云边端协同下的任务调度与资源分配方法

Task scheduling and resource allocation methods under cloud edge collaboration

为了优化多用户终端、多边缘节点以及云服务器之间的计算资源分配,找到最优的任务调度和资源分配方案。本文在算网融合的基础上提出了一种3层计算资源分配方案,考虑到通信网络带宽的有限性,设立了任务计算优先级和任务上传策略,并在多边缘节点的环境下,制定了多节点并行计算规则,以最大化计算资源的利用率,通过深度强化学习算法进行训练,以学习最优的任务调度和资源分配策略。实验结果表明,本文所提方法在缩短计算任务完成时间方面表现出色,并且在任务数据量增长的情况下,依然表现出良好的鲁棒性。

To optimize the allocation of computing resources between multi-user terminals,multiple edge nodes,and cloud servers,and find the optimal task scheduling and resource allocation scheme.This paper proposes a three-layer computing resource allocation scheme based on the integration of computing and network.Considering the limited bandwidth of the communication network,task calculation priorities and task upload strategies were established,and multi node parallel computing rules were formulated in an environment with multiple edge nodes to maximize the utilization of computing resources.Deep reinforcement learning algorithms were trained to learn the optimal task scheduling and resource allocation strategies.The experimental results show that the method proposed in this article performs well in shortening the completion time of computational tasks,and still exhibits good robustness even with an increase in task data volume.