Abstract

We examine theoretically and experimentally the properties of two electrooptic devices based on quantum confined Stark effect of excitonic transitions. These two devices are shown to be critical in implementation of the Hopfield model as well as other neural type networks for associative memories. P–i(multiquantum wells)–n structures using GaAs/AlGaAs provide a controller–modulator device which has integrating–thresholding properties required of neurons. The p–i–n structures also provide programmable modulators which can serve as a synaptic mask. Using Monte Carlo techniques we examine an all-optical architecture to implement the Hopfield network. No external feedback-thresholding circuitry is required in this implementation due to special design of the controller–modulator device. Speed and stability issues of this architecture are also addressed. The computer simulation results provide valuable insight into how the controller–modulator device should be improved for better network implementation.

© 1988 Optical Society of America

Holographic implementation of a Hopfield model with discrete weightings

H. J. White and W. A. Wright
Appl. Opt. 27(2) 331-338 (1988)

Electronic hardware implementations of neural networks

A. P. Thakoor, A. Moopenn, John Lambe, and S. K. Khanna
Appl. Opt. 26(23) 5085-5092 (1987)

Unipolar terminal-attractor-based neural associative memory with adaptive threshold and perfect convergence

Chwan-Hwa Wu and Hua-Kuang Liu
Appl. Opt. 33(11) 2210-2217 (1994)

Optical neural network with bipolar neural states

Xu-Ming Wang and Guo-Guang Mu
Appl. Opt. 31(23) 4712-4719 (1992)

Optical production systems using neural networks and symbolic substitution

Elizabeth Botha, David Casasent, and Etienne Barnard
Appl. Opt. 27(24) 5185-5193 (1988)