A Dynamic Simulated Annealing Algorithm with Self-adaptive Technique for Grid Scheduling

Due to the availability of the excess computation power, the heterogeneous grid systems have significant predominance in higher performance and low cost. In this paper, a self-adaptive algorithm is proposed based on Simulated Annealing (SA) for grid scheduling to minimize the make span, in response to the grid dynamic characteristic. During the process of scheduling, partial scheduling information is exploited to decide the next partial scheduling parameters set, which is self-adaptive to the stochastic job arrival time. We simulated the algorithm in GridSim environment, and the simulation results demonstrated that the proposed algorithm improved performance compared to existing scheduling techniques.