| We address job scheduling in decentralized data centers with considering distributed energy that is available near the data centers. Because it is sometimes better to delay job processing for using more distributed energy, the job scheduling problem is solved as a multi-objective optimization problem to minimize energy cost and delay time. Conventional methods to solve the multi-objective scheduling problem are fast but can not derive Pareto optimal solutions accurately. In this paper, we regard the job scheduling problem as a shortest path problem whose path length is defined by energy cost or delay time, and apply Ant Colony Optimization that is a well-known method to solve the shortest path problem. |