Research On Low Power Resource Scheduling Strategy Based On Master-slave Cooperation In Edge Computing

The emergence of cloud computing has greatly promoted the growth of applications,but cloud computing also has unsolvable problems.When using cloud computing,all data of the application will be processed through a centralized cloud computing platform.With the development of the Internet,more and more real-time applications have emerged,and the requirements for low latency are becoming higher and higher.Cloud computing cannot solve the limitation of long-distance transmission,so that the latency remains high.The real-time requirements of the application encountered a certain bottleneck.The user’s request encountered great challenges.To solve this problem,edge computing is proposed.It sets up some servers on the near user side,and allows some user requests to be returned to the user side after processing through these servers to solve the delay problem.At present,there are still two major bottlenecks for edge computing in the industry,one is how to reduce the delay of processing tasks through resource scheduling,and the other is to consider the endurance problem of the terminal,because many edge device devices are provided by portable terminals such as smart phones.This paper analyzes the current work related to task division in the field of edge computing resource scheduling,and believes that it is feasible to divide the task into many subtasks and offload these subtasks to different servers to solve the problem of computing power consumption of user equipment.In order to solve this problem,this paper designs a resource scheduling algorithm based on branch and bound method.On this basis,this paper also designs an algorithm based on greed,which can quickly solve a feasible sub-optimal solution.In the subsequent simulation,it can be seen that the algorithm based on branch and bound fully utilizes the task’s tolerable delay and the edge-cloud node resources,which effectively reduces the power consumption of user equipment.The minimum difference between the suboptimal solution and the optimal solution based on the greedy algorithm can reach 1.8%.For the optimization of edge node devices,this paper proposes a sleep-based task scheduling algorithm to reduce edge node power consumption.The algorithm adds the sleep mode of the edge node to the scheduling.By splitting the edge node into multiple small devices,the proportion of sleeping devices is increased.On this basis,through resource scheduling,the characteristics of low power consumption of the device during sleep are brought into play.In addition,the scheduling process also guarantees the maximum tolerance of the task delay.Experiments show that the algorithm can effectively reduce the overall power consumption of the system.In the test of relatively idle tasks,the highest energy consumption is reduced by 27.9% compared with the original algorithm.