Multi-objective cellular neural networks

Abstract

This paper presents a new kind of cellular neural network (CNN) called multi-objective cellular neural network (MCNN). Like CNN, it is a large-scale nonlinear analog circuit which processes signals in real time, and it is made of a massive aggregate of regularly spaced circuit clones, called cells, which communicate with each other directly only through their nearest neighbors. Unlike CNN, each cell of MCNN has multiple vectors denoting different cell feature, and one vector will represent the meaning of that cell against other vectors when the network reaches the equilibrium state. Multi-objective cellular neural networks have some characteristics of CNN like: its continuous time feature allows real-time signal processing found wanting in the digital domain, its local interconnection feature makes it ideal for VLSI implementation and its multiple vector characteristic makes it applicable in many fields like image processing.