A Spectral Approach to the Study of Propagation Phenomena in CNNs

The propagation phenomena occurring in cellular neural networks (CNNs) described by a one-dimensional template are investigated using a spectral technique. The CNN is represented as a scalar Lur'e system to which a suitable extension of the describing function technique is applied for predicting the existence and characteristics of spatiotemporal periodic oscillations. It is shown that such a method yields results that are in good agreement with those observed by time simulation of the system.