Hybrid BA-GA Approach for Cost-Effective SaaS Placement In Cloud Environment

Abstract

The increasing demand of software service in cloud environment needs strategic placement in the cloud infrastructure. Thus, in which the users use the service based on the service model of the provider and pays based on their use of the resources. These resources are storage, memory processing element and bandwidth. Efficient optimal placement is the main issue in order to provide a cost effective service to the user. In the Software task placement activity, the problem is a selection of virtual machine resource to place software task component that meets the requirement of the resources. However, selection of virtual machine resource to optimize the infrastructure resources is not an easy task to place software task component in the cloud environment. The problem is an optimization problem for initial software task placement. The aim of this study focus on how software task component can be placed in a cost-effective way to optimize the infrastructure resources. This research has proposed hybrid approaches to addresses the initial software task placement problem by exploring the advantage of both Bat algorithm (BA) and Genetic algorithm (GA), to make the initial ST placement processes optimum and cost effective. In order to evaluate the performance of the proposed hybrid algorithms, an experimental environment is configured using CloudSim simulation tool. The proposed solution performance has evaluated and compared with those existing placement algorithms such as genetic algorithm (GA) and particle swarm optimization algorithm (PSO). Accordingly, the performances of the algorithms are examined by varying the number of datacenters, virtual machines, and tasks. In the experimentation, the proposed algorithm has produced an optimum solution in both located and across datacenters that are compared with genetic and particle swarm optimization placement algorithms. The result from the hybrid algorithm has reduced the initial placement cost by up to 2 -13% on a cloud environment.