A genetic based workflow scheduling considering data transmission time

Many complex e-science and e-business applications are modeled as workflows. In this paper we consider workflow scheduling problem in distributed heterogeneous environments. Most of current researches have ignored the data transmission time when computing the optimal scheduling. In this paper, we show the side effect of this ignorance in data intensive workflows and high bandwidth heterogeneity situations. We propose a genetic based scheduling algorithm. The evolution process pays attention to minimize the workflow completion time containing both communication and computation time. Simulation results show significant reduction in makespan in different workflows.