An Improved Method for Scheduling of Scientific Workflows

Abstract: Cloud computing is the latest distributed computer paradigm where the resources are delivered over the Internet as services. It follows a pay-as-you-go model where users are charged with respect to their use. A lot of researches are being done on the scheduling in cloud computing and most of them are about workflow and job scheduling. A workflow is defined as a sequence of tasks interconnected via data or computing dependencies. Many scheduling policies have been proposed which aim to maximize the amount of work completed while meeting QoS constraints such as deadline and budget. However many of them are not optimal because they do not incorporate some basic principles of cloud computing such as the elasticity, heterogeneity of the resources. In this paper a meta-heuristic optimization technique, HDPSO is used to minimize the execution cost. Other parameters like makespan, execution time are considered within heterogeneous and homogeneous environment of virtual machines.

Keywords: Cloud Computing, HDPSO, QoS, Scheduling, SLA, Workflow.