Abstract

Threshold (or relative magnitude) search is traditionally performed iteratively in a bit-serial manner in optical data-base/knowledge-base machines, which results in an execution time proportional to the operand size. We present a single-step threshold search algorithm and its optical implementation. The proposed algorithm performs magnitude comparison in constant time, independent of the operand size, and consequently it greatly increases the performance of optical data-base/knowledge-base processing operations such as searching, selection, retrieving, and sorting.

© 1993 Optical Society of America

Optical implementation of a constant-time multicomparand bit-parallel magnitude-comparison algorithm using wavelength- and polarization-division multiplexing with application to parallel database processing

Abram Detofsky, Peng Yin Choo, and Ahmed Louri
Opt. Lett. 23(17) 1372-1374 (1998)

Optical content-addressable parallel processor for high-speed database processing

Ahmed Louri and James A. Hatch
Appl. Opt. 33(35) 8153-8163 (1994)

Multiwavelength optical content-addressable parallel processor for high-speed parallel relational database processing

Peng Yin Choo, Abram Detofsky, and Ahmed Louri
Appl. Opt. 38(26) 5594-5604 (1999)

Iterative processing on a hybrid optical parallel array logic system with a selectable coherent correlator

Daisuke Miyazaki, Sunao Kakizaki, Tsuyoshi Konishi, Jun Tanida, and Yoshiki Ichioka
Appl. Opt. 32(17) 3053-3058 (1993)

Constant-time parallel sorting algorithm and its optical implementation using smart pixels

Ahmed Louri, James A. Hatch, and Jongwhoa Na
Appl. Opt. 34(17) 3087-3096 (1995)