Many complex scientific problems can be formulated with a regular 2-D or 3-D grid. Direct interaction between the signals on various grid points is limited within a finite local neighborhood, which is sometimes called the receptive field. The original cellular neural network (CNN) paradigm was first proposed by Chua and Yang in 1988 [1, 2]. The two most fundamental ingredients of the CNN paradigm are: the use of analog processing cells with continuous dsignal values, and local interaction within a finite radius. Many results on the algorithm development, VLSI implementations of CNN systems are reported in the first three IEEE International Workshops on Cellular Neural Networks and Their Applications (Budapest, Hungary, 1990; Munich, Germany, 1992; Rome, Italy, 1994); the book entitled: Cellular Neural Networks, which was edited by T. Roska, J. Vandewalle; and papers published in IEEE Trans. on Circuits and Systems, and other IEEE journals and conference proceedings.