A DISTRIBUTED APPROACH TO ANT COLONY OPTIMIZATION

Swarm Intelligence(SI) is the emergent collective intelligence of groups of simple agents. Economy is an example of SI. Simulating an economy using Ant Colony algorithms would allow prediction and control of fluctuations in the complex emergent behavior of the simulated system. Such a simulation is far beyond SI's capabilities, which is still in its infancy. This paper presents a distributed approach implementing Ant Colony Optimization(ACO). We present our agent based architecture of ACO and initial experimental results on the Travelling Salesman Problem. The innovation of our work consists of: i)representing network nodes as software agents, ii) representing software agents as software objects that are passed as messages between the nodes according to ACO rules.