Relationship networks as a survivable and adaptive mechanism for grid resource location

Abstract

Grid technologies have emerged to enable large-scale flexible resource sharing among virtual organizations. Resource location of next generation grid systems needs to be fully decentralized, more importantly, be survivable and adaptive to uncertain factors in grid environments. This paper applies the concepts and principles of relationship networks in human society to design a dynamic location approach. The simulation results have proved that our approach can: form relationship clusters to improve the location performance effectively; well adapt to different resource distributions and user request patterns; and survive from the changes of dynamic environments such as partial failure of agents. © Springer-Verlag Berlin Heidelberg 2005.