Programmable Optical Quadratic Neural Networks

Abstract

This report details the results of the first year of analytical and experimental investigations of programmable optical quadratic neural networks. The investigations have included: (1) computer simulations and theoretical characterizations of the performances of first and second order Hopfield associative memories in terms of a signal-to-noise ratio parameter C; (2) a hybrid electro-optical, polarization-encoding-based technique for implementing a quadratic neural processor and (3) use of photorefractive BaTiO3 crystals to perform a vector-matrix-vector operation based on four-wave mixing. Details are summarized in this report and in the publications resulting from the research effort.

Open PDF

Document Details

Document Type
Technical Report
Publication Date
Dec 31, 1988
Accession Number
ADA204160

Entities

People

  • John F. Walkup
  • Thomas F. Krile

Organizations

  • Texas Tech University

Tags

Communities of Interest

  • Energy and Power Technologies
  • Human Systems

DTIC Thesaurus Topics

  • Argon Lasers
  • Character Recognition
  • Coding
  • Computer Simulations
  • Computers
  • Content Addressable Memory
  • Electrical Engineering
  • Engineering
  • Light Sources
  • Neural Networks
  • Notation
  • Optics
  • Pattern Recognition
  • Polarization
  • Signal Processing
  • Simulations
  • Wave Mixing

Readers

  • Neural Network Machine Learning.
  • Optical Physics and Photonics.
  • Parallel and Distributed Computing.

Technology Areas

  • AI & ML
  • AI & ML - Neural Networks