Toward a game-theoretic model of grid systems

Abstract

Computational Grid is a promising platform that provides a vast range of heterogeneous resources for high performance computing. To grasp the full advantage of Grid systems, efficient and effective resource management and Grid job scheduling are key requirements. Particularly, in resource management and job scheduling, conflicts may arise as Grid resources are usually owned by different organizations, which have different goals. In this paper, we study the job scheduling problem in Computational Grid by analyzing it using game theoretic approaches. We consider a hierarchical job scheduling model that is formulated as a repeated non-cooperative game among Grid sites, which may have selfish concerns. We exploit the concept of Nash equilibrium as a stable solution for our game which eventually is convenient for every player. © 2010 Springer-Verlag Berlin Heidelberg.