Henry John Caulfield

Eigenvector determination by noncoherent optical methods

An iterative method for finding the eigenvectors and eigenvalues of a matrix via incoherent optical matrix-vector multiplication and simple electronic feedback is described.

Holography for the new millennium

I. Overview: Where Holography is Going.- II. Display Holography: 1. Color Holograms. 2. Real-Time Autostereographic Three-Dimensional Displays.- III. Special Modalities and Methods: 1. Edge-Lit Holograms. 2. Subwavelength Diffractive Optical Elements. 3. Coherence Gated Holograms. 4. Sculpturing of Three-Dimensional Light Fields by Iterative Optimization.- IV. Applications: 1. Solar Holography. 2. Holographic Optical Memories. 3. Holography and Speckle Techniques Applied to Nondestructive Measurement and Testing. 4. Diffuser Display Screen. 5. Holographic Nonspatial Filtering for Laser Beams. 6. Particle Holograms. 7. Holographic Antireflection Coatings.- V. Physics and Holography: 1. Holography and Relativity. 2. Quantum Holograms.

Pipelined polynomial processors implemented with integrated optical components

Optical systolic pipeline processors for polynomial evaluation can be built using Horner’s rule. With integrated optics techniques it will be possible to fabricate large-order pipelines operating at very high speeds.

Feedback methods for optical systolic and engagement matrix processors

The matching of the feedback circuitry to the optical systolic or engagement processor permits simple pipelining of stationary iterative algorithms as well as on-the-fly scale adjustment similar in effect to floating-point calculation.

Suggested Integrated Optical Implementation Of Pipelined Polynomial Processors

Optical systolic pipeline processors for polynomial evaluation can be built using Horners rule. With integrated optics techniques, it will be possible to fabricate large order pipelines operating at very high speeds.

Efficient real number representation with arbitrary radix

Because most optical digital computers use only non-negative quantities, it is of great interest to find an efficient way to represent real numbers. For radix 2 (binary) numbers the twos complement method requires only one extra digit beyond that needed for non-negative numbers. We introduce here an arbitrary radix generalization.

Algorithm improvements for optical eigenfunction computers.

Prior iterative approaches to optical eigenfunction solution have at least three major problems: slow convergence (sometimes); decreasing accuracy after the first solution; and imperfect parallel renormalization (leading to poor use of system dynamic range and hence poor accuracy). We introduce new approaches and algorithms to solve these problems. The new algorithms lead to a tight error bound on eigenvalues and an automatic handling of degenerate or near degenerate eigenvalues. Applications are discussed.

Robust Optical Systolic Long-Code Processor

A new architecture for performing systolic optical recognitiori of long codes is introduced. The resulting system is quite robust in that it is insensitive to time of arrival of the code word and to errors in recognizing parts of the code. The advantages, besides robustness, of this system are the use of simple, available components (independent of code word length), high speed, and the small number of components required.

Optical Algebraic Processing Architectures And Algorithms

A new generation of opto-electronic signal processors has been developing during the past several years. These processors are designed to perform algebraic operations like matrix-vector and matrix-matrix multiplication. In this paper key architectural developments are reviewed and major algorithmic methods and problems are discussed.

Optical Singular Value Decomposition For The Ax = b Problem

Optical approaches to solving the Ax = b problem have suffered from four difficulties: (1) an inability to handle the problem for nonsquare A, (2) the necessity of insuring convergence for nonsingular A, (3) the inability to handle a singular A, and (4) inaccuracies due to an ill conditioned A. We show that these problems can all be solved or mitigated by singular value decomposition (SVD). An accurate approach to optical SVD is shown.