Authors: P. Karthikeyan; Rinta Soni
Addresses: Department of Computer Sciences and Engineering, Karunya Institute of Technology and Sciences, Coimbatore, India ' CSS Corp, Chennai, India
Abstract: The service to the end user in cloud computing is offered as virtual machine instances as demanded for a specified duration of time and billed on pay per use basis. A major problem faced in cloud computing is the virtual machine scheduling problem. The existing algorithms efficiently cannot satisfy the requirements with respect to resource utilisation, bandwidth utilisation, and cost. Also, most of them are of the fixed type which leads to wastage of resources. To overcome these problems, the hybrid particle swarm optimisation (HPSO) algorithm is proposed by efficiently allocating the resources to the users. This algorithm combines the genetic algorithm (GA) and the variable neighbourhood search (VNS) algorithm with the particle swarm optimisation (PSO) technique to increase the utilisation rate of the virtual machine as well as to minimise the total completion time. The proposed system is evaluated by performing simulations. The experimental results show that the proposed algorithm minimises the total completion time and increase the resource utilisation than PSO, GA and VNS algorithm.
Keywords: cloud computing; particle swarm optimisation; PSO; variable neighbourhood search; VNS; genetic algorithm; GA.
International Journal of Business Information Systems, 2020 Vol.34 No.4, pp.536 - 559
Received: 09 Apr 2018
Accepted: 06 Jul 2018
Published online: 17 Aug 2020 *