"Event-Driven Load Balancing of Partially Replicated Objects Through a
Swarm of Mobile Agents"

James Schwing
Central Washington University

Swarms of mobile software agents, imitating the behavior of insects, can solve
complex tasks. Such agents individually have simple behavior. However, as a
collective unit, constructive behavior emerges, as it does in insect colonies.
The problem we address is optimizing the distribution of objects with partial
replication over a computer network. We present a new approach, implemented as
a prototype, based on swarm intelligence. The system performs dynamic,
event-driven, distributed load balancing.  Every node of the network is
capable of producing new events and introducing them into the network for
computation.  The architecture is general and can be used as a support for
clustering and task allocation applications.