Research On Microservice Resource Scheduling Of Edge Cloud Platform Under Internet Of Vehicles System

With the rapid deployment of 5G networks,intelligent driving technology is ushering in new development opportunities.The intelligentization and networking of automobiles can provide safer,more energy-efficient,more efficient and more convenient personal travel comprehensive solutions,which is the development direction of driving technology in the future.Due to its traffic attributes,intelligent driving vehicles have high requirements for QoS(quality of service),including communication delay,fast computation and high bandwidth.In order to meet the QoS requirements of intelligent driving vehicles,computing resources should be deployed to the edge of the network.However,compared with the traditional cloud computing center,the resources of edge cloud computing are less,which cannot provide the reliability advantages brought by large-scale data centers.In edge cloud computing,application deployment using container-based microservice architecture can maximize the utilization of hardware resources and achieve high reliability,flexibility and performance under limited resource conditions.Therefore,it is very important to design the resource scheduling strategy of microservice reasonably to satisfy the service quality requirements of intelligent driving vehicles and to satisfy the sufficient edge computing resources.Based on the above-mentioned topics,this paper researches the problem of microservice resource scheduling on the edge cloud platform of the IoV system.The specific research contents are as follows:1.This paper first analyzes the overall architecture of the IoV system and the layered deployment architecture of microservices in the IoV system,and explains the importance and significance of studying the application service deployment problem on the edge platform.After that,the analysis is conducted of the micro-service resource scheduling problem of the IoV edge cloud platform.2.This paper establishes a multi-objective optimization model for microservice resource scheduling by analyzing the resource constraints and characteristics of microservice in cloud platform deployment.In this model,the comprehensive energy objectives include the shortest distance of microservice invocation,the highest utilization rate of physical machine cluster resources,and the best guarantee of load balance.3.This paper designs a new Multiple Fitness Genetic Algorithm(MFGA)to solve the multi-objective optimization model of microservice resource scheduling.MFGA integrates a gene evaluation function and a container dynamic migration strategy into traditional genetic algorithm to make it more suitable for solving micro-service resource scheduling problems.4.The container cloud evaluation environment ContainerCloudSim in CloudSim is used to simulate the microservice resource scheduling optimization model and algorithm designed in this paper.The simulation results show that the MFGA can not only improve the resource utilization rate of the physical machine cluster and ensure a certain load balance,but also effectively reduce the cross-physical host calls of microservices and obtain better overall performance.