A tunable Gaussian/square function computation circuit for analog neural networks
- 1 March 1998
- journal article
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Circuits and Systems II: Analog and Digital Signal Processing
- Vol. 45 (3) , 441-446
- https://doi.org/10.1109/82.664259
Abstract
A Gaussian/square function computation circuit suitable for analog neural networks is proposed. It can realize Gaussian and square functions when operating in weak and strong inversion region, respectively. It is shown that the center, width, and peak amplitude of the dc transfer curve can be controlled separably. Measurement results on 3-μm CMOS fabricated chips confirm theoretical and simulation findingsKeywords
This publication has 14 references indexed in Scilit:
- A radial basis function neurocomputer implemented with analog VLSI circuitsPublished by Institute of Electrical and Electronics Engineers (IEEE) ,2003
- Analog CMOS integration and experimentation with an autoadaptive independent component analyzerIEEE Transactions on Circuits and Systems II: Analog and Digital Signal Processing, 1995
- A Gaussian synapse circuit for analog VLSI neural networksIEEE Transactions on Very Large Scale Integration (VLSI) Systems, 1994
- Simple precision bias circuit for medium-power amplifiersIEEE Journal of Solid-State Circuits, 1994
- Programmable analogue VLSI for radial basis function networksElectronics Letters, 1993
- Ratiometric temperature stable current referenceElectronics Letters, 1993
- Two quadrant analogue squarer circuit based on MOS square-law characteristicElectronics Letters, 1991
- A neural network which learns decision boundaries with nonlinear clusteringPublished by Institute of Electrical and Electronics Engineers (IEEE) ,1991
- A class of analog CMOS circuits based on the square-law characteristic of an MOS transistor in saturationIEEE Journal of Solid-State Circuits, 1987
- Direct-coupled MOS squaring circuitIEEE Journal of Solid-State Circuits, 1979