Research On Resource Management Algorithm Based On Deep Q-learning In Cloud Computing Environment

Task scheduling,which plays an important role in cloud computing,is a critical factor that determines the performance of cloud computing,which is defined as an NP-hard problem.From the explosion of information processing to the increasing needs of Quality of Service(QoS)in the internet,dynamic task scheduling problem has attracted more and more attention.Additionally,the dynamic online task scheduling usually handles tasks in a complex environment,which makes it even more difficult to balance benefits from each and every aspect of cloud computing.In addition,the surge in data processing has led to the rapid expansion of computing resources.Larger scale clusters present a heavy burden on cloud service providers and the environment.High energy consumption decreases the economic benefits of cloud service providers,while great power demand puts pressure on the environment.Dynamic consolidation of virtual machines(VMs)using live migration technology to improve resource utilization and reduce energy consumption is an effective way to save energy consumption while ensuring high performance with the desired level of quality of service(QoS)between cloud providers and users.In view of the two difficult problems faced by cloud computing,this paper uses a new machine learning method,called deep Q-learning(DQL)to try to solve the following two problems.This paper proposed a novel artificial intelligent algorithm called deep Q-learning task scheduling(DQTS).This new method is aimed to solve the problem of handling directed acyclic graph(DAG)tasks in a cloud computing environment.Based on WorkflowSim,experiments are conducted comparatively with the variance of Makespan and load balance in task scheduling.Both simulation and real-life experiments are conducted to verify the optimization and learning abilities of DQTS.The experiment shows that compared with several classic algorithms pre-coded in WorkflowSim,DQTS has advantages in terms of learning ability,containment and expansibility.Our research in this paper has successfully provided a new method for task scheduling in cloud computing.Besides,this paper proposed a novel machine learning algorithm called deep Q-learning based VM consolidation(DQLVMC) that combines the Q-leaning approach and deep learning to find an approximate optimality solution.Furthermore,based on the real workload trace in the cloud,the experiment shows that DQLVMC effectively reduces energy consumption while meeting the high performance of QoS requirements.