Sensitivity study and improvements on a nonlinear resistive-type neuron circuit

Author

Djahanshahi, H. ; Ahmadi, M. ; Jullien, G.A. ; Miller, W.C.

Author_Institution

Dept. of Electr. & Comput. Eng., Windsor Univ., Ont., Canada

Volume

147

Issue

4

fYear

2000

fDate

8/1/2000 12:00:00 AM

Firstpage

237

Lastpage

242

Abstract

A generalised VLSI circuit realisation for a nonlinear active resistor-type neuron is proposed that implements a saturating sigmoidal-like function by combining the nonlinear characteristics of NMOS and PMOS transistors. The circuit design is based on using a parameter sensitivity analysis to develop a robust design that will be relatively insensitive to process-parameter variations over the area of the die. The nonlinear resistor has been integrated into a module which realises a programmable digital synaptic weight capability. A neuron is effectively formed from the parallel interconnection that takes place as multiplier outputs are connected to create an input node to the resultant distributed neuron. Designs in 0.35 and 0.8 μm processes are compared with a conservative 1.2 μm CMOS implementation

Keywords

CMOS integrated circuits; VLSI; integrated circuit design; mixed analogue-digital integrated circuits; neural chips; programmable circuits; sensitivity analysis; 0.35 micron; 0.8 micron; 1.2 micron; CMOS implementation; NMOS transistors; PMOS transistors; generalised VLSI circuit realisation; multiplier outputs connection; nonlinear active resistor-type neuron; nonlinear resistive-type neuron circuit; parallel interconnection; parameter sensitivity analysis; process-parameter variations insensitivity; programmable digital synaptic weight capability; robust circuit design; saturating sigmoidal-like function; sensitivity study;

fLanguage

English

Journal_Title

Circuits, Devices and Systems, IEE Proceedings -

Publisher

iet

ISSN

1350-2409

Type

jour

DOI

10.1049/ip-cds:20000237

Filename

861416