边缘协作环境下最小化完工时间任务调度方法

Task scheduling method for minimizing completion time in edge collaborative environment

由于用户地理位置分布不均可能导致边缘服务器负载不均衡,难以为用户提供满意的服务质量。此外,边缘服务器可用资源有限,一些大任务可能难以全部卸载到边缘服务器。针对以上问题,利用多个边缘服务器之间的协作,结合任务部分卸载方式,提出一种边缘协作环境下最小化完工时间的任务调度方法。首先,结合边缘水平协作和任务部分卸载技术,考虑多用户多边缘服务器场景下用户和边缘服务器的位置关系,以最小化任务完工时间为目标,建立任务部分卸载调度模型;其次,提出基于改进分组教学优化算法的任务调度算法,联合优化边缘服务器计算资源分配、用户-边缘服务器关联决策、任务卸载比例以及执行位置决策,以最小化任务完工时间为目标,实现边缘计算环境下任务的高效调度;最后,通过实验将提出的任务调度算法与其他算法在多个指标下进行对比。实验结果表明,所提方法能够有效降低任务完工时间。

The uneven geographical distribution of users may lead to unbalanced load on edge servers,which makes it difficult to provide satisfactory service quality for users.In addition,the available resources of the edge server are limited,and some large tasks may be difficult to offload to the edge server.To solve the above problems,this paper proposes a task scheduling method to minimize the completion time in the edge collaboration environment by utilizing the collaboration among multiple edge servers and combining the task partial offloading technology.First,by combining the edge of horizontal collaboration and task partial offloading technology and considering the position relationship between users and edge servers in multi-user multi-edge server scenario,a task partial offloading and scheduling model is established to minimize the task completion time.Second,a task scheduling algorithm based on the Improved Group Teaching Optimization Algorithm(IGTOA)is proposed to jointly optimize the edge server computing resource allocation,user-edge server association decision,task offloading ratio and execution location decision.With minimizing the task completion time as the goal,efficient task scheduling is achieved under edge computing environment.Finally,the proposed task scheduling algorithm is compared with DTOSO,HJTORA and ACS algorithms under multiple indexes.Experimental results show that the proposed method can effectively reduce the task completion time.