Chaotic social spider algorithm for load balance aware task scheduling in cloud computing

In recent years, the revolution of cloud computing has taken the IT business to greater heights with the rapid sharing of vast web resources over the internet. Proficient task scheduling and balanced task distribution is still exists as a major challenging issue in cloud computing system due to dynamic heterogeneous nature of resources and tasks. It is a NP-hard problem where the scheduler needs to find the best optimal virtual machines with minimum makespan and proper resource utilization. The major part of this problem is to design an efficient intelligent searching pattern to schedule the tasks in best virtual available machines. In this paper we propose a meta heuristic algorithm called chaotic social spider algorithm inspired by social spider to tackle the problem of task scheduling in various heterogeneous virtual machines. This paper focus on minimizing overall makespan with effective load balancing by modelling the swarm intelligence of social spider with chaotic inertia weight based random selection. The proposed algorithm prevents the local convergence and explores the global intelligent searching in finding the best optimized virtual machine for the user task among the set of virtual machines with minimum makespan and balanced resource utilization. We have made the simulation and performance evaluation using cloudsim toolkit and compared the results with other swarm intelligent based algorithms such as GA, PSO and ABC. The evaluation results show that there is a major improvement in minimizing the makespan with balanced task distribution.