Research On Optimization Method Of Service Operation Based On Edge-cloud Collaboration

In recent years,the Internet of Things technology has developed rapidly,and the number of smart terminal devices has continued to increase.The large amount of data generated at the edge of the network makes traditional cloud computing centers unable to efficiently process massive streaming Io T data,and it is difficult to guarantee the real-time performance of data processing by applications..In order to cope with the above-mentioned challenges,edge computing and edge-cloud collaborative data processing technologies have emerged.For this reason,based on the edge-cloud collaboration technology,this article has launched a research on the optimization method of streaming data processing services in the Internet of Things environment.The main work of the thesis includes:For stream data processing based on the edge-cloud collaboration framework,there are currently two methods as follows:1.Aiming at how to reduce the amount of data transmitted by the edge cloud,a task offloading method based on the FF(Ford-Fulkerson)algorithm under constrained conditions is proposed.This method abstracts the stream data processing service into a formal representation of a "flow graph",establishes a resource model of the stream data processing service under resource constraints,and cuts the flow graph through the FF minimum cut algorithm.Since the cutting result may have multiple cutting methods,this article combines edge cloud resources to choose a cutting method,and then extracts the edge execution task sub-graph and the cloud task sub-graph separately,reducing the task sub-graphs executed from the edge to the cloud The data transmission volume of the task subgraph reduces the service delay.2.Aiming at the problem of how to optimize task scheduling,a DACBPIO scheduling optimization algorithm based on Pigeon-inspired Optimization(PIO)is proposed.By dividing and conquering the DAG task scheduling problem,it avoids the dependency on the task allocation.The influence of the algorithm.By improving the resource utilization rate of a single divided task,the overall resource utilization rate is optimized,and the service operation speed of edge-cloud collaboration is improved.3.Evaluate,analyze and verify through experiments and prototype systems.The results of simulation experiments proved the effectiveness of the above method to a certain extent,and a prototype system for stream data processing service operation optimization based on edge-cloud collaboration was designed and implemented.