Cloudlet scheduling using merged CSO algorithm

Abstract

Cloud computing is used to provide information technology services to users over the internet. Scheduling is a very important aspect of cloud computing to manage resources and user demands. A numerous number of users are requesting the Service Providers for various requests which needs to be scheduled for an efficient optimization of various constraints. There exist a number of algorithms for Cloud Task Scheduling based on Swarm Intelligence. Two of them are Particle Swarm Optimization (PSO) algorithm and Cat Swarm Optimization (CSO) algorithm. Both the algorithms have various merits and demerits. In this paper, a merged CSO(MCSO) is presented with the aim to combine the merits of CSO and PSO to obtain better results. The basic purpose of developing this algorithm is to minimise the execution cost and time for running the algorithm to reach that optimised cost mapping. The algorithm steps are described, and it is implemented in CloudSim simulator and much-improved results are obtained. The algorithm performs comparatively well as compared to both PSO and CSO in terms of execution time and convergence. © 2018 IEEE.