Cloud Computing Resource Schedule Policy Based On Ant Colony Algorithm And Time Series Prediction Model

Cloud computing is a promising new commercial computing model.It realizes that computing capacity,storage and information services carl be easily used on demand and be charged flexibly like water,electric power,and gas.By using virtualization technology,Cloud computing makes physical resource pooling and this Can simplify system management,optimize resource utilization and reduce energy consumption.Because of its huge benefits and great prospects,Cloud computing gets the attention of many companies and becomes the hot issues of academic research.Task scheduling problem is the keystone to Cloud computing.Cloud computing system faces with different demands from many users.So task scheduling policy concerns system efficiency and users quality of service.As a NP-hard problem,task scheduling problem usually Can be solved by heuristic algorithm.ACO(Ant Colony Optimization)is a heuristic algorithm which finds optimum solution by its positix7 e feedback and distributed cooperation mechanism.With strong parallelism and global searching capacity,it has great advantages in solving task scheduling problem in Cloud environment.In this thesis we present a task scheduling policy called TSACO(Task Scheduling Based On ACO)in Cloud environment based on improved ACO.A balance of the minimum execution time and load balance of task schedule are gotten of this algorithm.The TS-ACO also absorbs the advantages of some refine ant colony algorithms occurred recent years.Considering that problems of the host switch machine fluctuation,we present a cloud computing resource scheduling strategy based on time series prediction.We use the time series prediction module to make real time prediction of load variation trend in the future on cloud computing data center or designated host machine.The result is reported to the physical machine power management module.The power management module analyzes the information collected from the time series prediction module,and determines whether to shut down the physical machine or leave it in a sleep state.On the basis,we implement the proposed TSACO and build the simulation,experiment environment based on Cloudsim.We compared TSACO with the polling scheduling policy in Cloudsim from different aspects through the simulation experiments.The results demonstrate the effectiveness of TSACO.