Computing resource requisition method based on game model

Abstract

The present invention relates to a computing resource expropriation method based on a game model, including: A. Formulating user service quality indicators in the game model, including task response time, user price overhead, system effectiveness, reliability and credibility; B. Corresponding allocation of units and tasks, establishing particle encoding model; C. Optimizing particle swarm with global income as optimization goal; D. Calculate the maximum value of global income according to the gravity search algorithm, and judge whether the number of iterations in the algorithm reaches the maximum, if so, then End, otherwise continue iteration from step C. The present invention can accurately represent the optimization target of the distributed resource requisition method, significantly improves the accuracy of the algorithm, and maximizes the global benefit by maximizing the multiple goals of resource requisition, which greatly improves the Resource requisition optimization degree.

Description

technical field [0001] The invention relates to a method for requisitioning computing resources, in particular to a method for requisitioning computing resources based on a game model. Background technique [0002] The distributed system can effectively manage various resources scattered in the network by realizing the management of distributed resources, so that multiple users can share and use the same resource in the network. In order to meet the user's task application, the distributed system needs to requisition the resources in the network to ensure the smooth completion of the task. Therefore, which computing resources are requisitioned will directly affect the user's QoS (Quality of Service) and the overall energy consumption of computing units (physical machines). The ensuing problem is how to establish a model to describe the relationship between user QoS and energy consumption of computing resources, and search for an optimal computing resource allocation result....

AI Technical Summary

Problems solved by technology

Some existing algorithms often only focus on cost control and resource utilization, ignoring the impact of load balancing on system performance. Although the contradiction between saving resources and user needs is alleviated to a certain extent, the scale of resources in distributed systems Problems such as large differences and complex composition directly lead to the waste of distributed system resources. Nowadays, there is no good distributed resource requisition method to quickly realize the load balancing of distributed systems. Therefore, it is important to study advanced distributed resource requisition methods. Practical and Academic Significance

Images