Research On Task Scheduling Algorithm In Edge IoT Agent

The task scheduling problem of the edge IoT agent in the power Internet of Things is actually the resource allocation problem between the computing load demand and the supply of computing resources.The allocation process needs to consider the comprehensive needs of the actual tasks of the power Internet of Things,the resource characteristics of the edge IoT agent node,and then use A reasonable allocation strategy can obtain the optimal resource allocation scheme,thereby maximizing the value of a task scheduling scheme.The research on the computing load model of the power Internet of Things mainly considers the single operation resource demand of the container,while the research in the scheduling strategy mainly focuses on the balance,delay,communication cost,etc.,ignoring the relationship between the time-series logic of the aggregated service load and the mathematical description method,which is easy to This causes the resource fluctuation of the scheduled container in some time periods and the long delay of the key microservice container,which is not conducive to the smooth operation of the entire IoT edge side.Based on the above problems,this paper analyzes the description method of power microservice time series computing load characteristics,constructs a comprehensive time series scoring strategy,and considers the replica constraints of migration to improve the solution strategy of container migration.The specific research content of the paper is as follows:First,a quantitative strategy of a common container cloud scheduling method is proposed,and its scheduling effect is verified.The simulation results show that the proposed algorithm can effectively reduce the delay,resource fragmentation rate and improve balance compared with moving average and random strategies.Secondly,a matrix description method based on microservice temporal logic is proposed,and the temporal logic relationship is obtained through temporal matrix transformation.On the basis of considering the evaluation attributes such as affinity,connection relationship,and importance of microservices in the calculation load sequence,a scheduling scoring strategy that conforms to the characteristics of power microservices is designed.Taking the new power distribution station area control aggregation service as an example,and conducting experiments through the Edgcloudsim simulation platform,the simulation results show that considering the load sequence logic can effectively reduce the volatility of node resources and ensure the running time of key containers for power edge computing.Finally,in view of the problems of excessive resource usage deviation,resource fragmentation,and migrating copies,a container migration scheme is proposed that considers the transient characteristics of containers,computational load timing logic constraints,and balanced costs.It is faster to improve the worst reduction strategy.To solve the container migration problem in this paper,and to avoid intensive scheduling with high weights,decentralized scheduling operations for microservices with higher weights are more in line with the actual operation scheme.