Research On Task Scheduling Optimization Of Power Distribution LoT Terminals Under Cloud-edge Collaboration

With the construction of the Internet of Things for power distribution,the number of terminal devices and tasks has increased rapidly.For emerging business increments,edge devices use microservice architecture and container technology to realize the processing of different services by one physical device.As new services continue to expand,device computing,storage,and communication resources may be insufficient.At the same time,the task may fail to compute and transmit when the device resources are limited.The use of the independent characteristics of microservices,according to the differentiated needs of application services,carry out task scheduling and resource adjustment,which is of great significance for making full use of the advantages of the cloud-edge and improving the implementation efficiency of power tasks.However,most of the existing research on power task scheduling does not consider the topology and differentiation characteristics of tasks,as well as the impact of changes in equipment and container resources and service deployment on scheduling decisions.To this end,this paper conducts research on the problems existing in the existing work:Task processing delay is the most critical indicator to decide whether a task can be completed or not.In order to satisfy more task requests,a task cloud-edge collaborative scheduling optimization algorithm based on greedy policy is proposed.First,the task scheduling problem is analyzed from the task perspective and the device perspective,and multiple dependent microservices are modeled by directed acyclic graph(DAG).The processing of microservices in containers is subdivided and the corresponding mathematical model is established considering the limited computing resources of the edge devices.Then the individual microservices are ranked by microservice priority to form the scheduling sequence of tasks.The priority of power tasks is calculated by fuzzy logic algorithm,so as to distinguish the importance of different tasks.Finally,the different tasks are scheduled according to whether the system resources are constrained or not.The simulation results show that the proposed strategy has higher task execution efficiency.The execution of tasks generates energy consumption,and different tasks have different emphasis on latency and energy consumption.In order to balance task latency and energy consumption,a task cloud-edge collaborative scheduling optimization algorithm based on NSGA-II is proposed.In order to get closer to the real scene,consider the restrictions on the computing and storage resources of the edge equipment,and establish a microservice processing model based on the service configuration mechanism.By analyzing the constraints in the task scheduling,priority constraints,ability constraints and execution time constraints are pointed out.The decision-making problem of delay and energy consumption is transformed into a multi-target optimization problem with constraints,and the NSGA-II algorithm is solved,and then the scheduling scheme is selected through the multi-standard decision method based on fuzzy logic.The simulation results show that the strategy mentioned effectively reduces the delay of task processing and system energy consumption.As requirements and services change,task scheduling is taking on an increasingly important role,as a key supporting technology of edge computing,it will further promote the development of edge computing in the distribution Internet of Things.