Cost Adaptive VM Management for Scientific Workflow Application in Mobile Cloud

In this paper, to guarantee Service Level Agreement composed of the deadline and budget given by users for workflow application services in mobile cloud, we propose the two-phases algorithm with a cost adaptive VM management. Firstly, the greedy based workflow co-scheduling phase schedules a workflow by using a resource consolidation in a parallel manner to decrease a cost with the deadline assurance. Secondly, the resource profiling based placement phase locates a VM to a certain physical host in the multi-cloud using the profile on the property of clouds in order to comply with the budget while maximizing the service quality. We implement mobile cloud brokering system with the two-phases algorithm and demonstrate that our proposed system outperforms traditional cloud systems through several experimental results.