Genetic Algorithm Based Scheduling Method for Efficiency and Reliability in Mobile Grid

Abstract

A mobile grid is emerging as a new opportunity to integrate and utilize the exploding population of mobile resources. However, the mobile grid is not as stable as the traditional fixed grid, and the scheduling scheme plays a crucial role for the efficient job execution. In this paper, we take into account the characteristics of mobile devices such as inferior reliability and limited capacity in scheduling. With the replication scheme adopted to enhance reliability, we propose a novel scheduling scheme for the efficient execution of replicated jobs. Our proposed method is based on genetic algorithm. It strives to minimize the completion time of grid applications and reduce the replication overhead. Our method provides the efficient utilization of mobile grid resources by avoiding executions of useless replica. Experiment results show that our method outperforms the other conventional methods in terms of completion time and wasted resources.