Synchronized chaos in coupled neuromodules of different types

We discuss the time-discrete parametrized dynamics of two coupled recurrent neural networks. General conditions for the existence of synchronized dynamics are derived for these systems, and it is demonstrated that also the coupling of totally different network structures can result in periodic, quasiperiodic as well as chaotic dynamics constrained to a synchronization manifold M . Stability of the synchronized dynamics can be calculated by Lyapunov exponent techniques. In general, in addition to synchronized attractors there often co-exist asynchronous periodic, quasiperiodic and even chaotic attractors. Simulation results with respect to a minimal coupling scheme for neuromodules of different type are presented. ∗in: Proceedings of the International Joint Conference on Neural Networks (IJCNN’99), Washington, July 10-16, 1999, IEEE-CD, Piscataway 1999.