to leave the group independently.How to return task results and allocate computing tasks efficiently is a key problem to be further studied.A task allocation algorithm is proposed,which considers the factors that may cause vehicles to leave the group and the task partition that each vehicle in the group can provide different computing resources.In order to minimize the total completion latency,the task allocation problem is formulated and solved by simulated annealing algorithm.

关 键 词：车联网 移动边缘计算 车辆成组 车辆间通信 

分 类 号：TN929.5[电子电信—通信与信息系统]