云计算环境下基于多目标优化的虚拟机放置研究

VIRTUAL MACHINE PLACEMENT BASED ON MULTI-OBJECTIVE OPTIMIZATION IN CLOUD COMPUTING ENVIRONMENT

云数据中心的规模日益增长导致其产生的能源消耗及成本呈指数级增长。虚拟机的放置是提高云计算环境服务质量与节约成本的核心。针对传统的虚拟机放置算法存在考虑目标单一化和多目标优化难以找到最优解的问题,提出一种面向能耗、资源利用率、负载均衡的多目标优化虚拟机放置模型。通过改进蚁群算法求解优化模型,利用其信息素正反馈机制和启发式搜索寻找最优解。实验结果表明,该算法综合性能表现良好,符合云环境对高效率低能耗的要求。

The growing size of cloud data centers has led to an exponential increase in energy consumption and costs.The placement of virtual machines is the core to improve the quality of service and cost savings in the cloud computing environment.Aiming at the problem that traditional virtual machine placement algorithm considers that target singularity and multi-objective optimization are difficult to find the optimal solution,this paper proposes a multi-objective optimization virtual machine placement model for energy consumption,resource utilization and load balancing.The improved ant colony algorithm was used to solve the optimization model.It found the optimal solution through its pheromone positive feedback mechanism and heuristic search.The experimental results show that the ant colony algorithm based on multi-objective optimization performs well and meets the requirements of high efficiency and low energy consumption in the cloud environment.