Research For Energy-optimized Resource Scheduling Algorithm In Cloud Computing

Cloud computing is becoming a popular network computing model. It can help the users to ignore the complex calculation process. The users and enterprises need not to pay for lots of hardware devices and can obtain numerous computing abilities and service resources through the application program interfaces. With the amount increase of global servers provided by the cloud computing business enterprises, the corresponding energy consumption gradually attracted people’s attention. Unfortunately, the optimization problem of energy consumption in the resource scheduling of cloud computing has rarely been taken into account. Therefore, through the reassignment of system resources and tasks scheduling, some resource scheduling algorithms of energy consumption optimization in the process of cloud computing resource scheduling are proposed in this paper.Firstly, the research background and significance of the subject are introduced. The cloud computing system is studied in-depth, and the development situations of current cloud systems both at home and abroad are summarized. The cloud computing services, service capabilities and implementation mechanism are studied systematically. Then the resource scheduling framework of energy optimization are designed from perspectives of the task level, resource scheduling layer, the virtual resource and physical link layer. The survey of cloud computing resource scheduling and energy consumption optimization are analyzed in-depth and the existing problems are proposed as well as the ideas to solve them.Secondly, the resource scheduling power and time of random tasks and the resource scheduling strategies of energy optimization in cloud system are analyzed in-depth based on the energy optimization frawework of resource scheduling, and the problem model of cloud computing is proposed as well. For the relatively small random tasks, the resource scheduling energy optimization model based on the time prediction of clound computing is proposed to optimize the low-efficent utilization. According to this model, the resource scheduling algorithm based on the time prediction is designed, in which the time predictors are proposed to real-time adjust the accuracy of every time prediction and a historical time slot algorithm is designed to store and update the time nodes. The algorithm is validated in experiments. The experiments show that this scheduling algorithm can effectively save the energy consumption when the cloud system is in low efficency.Thirdly, for the relatively large random tasks in cloud computing system, the resource scheduling algorithm of energy consumption optimization based on the task patience value is proposed. The executed tasks of cloud system are studied. The concept of relaxation time is brought forth and the task time-out patience function is defined. The waiting time of the tasks are reduced by adjusting the allocation of resources. The patience function value can be calculated through the current network important level and the duration of time of task execution in cloud system. The energy consumption model of resource scheduling based on the task patience value is proposed and the resource scheduling algorithm is designed to optimize energy consumption. The experiments show that the proposed algorithm can optimize the energy consumption much better than those previously mentioned algorithms.Furthermore, for the relatively middle random tasks in cloud computing system, in order to avoid much waste of energy consumption in the data transmission between multiple data centers, the resources relevancies between the cloud system data centers are studied and the resource-related concept is proposed. The tasks can be scheduled by calculating the size of the resource relevancy. Then the scheduling model of cloud computing based on resource relevancy is designed to decrease the energy consumption with the increase of resource relevancy of energy in scheduling. The resource scheduling algorithm based on resource relevancy is proposed to optimize the energy consumption in the process of transmission. The results of experiments show that the propsed algorithm optimizes the energy consumption generated in cloud system and improves the system efficency much better than other algorithms.Finally, for the various sizes of random tasks, the possible cost is constrained in the process of cloud resources scheduling. The electricity price of the worldwide dynamic time zones is proposed based on the different electricity price of the world time zone characteristics, the network bandwidth and load levels. The optimization model of energy consumption of could system with execution cost as constraint condition is proposed, which optimizes the energy consumption of cloud system through the load level, electricity price and other factors in the resource scheduling process. In this model, the task hierarchical strategy is designed to realize the hierarchical task energy consumption. The self-adaptive matching resource scheduling algorithm is proposed and it can switch to select the reasonable choice between the designed scheduling algorithms by judging the system resource utilization and other factors when meeting the constraint conditions of scheduling process. Thus the scheduling algorithm of energy optimization with cost constraint is proposed. The results of experiments show that the algorithm can both optimize the energy consumption and reduce the service cost.