NSGA-II with Local Search for Multi-objective Application Deployment in Multi-Cloud

Cloud computing allows software application or service providers to purchase computational resources for hosting software applications or services. There is an increasing number of cloud service vendors, each offering different types of computational resources, such as virtual machines (VMs), in distinct physical locations, meaning that the chosen computational resources directly affect the cost and performance of the deployed applications. Given the increasing number of cloud providers and types of computational resources, the process of selecting the most suitable resources to meet multiple quality of service (QoS) requirements becomes quite challenging. Thus, the aim of this paper is to propose an efficient and effective approach to the problem of application deployment in multi-cloud with the objective of minimizing overall deployment costs and response time. This is accomplished by employing a hybrid NSGA-II approach with a local search method. Experimental evaluations with benchmark datasets demonstrate that our proposed hybrid approach outperforms NSGA-II and SPEA2.