Research on Task Scheduling Algorithm for Edge Computing in Low-voltage transformer the area network

Abstract

Aiming at the task scheduling strategy and performance optimization problems in the edge computing of the power Internet of Things, to meet the needs of refined and intelligent power distribution in the Low-voltage transformer the area, this paper studies how to effectively schedule applications to achieve task operation, processing time, and task delay And the trade-off between system performance. Taking into account the local advantages of the edge of the Internet of Things agents in the Low-voltage transformer the area(LV-TAN), as well as the characteristics of the edge intelligent computing tasks in the power distribution domain of the Low-voltage transformer the area, and the physical static characteristics of the Low-voltage transformer the area, it is established to minimize task running time as the goal, while considering the stability of the task queue, Balance, priority, and propose a scheduling strategy for edge computing based on container scheduling algorithm. This strategy decomposes the optimization problem into a series of sub-problems, and distributes tasks according to the current status of the queue backlog and unloading target nodes to meet the optimization goals of the distribution Low-voltage transformer the area and ensure the stability of the system. The simulation results show that the algorithm effectively reduces the task scheduling time, increases the CPU utilization rate, and reduces the jitter of task processing.