A DQN-based workflow task assignment approach in cloud-fog cooperative considering terminal mobility
Dongyue Huo, Hailiang Wu, Baiyang Wang, Yuyun Kang, Dongping Chen · 2022
When the terminal device moves under the cloud-fog cooperative, reasonable task assignment between fog nodes and cloud servers is a difficult problem and uncompleted tasks migration to maintain the continuity of tasks is another difficult problem. To solve these two problems, a workflow task assignment and migration decision algorithm based on Deep Q Network(DQN) is proposed. Firstly, a workflow task assignment optimization model for the minimized latency is constructed. Secondly, a DQN-based algorithm is used to assign tasks reasonably. When the terminal device is moving, a task migration decision program is activated to execute task migration. Finally, a simulation experiment is proposed to verify the approach and the results indicated that the approach reduced the time latency reasonably and effectively by allocating the computing resources.