CURRENT-MODE TECHNIQUES FOR THE IMPLEMENTATION OF CONTINUOUS-TIME ANDDISCRETE-TIME CELLULAR NEURAL NETWORKS

This paper presents a unified, comprehensive approach to the design of continuous-time (CT) and discrete-time (DT) cellular neural networks (CNN) using CMOS current-mode analog techniques. The net input signals are currents instead of voltages as presented in previous approaches, thus avoiding the need for current-to-voltage dedicated interfaces in image processing tasks with photosensor devices. Outputs may be eithe r currents or voltages. Cell design relies on exploitation of current mirror properties for the efficient implementation of both linear and nonlinear analog operators. These cells are simpler and easier to desi gn than those found in previously reported CT and DT-CNN devices. Basi c design issues are covered, together with discussions on the influenc e of nonidealities and advanced circuit design issues as well as desig n for manufacturability considerations associated with statistical ana lysis. Three prototypes have been designed for 1.6-mum n-well CMOS tec hnologies. One is discrete-time and can be reconfigured via local logi c for noise removal, feature extraction (borders and edges), shadow de tection, hole filling, and connected component detection (CCD) on a re ctangular grid with unity neighborhood radius. The other two prototype s are continuous-time and fixed template: one for CCD and other for no ise removal. Experimental results are given illustrating performance o f these prototypes.