Research On Task Scheduling Strategy Of Cloud Computing Based On Group Intelligence Optimization Algorithm

With the rapid development of IT industry, the development of information technology has entered the new era of big data, cloud computing, data mining.Cloud computing is a new service to users as the center of business model, using virtualization technology which links different geographic locations and different functions of the computing node resources into a virtual resource pool for use by the users. Due to the cloud computing refers to the efficient computation and data communication and storage, virtualization technology, parallel. What is more, the status of cloud computing system is constantly changing. Therefore, cloud computing task scheduling strategy is the key to the realization of cloud computing. However, on the one hand, the cloud task scheduling algorithms in the present only focus on how to shorten the task completion time and ignore the needs of users to the actual service quality; On the other hand, they only pay attention to the cloud task efficiency, while ignoring cloud services resources utilization and economic benefits. Hence, it is extremely important significance to carry out scheduling strategy for research to improve the cloud computing service ability and broaden the cloud computing applications.This paper has done the following work which is relevant to cloud computing user QoS constraints of load balancing problem and cloud computing task execution cost budget problem based on swarm intelligence optimization algorithms.(1) Through the feasibility analysis of credibility to cloud computing task scheduling, the credibility is introduced into the ant colony, and the credibility of the virtual machine is introduced to the transfer probability formula of ants. Finally, This paper present a cloud task scheduling strategy based on the credibility ant colony optimization algorithm.(2) In view of the need for cloud computing task scheduling to meet user requirement which is load balance and users’ QoS. The paper introduces a design for independent tasks considering time and load balancing based on credibility of ant colony algorithm. So cloud computing system can be maintained in a relatively balanced state. For cloud computing services to the user point of view, the users only pay attention to the service results and execution cost. This paper proposes a new approach which meet the prerequisite of completion time of task execution, consider the cost factor and based on improved particle swarm optimization algorithm for workflow tasks.(3) The application of CloudSim simulation environment for the simulation task scheduling strategy. The results show that the strategy of task scheduling algorithm based on CTLBACO not only improves the cloud computing task execution efficiency, and the system load is relatively balanced and hold high resource utilization rate. The task scheduling strategy of based on improved particle swarm algorithm optimization shorten the task completion time and reduce the cost of task execution.