Research On QoS-based Application Deployment Optimization In Fog Computing Environment

As smart devices become more popular in people’s daily lives and user experience requirements continue to improve,more and more complex applications are emerging in the field of big data and machine learning,which will be data-intensive,delay-sensitive and realtime.Traditional cloud computing architectures deploy cloud servers far away from generating application data,and some latency-demanding applications cannot tolerate the huge and unpredictable delays that cloud computing brings.As a new computing paradigm,fog computing provides a potential solution to the above challenges.As an extension of cloud computing,fog computing extends cloud services to the edge of the network by providing storage,computing and network resources at the fog nodes attached to terminal devices,so as to provide services to end users nearby.However,the fog nodes in the fog computing system are limited by their own computing capacity and storage capacity,and the number of application services that can be deployed is limited.Facing the complex and diverse user needs in the Internet of Things environment,providing quality services is the key to the recognition and promotion of fog computing.Therefore,based on the characteristics of distributed deployment of application services in fog computing,it is a key research issue in the field of fog computing to find a reasonable deployment scheme to improve user experience and Quality of Service(Qo S).In this paper,based on the architecture characteristics and key technologies of fog computing,Qo S problems in fog computing application deployment are studied and reasonable solutions are proposed.Firstly,the quantifiable Qo S index system of fog computing is studied by discussing the key issues of Qo S guarantee in fog computing.Then,respectively from the independent node model and clustering model to conduct the thorough research to the application deployment problem in system: independent node model,according to the diversity of user application service requests,respectively,to optimize the service model is established with target of delay and energy consumption of resources,design the application deployment algorithm based on greedy strategy,the performance of the online algorithm is analyzed theoretically;In the cluster mode,according to the complex and heterogeneous characteristics of fog node resources in the cluster,the application deployment optimization problem is proposed from the perspective of improving resource utilization rate.To solve this problem,an effective heuristic algorithm is proposed,which improves the utilization rate of fog resource and guarantees the quality of service for users.Finally,through i Fog Sim fog computing simulation platform,the fog computing application deployment optimization scheme in independent node and cluster mode in this paper are verified and compared,which proves the feasibility and effectiveness of the algorithm proposed in this paper.By comparing the proposed scheme with other relevant schemes,the performance advantages of the proposed scheme are illustrated.The optimization of operating cost and resource utilization while guaranteeing Qo S is realized,which provides an effective reference scheme for application deployment research in fog computing.