Ant inspired server population management in a service based computing environment

This paper presents an approach, experiments and a practical application to balancing service availability for a multi-agent system deployed on a wireless mobile ad hoc network. Finding the optimal number of services required to achieve the desired effect is a difficult and problem-specific task. We propose an approach inspired by the quorum sensing behavior of the leptothorax albipennis ants. Similar to a real ant colony, the solution exhibits properties of emergent stability, decentralized control, and resilience to possible disturbances. The experiments confirm the validity of our technique in wireless ad hoc network environments.