Resource Allocation And Task Scheduling For Multi-objective Optimization In Edge Computing

With the rapid development of mobile Internet,Internet of things and other technical fields,cloud computing technology has also made great progress.More and more applications are deployed on cloud servers to provide services for low-cost terminal devices.In the intelligent era of "everything connected",more and more terminal devices will access to the Internet,and cloud computing network is under increasing pressure.According to Cisco's forecast,there will be more than 50 billion devices connected to the network in 2020,which means that a large amount of data will be sent to the cloud computing center through the Internet.In the face of massive data processing,cloud computing is under great pressure.The development of Internet of things technology is facing a bottleneck,so it is urgent to provide a low latency computing service.As the supplement of cloud computing technology,edge computing technology provides computing services by deploying devices with computing power on the edge of the network near the terminal device,which makes up for the deficiency of cloud computing.The concept of edge computing is to extend cloud computing to network edge processing,so that the data obtained from edge terminals can be calculated near the user terminal,at the same time,the data that does not need to be stored for a long time does not need to be transmitted to the cloud service center for backup,which relatively reduces the occupation of network bandwidth,reduces the power consumption and load of the cloud.By analyzing the architecture and current situation of edge computing network,this paper finds out the challenge of the development of edge computing technology: resource allocation and task scheduling in edge network.In the cloud computing technology,the computing task is concentrated on the cloud server,and all nodes in the network only have the function of forwarding data.Edge computing technology gives computing power to gateway,router and other nodes in the network,expecting to provide computing services through these nodes.However,how to allocate a large number of computing resources and how to schedule computing tasks in the edge network has become a difficult problem of edge computing technology.In this paper,the problems in resource allocation and task scheduling of edge computing are studied as follows:First,we analyzed the characteristics of Internet of things applications and the distribution and connection of servers in edge computing network.The delay model of application program and the power consumption model of system in edge computing network are proposed,and the multi-objective optimization model is proposed for the two models to analyze and evaluate the performance of edge computing system.Secondly,an edge computing dynamic task scheduling strategy with delay awareness is proposed for the edge cloud collaboration computing model.This strategy can make full use of the resources in the edge computing system by dynamically creating virtual machines in the edge computing network,and reduce the use of the network on the basis of ensuring the low latency of the system.Thirdly,in order to meet the needs of low latency and low power consumption in edge computing system,a resource allocation scheme sap,which combines simulated annealing algorithm,is proposed.This scheme takes low delay and low power consumption as the optimization goal.With the advantages of simulated annealing algorithm,it can search the global optimal solution quickly and find the approximate global optimal solution,so as to reduce the overall energy consumption of the system to the maximum extent while maintaining the low delay of the system.Finally,in order to verify the effectiveness of the strategy proposed in this paper,the i Fog Sim simulator is used to simulate some scenes.The experimental results show that compared with the traditional static task scheduling strategy,the dynamic task scheduling strategy proposed in this paper can significantly reduce the response delay of the application program,which can reduce 33.07% at most,and 66.22% at most.Compared with the traditional resource allocation strategy,SAP algorithm also greatly reduces the application delay time while maintaining low energy consumption.