端边云协同环境下能耗感知的工作流实时调度策略

Energy-aware workflow real-time scheduling strategy for device-edge-cloud collaborative computing

针对端边云协同环境下工作流应用场景中,智能终端可移动、边缘服务器服务范围有限、用户实时性要求高和终端能耗等问题,建立了基于端边云异构资源有效协同的工作流任务执行时间模型和终端能耗模型,在此基础上构建端边云环境下能耗感知的工作流实时调度模型,并提出能耗感知的工作流任务调度算法。该算法首先根据工作流特性划分子任务优先级;其次根据终端初始位置信息,利用改进粒子群优化算法找到一个最优的资源调度方案;然后根据终端移动轨迹筛选可迁移资源,并为每个任务动态选择最优迁移决策。仿真结果表明,与已有策略相比,新策略能够在满足时间延时的约束下降低终端能耗,获得最优系统适应度值。

In view of the challenges such as mobility of smart terminal, limited service scope of edge servers, high real-time requirements of users and terminal energy consumption in end-edge-cloud workflow application scenarios, a workflow task execution time model and a terminal energy consumption model based on the effective collaboration of heterogeneous resources were established. On this basis, an energy-aware workflow real-time scheduling model in the end-edge-cloud system was constructed, and a real-time scheduling algorithm named Energy-Aware Workflow Scheduling Algorithm(EAWSA) was proposed. The algorithm prioritizes subtasks according to the workflow characteristics. According to the terminal initial location information, an optimal resource scheduling scheme was found by using the improved particle swarm optimization algorithm. Then, according to the terminal moving trajectory, the migration resources were deleted and selected, and the optimal migration decision was dynamically selected for each task. The simulation results showed that the new strategy could reduce the terminal energy consumption and obtain the optimal system fitness value under the constraints of time delay by comparing with the existing strategies.