Research Of Placement Strategy Of Virtual Machine For Multi-objectiive Optimization Based On Improved Ant Colony Algorithm

Cloud computing is a new model and business mode,which using distributed physical machine cluster to integrate computing resources and information services,making cloud service providers to provide users with high performance computing services.With the development of cloud computing services,the physical machine cluster that supports these services also need to be expanded.The expansion of the system scale caused by the system will lead to the complexity of system management and the inherent dynamic of the system.Virtualization technology through the virtual machine in the form of physical machine cluster cutting independent management,in order to allocate the form of distribution of resources to support the cloud platform for resource sharing.How to build a reasonable and efficient physical machine virtual machine mapping relationship,to ensure that the cloud platform with a higher performance and resource utilization has become a key issue in the industry.This paper aims at the problem of cloud platform resource scheduling management.Analysis of the research status of cloud platform resource management at home and abroad,summarizes the problems existing in current research,the virtualization technology and the previous version of the ant algorithm is analyzed and summarized,and the emphasis is placed on the cloud platform virtual machine research.The main research contents and innovations of this paper divide into two parts:?.Virtual machine placement strategy and modelingAiming at the problem of service quality,energy utilization and energy consumption of the user,the strategy of virtual machine placement based on multi objective optimization is proposed.The concept of SLA compliance rate,load balancing rate and energy power is proposed,and the mathematical model is described in detail,and the mathematical model is established for the three optimization objectives.?.Improvement of algorithm(1)First,Redefine heuristic factors.Give more weight to heuristic factor,the path selection probability more influence,make the algorithm more rational in the choice of the path,and can be in multiple conflicting objectives between the compromise,the improved SLA performance rate and load balance rate,reduce energy consumption and increase energy power.(3)Second,Re formulating the pheromone update strategy.In the algorithm,the pheromone can be accumulated at the same time,but also according to the weight order.In addition,in order to solve the problem of "premature" and easy to fall into local optimum,the upper and lower limits of pheromone are set.(3)Third,Reformulate state transition rules.Three kinds of state transition rules are established for ants,which make the algorithm have a higher accuracy and a wider search space,and more suitable to find the optimal path of the target.(4)Last,Submit the catastrophe strategy.Drawing on the idea of variation in genetic algorithm,Then using small probability model,proposed catastrophe strategy.Pheromone concentration on the manual reset path,It is better to have better fault tolerance and robustness when the environment is dynamically changed and the local optimal is trapped in the algorithm.