The Research Of Task Scheduling Based On The Cloud Environment

Cloud services as a typical commercial service payment model, uses virtualization technology to integrate resources, providing high availability and high security services to users via the internet.In the Cloud environment,different users to submit multiple and complex tasks,and the cloud resources performance is not the same, so to design a rational and efficient resource allocation policy to shorten the task execution time, reduce the cost of system and users,achieve the load balancing of the entire cloud computing system has become a hot issue in cloud computing research area.1)Aiming at the disadvantage of the ABC algorithm(Artificial bee colony algorithm,ABC) is easy to fall into local optimum,the paper proposes an improved artificial bee colony algorithm(Improved Artificial bee colony algorithm, IABC).And optimized improvements in three areas.A) optimization probability selection model, the introduction of the maximum and minimum fitness value to avoid because of the continuing convergence of the optimal solution greed caused the population to narrow diversity;B) optimization the bee role reversal model,the bees will according to the ratio of earnings and the ratio of following to adjust the role of bees to avoid the unscientific of bee allocated number cause reduced efficiency optimization;C) optimization solution space search strategy model,according to the three parameters of current optimal solution, global optimal solution and learning factor to strengthen the search efficiency in the leading bee stage and following bee stages.2)Aiming at the past scheduling algorithm to achieve a minimized task completion time as a single target and ignore the problems of the QoS requirements of the user in the cloud environment, the paper proposes a multidimensional Qo S cloud task scheduling algorithm based on improved artificial bee colony algorithm.First, to meet the multi-dimensional Qo S of users,give full consideration to the completion time, costs of implementation and the reliability requirement when the task is executed,build a cloud computing model, a resource Qo S model, the criterion that basing on the user preference to evaluate the performance evaluation of virtual resources is proposed.Secendly,to improve the efficiency of resource allocation,the improved artificial bee colony algorithm(IABC)was applied to a cloud environment, selects a higher performance or more in line with the user preferences assigned to the task.Finally, the CloudSim simulation tool is extended by simulating a DAG Builder, by contrast with the other four algorithm,the experiments show that the algorithm can achieve high efficiency task allocation of resources and meeting the user's demand of Qo S.3) Because of the current scheduling algorithm ignores the problem of multi-attribute characteristics of the cloud user tasks, the paper proposes a cloud computing task scheduling algorithm based on fuzzy clustering and improved artificial bee colony.First, the fuzzy C-means clustering(Fuzzy C-Means, FCM) algorithm to classify tasks,than the task will be divided into computational--type task, bandwidth-type task and storage-type tasks, and task sorting between classes and classes in the order.Secendly,the criterion to evaluate the performance evaluation of virtual resources is proposed,and the improved artificial bee colony algorithm(IABC)was applied to a cloud environment, selects a higher performance or morein line with the user preferences assigned to the task.Finally, the simulation experiments,CloudSim platform can be extended to generate the required random simulation task instance,by contrast with the Min-Min algorithm and so on, the results show that the algorithm can achieve efficient resource scheduling tasks and can achieve more high resource utilization.