IIoT环境下基于聚类的工作流多雾协同调度算法  

Clustering-based multi-fog collaborative scheduling algorithm for workflows in IIoT environments

作　　者：吴宏伟 江凌云[1,2] 陈海峰 WU Hong-wei;JIANG Ling-yun;CHEN Hai-feng(School of Communication and Information Engineering,Nanjing University of Posts and Telecommunications,Nanjing 210003,China;Internet of Things Research Institute,Nanjing University of Posts and Telecommunications,Nanjing 210003,China;Cloud Computing Division,China Unicom Digital Technology Limited Company,Nanjing 210001,China)

机构地区：[1]南京邮电大学通信与信息工程学院,江苏南京210003 [2]南京邮电大学物联网研究院,江苏南京210003 [3]联通数字科技有限公司云计算事业部,江苏南京210001

出　　处：《计算机工程与设计》2025年第1期52-59,共8页Computer Engineering and Design

基　　金：江苏省重点研发基金项目(BE2020084-4)。

摘　　要：为解决在IIoT(industrial internet of things)环境下,现有的调度算法调度工作流中通信频繁、数据传输量大的任务所带来的完工时间上升、成本增加等影响的问题,提出一种基于聚类的工作流多雾协同调度算法。通过二分K均值算法对工作流中的任务进行聚类,基于聚类结果,在多个雾服务器之间使用改进的免疫粒子群优化算法进行任务调度。实验结果表明,该算法相比其它一些传统的调度算法在完工时间、成本、负载均衡方面都有一定提升。To solve the problem of rising completion time,increasing cost and other impacts of existing scheduling algorithms for scheduling tasks with frequent communication and large amount of data transmission in workflows in IIoT(industrial internet of things)environment,a clustering-based multi-fog collaborative scheduling algorithm for workflows was proposed.The tasks in the workflow were clustered using a bisecting K-means algorithm.Based on the clustering results,the tasks were scheduled among multiple fog servers using an improved immune-based particle swarm optimization algorithm.Experimental results show that the algorithm has some improvement in makespan,cost,and load balancing compared with some other traditional scheduling algorithms.

关 键 词：工业物联网 聚类 工作流 二分K均值算法 多雾 免疫粒子群优化算法 调度算法 

分 类 号：TP302[自动化与计算机技术—计算机系统结构]