Performance Evaluation of Self-organizing Agents for Grid Computing

A grid is a wide-area computing environment for cross-domain resource sharing and service integration. Software agents are one of mainstream technologies for management of geographically distributed grid resources. In this work, self-organization mechanisms are investigated for multiple agents to cooperate with each other so that an overall grid load balancing and resource scheduling can be achieved. In particular, an ant-like self-organizing mechanism is introduced and proved to be powerful to achieve overall grid load balancing through a collection of very simple local interactions. A performance modeling and evaluation environment is developed to enable performance of the ant algorithm to be investigated quantitatively. Simulation results included in this work illustrate the impact of different performance optimization strategies on the overall system load balancing level, speed and efficiency. Key Words—Grid computing, performance evaluation, multi-agent systems, load balancing, self organization, and ant algorithms.