Exploring Behavioral Dynamics in Systems of Systems

This paper summarizes the early stages of research into methodologies for understanding the governing principles in systems of systems. An approach is suggested utilizing computational intelligence, agent based modeling, and wireless ad hoc network simulation as a computational testbed for exploring the generalized dynamics of complex adaptive systems. Utilizing a standard 802.11 ad hoc protocol set, the network can be understood in terms of its capabilities and limitations. Agent based modeling facilitates the individual autonomy aspect of system of systems participants and is responsible for determining how a node will use the 802.11 protocol. Evolutionary algorithms can be used to evolve the protocol usage strategies of the individuals. It is proposed that the evolutionary algorithm and agent based model provide the flexibility and autonomy needed to simulate a representative system of systems while the clear definition of the 802.11 protocol provides a framework for the experiment.