A Multi-Agent Architecture for QoS Support in Grid Environment

Abstract

Grid computing is emerged as a new distributed computing technology that implements flexible and coordinated resource sharing among dynamic, heterogeneous, unpredictable and geographically distributed collections of resources owned by different individuals and organizations. Due to heterogeneous and dynamic nature of the grid, resource management and application scheduling is a complex undertaking. This paper proposed a multi-agent architecture that addressed resource management and application execution with support for Quality of Services (QoS) in grid environment. Five types of collaborative intelligent and mobile agents proposed to manage grid resources and applications in a decentralized, autonomous and intelligent manner. In the proposed architecture, negotiation, advanced reservation and QoS measurements handled by collaborative intelligent agents. Resources and applications are managed independently with respect to their defined policies. The simulation results showed that the multi-agent architecture is practical, flexible and effective.