Task scheduling in cloud computing using particle swarm optimization with time varying inertia weight strategies

Abstract

Cloud computing is an efficient technology to serve the requirement of big data applications. Minimizing the makespan of the cloud system while increasing resource utilization is important to reduce costs. In this case, task scheduling is a challenging task to meet the requirement because it requires both effectiveness and efficiency. This article proposes a task scheduler with several discrete variants of the particle swarm optimization (PSO) algorithm for task scheduling in cloud computing. In order to evaluate the performance, these approaches were compared with three well-known heuristic algorithms on task scheduling problems. Experiment results demonstrate the efficiency and effectiveness of the proposed approaches. For the proposed PSO-based scheduler, an appropriate choice is to use the logarithm decreasing strategy to provide an optimal scheduling scheme. The average makespan of the proposed PSO-based scheduler that adopts logarithm decreasing strategy is reduced by 19.12%, 21.42% and 15.14% relative to the compared gravitational search algorithm, artificial bee colony algorithm and dragonfly algorithm respectively.