Optimal VM Coalition for Multi-Tier Applications Over Multi-Cloud Broker Environments

In the current cloud market tradition, multi-cloud broker environments have an important role to support multitier cloud applications where a set of interconnected virtual machines (VMs) are required for the deployment and execution of the application. There are several cloud service providers (SP) who provide services in the form of different VM instances. For different SPs, pricing of different configurations is different based on various technical, service level as well as economic factors. Also, these SPs have multiple data centers located at different geographical locations with diverse resources and pricing schemes due to differences in electricity costs, maintenance costs etc. in different regions. Considering these facts, in this paper, we propose a multi-cloud broker environment to select an optimal VM coalition for multi-tier applications from an SP with minimum coalition pricing and better quality of service. We model the problem as a bi-objective optimization problem and solve it using ant-colony optimization based technique. The entire interaction of SPs is modeled using game theory and results from simulation show promising behavior in achieving the trade-off between the aggregate cost and the QoS.