Multitasking Multi-objective Scheduling Model Construction And Its Optimization Solution In Edge Environment

With the widespread popularity of mobile devices and the increasing number of mobile applications,the development of mobile edge computing has been driven.As a new mobile computing paradigm,mobile edge computing has the advantages of low latency,higher security,more reliability,and scalability.At the same time,it also has various challenges,especially service deployment issues and task offloading issues.Both task offloading and service deployment have been studied,but most of them are analyzed separately for these two independent problems.In contrast,in real life,task offloading and service deployment have a highly coupled relationship,which is crucial for MEC systems,and few current studies can consider both at the same time.Therefore,the main work of this paper is presented as follows:(1)For the two major challenges faced by edge environments,service deployment problem and task offloading problem,currently,the single problem of task offloading is solved in edge environments,and the service deployment problem is rarely considered at the same time.Since service deployment and task offloading are highly coupled,considering only one of these problems has limitations and can result in some wasted resources and large latency,which affects the user experience.Therefore,in order to solve these problems simultaneously,a multitasking multi-objective model is constructed,treating each optimization problem as a task,and some classical evolutionary algorithms are selected to solve the model to verify the effectiveness of the model.(2)To address the problem that the multitasking multi-objective model in the edge environment cannot satisfy both the quality of experience of users and the demand of service providers,this paper constructs a multitasking many-objective model,in which each task contains four objectives.Firstly,for task 1(service deployment problem),the service deployment problem is considered comprehensively from the perspective of quality of experience(latency,network reliability)and service providers(energy consumption,deployment cost).Secondly,for task 2(task offloading problem),the service provider takes the user’s offload cost as the measurement index in order to improve the user’s satisfaction,and considers offload delay,energy consumption,load balancing and offload cost.Finally,this paper also designs a multitasking many-objective algorithm based on multiple selection policies for this model and solves the model.A pool of environment selection policies is constructed,integrating a variety of excellent selection policies,so that the convergence and diversity of the solution set are guaranteed,and the effectiveness of the algorithm is verified by simulation experiments.(3)In this paper,a MATLAB-based multitasking scheduling system in an edge environment is designed based on the constructed model and algorithms.The system implements a user registration and login module,a user management module,a server management module,and a service deployment module and a task unloading module implemented based on a GUI interface.The above functions are implemented and the final results are presented through the running test of the system.