Cloud Computing: Optimization using Particle Swarm Optimization to Improve Performance of Cloud

Abstract

INTRODUCTION: In the contemporary world cloud computing is acknowledged as advanced technology to manage and store huge amount of data over the network. To handle the network traffic and effective task scheduling some efficient load balancing algorithm should be implemented. This can reduce the network traffic and overcome the problem of limited bandwidth. The various research articles represents ample amount of optimization techniques to overcome the transfer of data with limited bandwidth. Among all, few solutions has been chosen for current research article such as – optimization of load distribution of various resources provided by cloud.
 OBJECTIVES: In this paper, Comparative analysis of various task scheduling algorithms such as (FCFS, SJF, Round Robin & PSO) have been proposed in current research article to accumulate the outcome and evaluate the overall performance of cloud at different number of processing elements (pesNumber) .
 METHODS: Overall performance of task scheduling is significantly enhanced by PSO Algorithm implemented on cloud in comparison of FCFS, SJF and Round Robin. Outcomes of optimization technique has been implemented and tested over the CloudSim simulator.
 RESULTS: The comparative analysis conducted based on scalability for increasing the number of processing elements over the cloud. The major insight of proposed algorithm has shows that results are still better when number of VMs is increased and it successfully minimizes waiting time and turnaround time and completion time by 43% which is significantly high than outcomes of existing research articles.
 CONCLUSION: To optimize the task scheduling in cloud computing, comparative analysis of various task scheduling algorithms has been proposed, including Particle Swarm Optimization algorithm.