Excavating
Patent
1991-06-25
1994-10-18
Canney, Vincent P.
Excavating
371 36, G06F 1100
Patent
active
053575288
ABSTRACT:
The logical comparison and arithmetic addition functions are optimally constructed in depth-2 threshold logic circuits employing majority elements arranged into structures corresponding to sparse delta polynomials. A delta polynomial is a polynomial having a relatively large value for a particular set of variable values and having a relatively small value for all other sets of variable values. A delta polynomial can be constructed through a column-by-column consideration of an error-correcting code generator matrix. The sparseness of a delta polynomial constructed in this manner means that the delta polynomial expression for implementing a threshold logic circuig which combines n-bit numbers contains no more than n.sup.c terms. A further benefit of such a delta polynomial is the low values of its coefficients.
REFERENCES:
patent: 3439328 (1969-04-01), Winder
patent: 3529141 (1970-09-01), Reed
patent: 3562502 (1971-02-01), Kautz
patent: 3609329 (1971-09-01), Martin
patent: 3780276 (1973-12-01), Carter et al.
patent: 3780312 (1973-12-01), Lin et al.
patent: 3900742 (1975-08-01), Hampel et al.
patent: 4964126 (1990-10-01), Musicus et al.
IBM Research Report On The Dynamic Range of Linear Threshold Elements, Siu & Bruck, Jan. 4, 1990, pp. 1-15.
1990 Society for Industrial and Applied Mathematics, "Harmonic Analysis of Polynomial Threshold Functions", J. Bruck, pp. 168-177.
"Threshold Circuits of Bounded Depth", Proceedings IEEE 28th FOCS, 1987, pp. 99-110, Hajnal et al.
"The CCD Neural Processor: A Neural Network Integrated Circuit with 65536 Programmable Analog Synapses" A. J. Agranat, et al. 1990 IEEE, pp. 1073-1075.
An Electrically Trainable Artifical Neural Network with 10240 "Floating Gate" Synapses, Holler et al.
Digital-Analog Hybrid Synapse Chips for Electronic Neural Networks Moopenn et al, pp. 769-776, Advances in Neural Information Processing Systems 2.
"Pulse-Firing Neural Chips for Hundreds of Neurons" Brownlow et al, pp. 785-792, Advances in Neural Information Processing Systems 2.
"A Purely Capacitive Synaptic Matrix for Fixed-Weight Neural Networks", IEEE, vol. 38, No. 2, Feb. 1991, Cilingiroglu, pp. 210-217.
"Artificial Neural Networks Using MOS Analog Multipliers" Hollis et al, IEEE, vol. 25, No. 3, Jun. 1990, pp. 849-855.
Alon Noga
Bruck Jehoshua
Canney Vincent P.
International Business Machines - Corporation
LandOfFree
Depth-2 threshold logic circuits for logic and arithmetic functi does not yet have a rating. At this time, there are no reviews or comments for this patent.
If you have personal experience with Depth-2 threshold logic circuits for logic and arithmetic functi, we encourage you to share that experience with our LandOfFree.com community. Your opinion is very important and Depth-2 threshold logic circuits for logic and arithmetic functi will most certainly appreciate the feedback.
Profile ID: LFUS-PAI-O-2378249