Research On QoS-Aware Workflow Scheduling Algorithm In Mobile Edge Computing

In recent years,with the rapid development of Mobile Edge Computing(MEC),the workflow scheduling problem in mobile edge environment has gradually become a new hotspot.How to schedule the tasks of workflow applications are directly related to the Quality of Service(QoS)of mobile applications.The current research on workflow scheduling in mobile edge environment mainly focuses on the optimization of application execution time,energy consumption and other single goals,which cannot meet the diversified scheduling requirements.At the same time,the computing resources in the MEC environment are prone to problems,such as task anomalies and software failures.How to ensure the workflow scheduling under the condition of resource failure is very necessary.Aiming at the above problems,this paper proposes a quality-oriented optimal multi-objective genetic algorithm and a fault-tolerant scheduling algorithm for edge computing resources.Finally,the effectiveness of the proposed algorithm is verified by experiments.The innovations and main work of this paper are reflected in the following aspects:Firstly,aiming at the QoS scheduling optimization problem of workflow tasks in mobile edge environment,a quality-oriented Optimal Multi-Objective Genetic Algorithm(OMOGA)was proposed.Firstly,the algorithm considers the cost of dependent data transmission between workflow tasks,the task preprocessing of user-submitted workflow was carried out.The task with high data dependent cost was scheduled to the same device,which can effectively reduce the energy consumption of the system.Secondly,the algorithm also combines the optimized heuristic multi-objective genetic algorithm,reasonable genetic coding and genetic operation for workflow tasks.A scientific fitness evaluation function is established,which improves the optimization ability of workflow scheduling scheme and effectively improves the QoS of users.Secondly,a reliability-oriented Fault-tolerant Space Time Genetic Algorithm(FT-STGA)for resource failure in mobile edge environments is proposed.Firstly,combined with the characteristics of computing resources in the edge environment,the resource fault and task execution time model are established.Then,the task replication policy is used to copy the multi-version tasks to different computing resource nodes.When the resource nodes of the main version task are invalid,the replica tasks on other resource nodes can still be executed.In addition,this method also takes into account the execution time delay caused by task failure.Combined with optimized spatio-temporal genetic algorithm,on the premise of ensuring the reliability of workflow computing resources,the completion time of workflow should be reduced as much as possible.Thirdly,through the workflow simulation software WorkflowSim experimental platform,the OMOGA and FT-STGA are implemented.The effectiveness of OMOGA algorithm in task execution time,workflow reliability and energy loss were verified by setting up scientific evaluation indexes through multiple experimental analyses.Through comparative experiments,it is found that for the uniformity of multi-objective solution sets,OMOGA algorithm proposed in this paper is 5.8%,12.6%higher than NSGAII algorithm and CMOHEFT algorithm,respectively.Under the premise of ensuring resource tolerance,the FT-STGA algorithm has improved the success rate of task execution by 15.6%and 21.8%compared with the traditional algorithms CCRH and IRW.The research on QoS-aware workflow scheduling under mobile edge environment proposed in this paper can further meet the diversified needs of workflow scheduling,to improve the service quality of mobile edge computing and enrich the workflow scheduling theory and method system under the mobile edge environment.And it has a good application scenario.