Research On Task Offloading Strategy Technology In Mobile Edge Computing Environment

With the continuous increase of 5G base stations and the increasing computing performance of mobile terminal equipment,the types and numbers of tasks that can be completed under the mobile Internet network are increasing,the total amount of data that needs to be processed and transmitted is also increasing,and the transmission rate is also increasing.Growing year by year,these advancements have brought a better app experience to users.In the future,the evolution from 2D images and videos to 3D data applications will be an irresistible trend.Under this premise,the level and capability of mobile edge computing will continue to improve to meet the everincreasing demands of the mobile Internet.At present,the existing mobile edge computing research mainly focuses on two aspects.On the one hand,it is the research on the task offloading strategy of mobile terminals,focusing on improving the comprehensive indicators of task delay,energy consumption or delay + energy consumption of mobile terminals;most of the existing researches ignore the topological structure of mobile terminal task flow in mobile terminals and mobile terminals.Distribution in edge servers,and rigid requirements for user privacy protection.On the other hand,it is the research on the resource allocation strategy of edge computing servers,focusing on improving the utilization of bandwidth resources,storage resources,computing resources and task delay indicators.Existing research is to make offloading decisions under the premise of a given number of task streams,a fixed number of servers,and no consideration of server resource competition and network congestion.In reality,the number of task flows changes in real time,the number of servers can be adjusted,and server resource exhaustion and network congestion occur from time to time.In response to the above two problems,this paper carried out the following research:(1)Research on mobile edge computing offloading strategy considering differences in task flow topology and user privacy protection requirements.In this application scenario,a task flow generated by a single user is interconnected with a single server.When formulating a computing offloading strategy,issues such as task flow multi-path selection,limited bandwidth resources,user privacy,and user delay sensitivity need to be considered.Focus on the difference of task flow topology and user privacy protection requirements,and take other conditions as auxiliary constraints.Therefore,we formulate workflow offloading as a constrained multiobjective optimization problem,with the constraints of task flow multi-path selection,limited bandwidth resources,user privacy,limited server computing resources,and user delay sensitivity.The Artificial Fish Swarm Algorithm(GAFSA)to solve the optimal policy for energy saving in application task offloading of workflow.Experimental results show that our strategy outperforms related methods applied to similar problems.(2)Research on the offloading strategy of mobile edge computing considering the changes in the number of task flows and server resources.In this scenario,multiple users generate task flow and multiple servers are interconnected.This is not a simple increase in the number of tasks.The multi-user task flow will increase the problem of task flow heterogeneity and random high concurrency.In addition,highconcurrency environments also need to consider situations such as variable number of servers,limited computing resources,and possible server congestion.When formulating a computing offloading strategy,it is necessary to consider the real-time change of the number of task flows,the adjustable number of servers,the exhaustion of server resources,and the irregular occurrence of network congestion.Therefore,considering the characteristics of task flow(heterogeneous task flow,random high concurrency)and server characteristics(limited computing resources,different congestion conditions,and variable number of servers participating in the calculation),this paper proposes a dynamic server computing offloading strategy(DSCOS).The algorithm involved in this strategy is based on and improved on genetic algorithm,which aims to minimize the weighted sum of delay and energy consumption for all tasks to be executed in a multi-user multi-server system,so that computationally intensive and delay-sensitive applications can work normally.run.Finally,the effectiveness of the algorithm in this paper is verified by simulation experiments and comparative experiments.