Research On Cloud Resource Scheduling Method Across Cloud Computing Federation

Cloud computing becomes the fifth utility after water, electricity, gas and communication. With the continuous development of cloud computing, the resource of a single cloud provider is limited, when the number of users increases, a single cloud provider may not be able to meet all the user’s request and can’t guarantee the quality of service to the resources they provide, coupled with the increasing number of cloud service provide cloud providers, the cooperation among different cloud providers and the cloud computing federation may become true. The number of cloud users and cloud service providers in cloud computing federation is much more than a single cloud provider platform. It is a problem for the cloud computing federation to choose the servise privider’s resourse.It is the scheduling problem of the cloud computing federation. Existing research about the cloud computing federation is less, some research ablout cloud computing federation just discuss the architecture of cloud computing federation, nearly nobody studies on the research about resource scheduling of cloud computing federation.Based on above problems, this paper established the cloud computing federation resource scheduling model and put forward the user interests maximization resource scheduling method based on genetic algorithm and suppliers benefit biggest resource scheduling method based on ant colony, then implemented the resource scheduling algorithm based on CloudSim. The main work includes:(1) The cloud resource scheduling model is built. Cloud computing resource scheduling model consists of three parts:cloud users, cloud service providers, cloud federation coordinator. Cloud users drive the cloud computing federation, cloud users use the resources that the cloud providers provide and interact with them. Cloud service providers are provided by a series of different service providers. The resources provided by different providers are heterogeneous, which have different resource parameters and performance, so the resource cost and the price for cloud users are different. Cloud federation coordinator includes the trading strategy knowledge base and cloud service provider information knowledge base. Trading strategy knowledge base is used for market trading strategy; the scheduler needs to choose different strategy in the knowledge base according to user’s different needs.(2) The user interests’ maximization resource scheduling method based on genetic algorithm. From a user perspective, considering the vm size and the resource of provider’s price, the cloud computing federation chooses the vm and cloud provider to make the total task cost minimum. The implementation of scheduling uses the genetic algorithm, the chromosome of genetic algorithm is divided into two parts, the first part of the chromosome presents tasks to virtual machines, the second part of the chromosome implementation presents vms to cloud providers, and find the optimal solution for the task-virtual machine-data center.(3) The cloud provider interests’ maximization resource scheduling method based on ant colony Optimization. Task scheduling considers from the cloud provider’s point of view to make the data center on the supplier’s cost lower and resources’s load balancing.This article considers the lowest total cost of the providers and the task scheduling in different data centers to a different supplier of task scheduling for suppliers, the executing cost difference is small, the transport cost is larger, so this paper considers the transport cost of tasks to the data center, and load rate will produce a correction factor, load rate in the process of ant colony algorithm affects the convergence direction of the ants, the optimal result satisfes the cost and load balencing.(4) The Realization of cloud resource scheduling algorithm based on CloudSim. CloudSim is used to realize the simulation of the scheduling process, through rewriting the class DatacenterBroker under the bag of CloudSim class in the resource to realize the user interests maximize resource scheduling method based on genetic algorithm and the cloud provider interests maximization resource scheduling method based on Ant Colony Optimization.lt can be observed that the resource scheduling execution process and output the final result.