Research and Simulation of Task Scheduling Algorithm in Cloud Computing

Abstract

This paper focuses on the task scheduling algorithms based on comprehensive QoS and constraint of expectation. Under the environment of dynamic cloud computing, efficiency improving of task scheduling and load balancing are eternal problems. For users, however, it's more important to meet their requirements of QoS. This paper relates to a benefit-fairness algorithm based on new Berger's model under the environment of dynamic cloud computing. According to the different type of task scheduling, we describe the priority of fairness, efficiency and the Balance between benefit and fairness respectively. We recompile the CloudSim and simulate the three task scheduling algorithms above on the basis of extended CloudSim respectively. The experimental results indicate that this algorithm dose not only meet the principle of giving priority to benefit with due to consideration to fairness, but also meet users' needs of synthesized QoS. DOI: http://dx.doi.org/10.11591/telkomnika.v11i11.3513