Study of Grid Task Scheduling Algorithm Based on Benefit Function

In the dynamic, heterogeneous and widely distributed grid environment, task scheduling is a complex and challenging problem. Because in DBC algorithm, QoS requests are static, either based on time optimisation, or budget optimisation, resource allocation in grid is irrational. To solve the problem, one QoS model based on benefit function is presented in this paper. Also, two algorithms based on benefit function are given. Using the GridSim simulator, these two new algorithms and DBC algorithm are simulated and compared. The experiment results indicates that the two algorithms can achieve better performance in terms of time utilized, budget spent and gridlets completed. Otherwise, the second algorithm is better than the first one.