基于电力物联网的边缘计算任务卸载优化

为了解决传统卸载模型仅涉及用户设备和边缘计算资源,而在云端资源利用上存在局限性的问题,通过有效利用计算任务时延、能耗及计算资源配置,提出了基于深度强化学习算法的计算任务卸载策略和资源配置优化算法,建立了边云协同的时延、能耗及能效模型,研究了用户设备数量、任务量、任务优先级等对时延、能耗及能效的影响。结果表明：边缘计算服务器资源配置为30 GHz较为合理；高级计算任务优先处理策略和计算资源优化分配,使得时延、能耗均较低；本文所提出的优化算法在时延、能耗及能效方面均优于其他3个对比算法,表明针对不同用户设备数量和计算任务量场景,本文所提出的优化算法和建立的模型能够更有效的实现基于电力物联网的计算任务卸载策略和资源配置优化。

Offloading Optimization of the Edge Computing Task Based on the Power Internet of Things

In order to solve the problems of traditional offloading algorithm only involves user equipment and edge computing resources, and there are some limitations in the utilization of cloud resources. By effective use of computing task delay, energy consumption and computing resource allocation, the computing task offloading strategy and resource allocation optimization algorithm based on deep reinforcement learning algorithm was proposed, and the models of the edge cloud collaboration delay time, energy consumption and energy efficiency were established. The influence of the number of user equipment, task quantity, and task priority on the delay, energy consumption and energy efficiency was studied. The results show that the edge computing server resource is reasonably configured to 30 GHz. The advanced computing task priority processing strategy and computing resource optimization allocation result in low delay time and energy consumption. The proposed optimization algorithm is better than the other three comparison algorithms in terms of delay time, energy consumption and energy efficiency, and the optimization algorithm and the established model proposed in this paper can more effectively realize the computing task offloading strategy and resource allocation optimization for the power Internet of Things in scenarios of different users devices and calculation task volume.