Implementation considerations of symbolic substitution in digital optical computing

摘要

An important area of current focus in optical computing research is the technique of Symbolic Substitution (SS), proposed by Huang in 1983. The technique is successful by virtue of its parallel signal processing ability. Coupled with modified signed digit (MSD) arithmetic and/or high-radix arithmetic, several researchers have been able to achieve high speed of parallel operations in digital computing. Although dual-rail logic for SS avoided the troublesome need for recognition of both bright and dark pixels, with the introduction of MSD, the dual recognition problem reappeared. A triple-rail logic which is compatible with both MSD and optical computing requirements of SS has been suggested. Besides explaining the scheme of triple-rail coding for MSD, techniques for its implementation in digital optical computing are presented and discussed. Possible extension of the scheme to high-radix arithmetic is indicated.