Abstract
The more simulation becomes an established tool in the design of
multi-agent systems, the more urgent the question becomes how valid
the induced properties and behavior patterns are. Answering this
question depends on the validity of the models, the correctness of
the simulators, and the simulations. In all of these aspects, reuse
and a declarative representation plays a crucial role. With James
II, we developed a customizable framework for modeling and simulation.
Its component-based architecture supports a reuse of models, simulators,
and experimental settings. The benefits of this architecture for
agent-based modeling and simulation will be illuminated by an excerpt
of a simulation study about trading strategies for mobile ad hoc
networks.
Users
Please
log in to take part in the discussion (add own reviews or comments).