A Load Balancing VMs Migration Approach for Multi-tier Application in Cloud Computing Based on Fuzzy Set and Q-Learning Algorithm

Cloud computing technology provides computing resource as utilities which face several resources management challenges. In the infrastructure service cloud, how to efficiently manage VMs to balance both economical cost and ensure QoS is focused by many researchers. VMs migration solution for cloud computing aims at different purposes such as cost, energy, performance, stability, availability, and depending on stakeholder’s objectives such as service providers, customers, and end users. However, its objectives are usually conflicted with each other. In addition, the optimal resource utilization problem in cloud infrastructure is usually the NP-Hard or NP-Complete class. In this paper, we formalize the VMs migration problem and develop a corresponding algorithm using Markov decision process and Q-Learning. The results of the simulation demonstrate the efficiency of our proposal. The proposed VMs migration algorithm has been benchmarked to the Round-Robin algorithm in order to highlight their strength and feasibility.