Time Optimized Hybrid Scheduling Algorithm for Cloud Computing Environment

Abstract

Recently, cloud computing has become a new global trend in computing. It's an innovativeway to use the power of the Internet to provide remote resources. It is a new solution andstrategy to achieve high availability, flexibility, cost reduction and expansion in demand.However, cloud computing has many challenges such as misuse of resources which has aprofound effect on the performance of cloud computing. These problems are caused byhuge amount of information. Thus, the need for efficient and powerful cloud computingtask scheduling algorithms to improve cloud computing performance is a major issue inthis field. Many researchers have proposed different algorithms for scheduling cloudcomputing tasks; there is still some insufficiency in system performance and low resourceusage. Therefore, this study proposes a hybrid task scheduling algorithm to improve theperformance and efficiency in a heterogeneous cloud computing environment. Theproposed algorithm is called Time Optimized Hybrid Scheduling Algorithm for CloudComputing Environment. It works based on the principle of Round Robin (RR) andShortest Job First (SJF) algorithm with dynamic quantum time. . In this work we take theadvantages of both RR and SJF algorithms. We take time quantum as 80 % of themaximum burst time and a task lesser than the quantum time will be assigned in SJFmanner and tasks larger than the quantum time will be assigned as their arrival time; if anew task arrives the scheduler recalculates the time quantum. The algorithm is evaluatedusing the CloudSim simulator and compared with RR, SJF and EDRR algorithms. As aresult, we found that average turnaround time and average waiting time were minimizedand performance was improved compared to other algorithms.