Stability assessment of the leech swim oscillator using a VLSI-based model

VLSI-based artificial neurons are used to realize neuronal oscillatory networks. A comprehensive electronic model of the Hirudo swim oscillator is recreated and subjected to numerous parametric variations. This, combined with similar analyses of the mutually and cyclically inhibited subnetworks embedded in the network, provides a good indication of the origins of network stability. Test data indicate that different sub-network types exhibit immunity to contrasting parameters, and that the overall swim oscillator network is more stable than either sub-network type alone.