Adaptive Optical Linear Algebra Processors
Abstract
The final report on research into adaptive optical linear algebra processors includes 3 processors. These include: a space integrating frequency- multiplexed processor, a hybrid space and time integrating processor, and a heterodyned linear analog processor. We also address number representation work using twos complement and negative base representation, plus fundamental new concepts such as matrix and bit partitioning. The applications addressed for these various systems include: the solution of linear and nonlinear algebraic equations, partial differential equations, computational fluid dynamics, finite element problems, adaptive signal processing (including adaptive phased array radar processing), and the potential for use in various adaptive neural etc. processing. We also include data to demonstrate and quantify that a high- accuracy digital multiplication by analog convolution optical architecture is superior to a digital processor.
Document Details
- Document Type
- Technical Report
- Publication Date
- Nov 15, 1988
- Accession Number
- ADA202996
Entities
People
- B. V. Kumar
- David P. Casasent
Organizations
- Carnegie Mellon University