关键词: 端边云协同, 工作流调度, 能耗感知, 低时延

Abstract: 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.

Key words: end-edge-cloud collaborative computing, workflow scheduling, energy-aware, low latency