Lo Chua

Analog signal processing using cellular neural networks

The cellular neural network (CNN) is an example of very-large-scale analog processing or collective analog computation. The CNN architecture combines some features of fully connected analog neural networks with the nearest-neighbor interactions found in cellular automata. These networks have numerous advantages both for simulation and for VLSI implementation and can perform (though are not limited to) several important image processing functions. The important features which enable the CNN architecture to perform signal processing functions using standard VLSI technology are discussed. Circuit characteristics are outlined, and examples of cellular neural network signal processing are presented. Connected segment extraction is illustrated by examples, as is histogramming using a two-layer CNN.<<ETX>>

CNN universal chips crank up the computing power

New cellular neural network chips, with stored-program capability and analog-and-logic architecture, are poised to challenge all-digital processing. In this article, we highlight the key ideas leading to the CNN Universal Machine, using simple circuit interpretations. We also illustrate the system, software, and application aspects.

Some novel capabilities of CNN: game of life and examples of multipath algorithms

The authors show that the game of life algorithm, which is equivalent to a Turing machine, can be realized by cellular neural networks (CNNs). They also present a multipath CNN algorithm that demonstrates the capabilities of analog/logic (analogic) software. Complex image processing tasks can be realized by programmable dual computing CNNs. A specific example of blob-counting in a gray-scale image is presented.<<ETX>>

The CNN universal machine. I. The architecture

Various types of cellular neural networks (CNNs) are summarized, and a taxonomy of CNNs is given according to the different types of grids, processors, interactions, and modes of operation. The CNN universal machine is introduced. The architecture and the key features of the CNN universal machine are outlined. An exhaustive list of references is given.<<ETX>>

Programmable analog VLSI CNN chip with local digital logic

A new integrated circuit cellular neural network (CNN) implementation having digitally or continuously selectable template coefficients is presented. Local logic and memory are added into each cell providing a simple dual (analog and digital) computing structure. The variable-gain operational transconductance amplifiers (OTAs) are used as the voltage controlled current sources to program the template element values. This analog array processor can be applied to solve problems with a sequence of different templates. A 4-by-4 CNN circuit is realized using the 2- mu m analog CMOS process. To test the circuit a general purpose control system was designed with the microcomputer interface.<<ETX>>

The CNN universal machine. II. Programmability and applications

For pt.I, see ibid., p.1-10 (1992). The programmability (as a stored program) of the CNN universal machine is discussed. It is shown why and in which sense this machine is universal. The analogic type of algorithm is introduced. The application potential is reviewed and the biological relevance is analyzed. It is shown that the architecture is optimal not only for silicon implementations, but also for many biological information processing organs that have the same structure.<<ETX>>

The structure of continuous piecewise-linear functions

Continuous piecewise-linear functions from R/sup n/ to R/sup m/ are characterized by the geometry of their region boundaries. The notions of the nonlinearity component and of the order of boundary intersections are introduced. The theory developed makes it possible to determine the exact number of independent parameters for a given class of setups.<<ETX>>