A design for emergence method applied to recurrent cellular computing systems with multi-nested cells
Radu Dogaru, Tudor C. Ionescu, Manfred Glesner · 2004
Cellular Neural Networks (CNN) are a convenient paradigm for compact and fast multidimensional signal processing in mixed signal technologies. This paper proposes a novel CNN framework where the cell is a Boolean universal multi-nested neuron with a very compact VLSI implementation, and derives a design for emergence method to determine the genes (parameters) of the CNN cells such that meaningful computation emerges in the recurrent CNN system.