Research On Mobile Service Selection In Cloud And Edge Environment

Compared with traditional cloud computing,services provided by edge computing have several advantages such as high speed and low latency,which make edge services become the key technology of 5G.However,the number of edge servers,the computing capability of an edge server and the number of services deployed on an edge server are limited.Moreover,it is difficult for a single service to satisfy complex user requests.Therefore,researchers propose to combine edge computing with cloud computing and to aggregate services.Typically,as the number of services in both cloud computing and edge computing environments continues to increase and user mobility enhances,selecting the appropriate services to fulfill the complex needs of mobile users becomes a critical issue.How to select appropriate cloud and edge services with low response time and cost to meet complex needs of mobile users is a Non-deterministic Polynomial-Hard problem.For this reason,this dissertation focuses on the issue of selecting mobile services in the cloud-edge environment and its main research contents are as follows:(1)This study aims to research and implement a model for mobile service selection in cloud and edge environments.The model includes uploading requests,selecting services between cloud servers and edge servers,downloading results,and locating user positions using a suitable mobile path model.A new fitness function for service selection is proposed based on modeling user mobility and service combination patterns and calculating overall response time and cost with respect to the Quality of Service.(2)This study proposes a mobile service selection method that combines the Skyline and Firefly optimization algorithms in cloud and edge computing environments.The Skyline algorithm is used to pre-process the candidate services and filter out low-quality services.A Skyline-based Moth-Flame optimization algorithm is designed and implemented to select appropriate edge and cloud services to meet user requirements.(3)This study presents a method that integrates comprehensive the Quality of Service with an improved Moth-Flame optimization algorithm.Specifically,the order relation method is used to record the subjective weight of the user’s preferred service,the coefficient of variation method is used to calculate the objective the Quality of Service of the service,and the subjective and objective weights of different service attributes are combined to calculate the comprehensive the Quality of Service.Additionally,the Moth-Flame optimization algorithm is improved through differential evolution,and the best service set with the best comprehensive the Quality of Service is selected while satisfying functional requirements.In conclusion,this dissertation combines the optimization of service selection algorithms with cloud and edge computing,and improves user satisfaction by perceiving quality of service in the mobile state.Experimental results show that the proposed method improves algorithm query efficiency and reduces service usage costs,providing a reference for the practical application of Qo S-perceived service selection in cloud and edge computing.