面向车联网低时延需求的分布式路径计算方案  

Distributed Path Computing Scheme for Low-Latency Demand of Internet of Vehicles

作　　者：李瑞彪 任继军 任智源[2] LI Ruibiao;REN Jijun;REN Zhiyuan(School of Communications and Information Engineering,Xi’an University of Posts and Telecommunications,Xi’an 710121,China;State Key Laboratory of Integrated Services Networks,Xidian University,Xi’an 710071,China)

机构地区：[1]西安邮电大学通信与信息工程学院,西安710121 [2]西安电子科技大学综合业务网理论及关键技术国家重点实验室,西安710071

出　　处：《西安交通大学学报》2022年第10期91-100,共10页Journal of Xi'an Jiaotong University

基　　金：陕西省重点研发计划资助项目(2021GY-100)。

摘　　要：针对传统云服务应用于车联网大数据处理时存在通信时间消耗大、服务质量无法保证的问题,提出一种低时延的分布式路径计算方案。利用云雾网络架构将数据业务就近上传到雾网络设备以及云服务平台;为了让雾网络设备协同处理车载业务,采用一种基于该架构的任务映射策略(也称为业务卸载策略),将以有向无环图表示的车载业务流模型映射到以无向图表示的雾网络拓扑中;为了解决映射中存在的时延波动问题,采用一种改进的离散二值粒子群优化(NBPSO)算法进行优化;将该算法与负载均衡技术相结合得到最优映射方案。仿真结果表明:所提出的分布式路径协同计算方案可有效解决云计算模式下数据处理的高延时问题;当数据量为12Mb时,采用所提方案的时延比采用云计算的时延降低约2.95s。In this paper,a low-latency distributed path computing scheme is proposed to solve the problems of long communication time and unstable services during big data processing for the internet of vehicles using traditional cloud service.First,a cloud-fog network architecture is adopted to upload data tasks nearby to the fog network devices and the cloud service platform;second,a task mapping strategy based on this architecture,also known as task offloading strategy,is employed to map flow models of vehicle tasks represented by directed acyclic graph(DAG)to the fog network topology represented by undirected graph(UG),which enables fog network devices to collaboratively process in-vehicle tasks;then,a modified discrete binary particle swarm optimization(NBPSO)algorithm is employed to optimize the mapping in order to tackle delay fluctuation in the mapping;finally,a load balancing technique is used in combination with this algorithm to obtain the optimal mapping scheme.According to the simulation results,the distributed collaborative computing scheme proposed in this paper can effectively tackle the delay of data processing in the cloud computing mode;the latency under this scheme is reduced by about 2.95 s compared with cloud computing when the data size is 12 Mb.

关 键 词：车联网 云雾网络 路径计算 任务映射 

分 类 号：TN925.1[电子电信—通信与信息系统]