VM Performance Optimization Virtual Machine Migration Method Based on Ant Colony Optimization in Cloud

Virtual machine migration (VMM) is one of the most commonly used technologies in cloud platforms. However, the existing VMM methods did not try to optimize VM performance for cloud users when migrating VMs, so as to affect the users' experiences. In this paper, a VM Performance Optimization Virtual Machine Migration method (POVMM) is proposed, which can bring benefits to both cloud users and cloud service providers. It first takes the workload of PM into account to establish an improved workload-based VM performance model, and then uses the trained model to predict VM performance after migration. Then it formulates VMM as a multi-objective optimization problem, whose optimization objectives include maximizing VM performance, minimizing migration costs of all migrat-ed VMs and reducing the number of working PMs in cloud. Lastly, an ACO-based (ant colony optimization) algorithm, POVMM, is proposed to obtain the approximate optimal solution of the VMM problem. The simulation experiment is completed on the cloud simulation software, CloudSim. Through comparing with the other VMM algorithms, the POVMM algorithm has better results, which proves the effectiveness of the POVMM algorithm.