Performance of particle swarm optimization bin packing algorithm for dynamic virtual machine placement for the consolidation of cloud server. Issue 1 (March 2021)

• Record Type: Journal Article
• Authors: Pandiselvi, C.; Sivakumar, S.
• Is Part Of: IOP conference series. Volume 1110:Issue 1(2021)
• Publication Date: 2021-03
• Subjects: Virtual machine placement -- Energy consumption -- Resource utilization; Materials science -- Periodicals; 620.1105
• DOI: 10.1088/1757-899X/1110/1/012007
• Languages: English
• ISSNs: 1757-8981

Abstract

Abstract: Infrastructure as a service offered by the cloud computing is one of the most important service. It allows physical machines to get virtualized by creating many instances of virtual machines. Mapping virtual machines on physical machine has become the major challenge in cloud data centres. The dynamic virtual machine placement methods are used to solve this issue with objectives like maximizing the resource utilization, minimizing the energy consumption and maximizing the scalability of data centres. In this paper a virtual machine placement-based bin packaging algorithm is proposed and analysed with four different fitness strategies to obtain the optimal solution. The unimodal (Sphere, Step) and multimodal (Graywang and Rastridge) benchmark functions are used with proposed algorithm for the analysis and obtain the quantitative measurements. The results show that optimizing the mass of particles using the best fitting strategy reduces the energy consumption, resource utilization and improved the scalability of data centres.