Task Scheduling Algorithm And System Implementation In Dynamic Edge-Cloud Hybrid Environment

The purpose of this paper is to explore the edge-cloud hybrid task scheduling algorithm,which is a key technology of edge-cloud hybrid computing.With the continuous development of cloud computing and edge computing technologies,edge-cloud hybrid computing has received wide attention as a new computing model.Edge computing provides computing resources and services closer to users,while cloud computing can provide more efficient and larger scale computing resources.Edge-cloud hybrid computing integrates the advantages of both and provides more efficient and flexible computing services through rational utilization of both resources.Task scheduling algorithm,as the core technology to realize edge-cloud hybrid computing,can reasonably allocate tasks to edge nodes and cloud nodes,optimize resource utilization,improve computing efficiency,and reduce energy consumption,which has important practical significance in typical cloud-edge scenarios such as intelligent medical care and intelligent transportation.Therefore,the research of edge-cloud hybrid task scheduling algorithm is crucial for the development and application of edge-cloud hybrid computing.In this paper,we construct a pervasive edge cloud model with typical scenarios and propose a task scheduling algorithm that supports dynamic weight adjustment.The algorithm enables users to make trade-offs for costs such as latency and energy consumption according to different scheduling tasks,while taking load balancing into consideration,and describes uncertainties in edge cloud environments,providing a generalized solution to the task scheduling problem of cloud-edge hybrid scenarios.The main research components include.1.Modeling and analysis in conjunction with a dynamic cloud-edge hybrid scenario represented by a medical system.The dynamic edge-cloud hybrid scenario represented by a medical system can efficiently carry different task demands in an intelligent and digital medical system by leveraging the respective advantages of the edge and central clouds,and performs well in improving diagnostic efficiency,alleviating medical resource inequality,and promoting the development of smart healthcare.This paper describes the relationship between healthcare system and cloudside hybrid scenario by constructing system,task and cost models,and provides practical scenario support for solving dynamic cloud-side hybrid task scheduling algorithms.2.Proposed particle swarm optimization-based scheduling algorithm for dynamic edge-cloud hybrid environment.The particle premature problem existing in the standard particle swarm algorithm is discussed and the optimization design of the algorithm is carried out.The particle iteration is changed by updating the inertia weight coefficients and performing subpopulation division,and the diversity and convergence of the algorithm are optimized by defining the learning strategy,rejection strategy and variation strategy.The better performance of the proposed algorithm in the same scenario is verified by comparison experiments.3.Based on the algorithm research,this paper designed and implemented a task scheduling system that can schedule tasks in Kubernetes.The system is divided into functional modules according to the use case,including environment configuration,algorithm training,task scheduling,and real-time monitoring,and supports dynamic resource weights and scheduling result visualization.The system follows the software engineering design process,has good scalability,and has a simple and convenient interface interaction.It has been verified by all test cases.