Sum-of-product neural network and its hardware design

Chau-Shing Wang, Chun Shin Lin

Research output: Contribution to conferencePaper

Abstract

This study investigates the possible hardware realization of the sum-of-product neural networks (SOPNN). A SOPNN is a memory-based neural network, whose output has a sum-of-product form. The neural network structure consists of several submodules, of which each receives inputs and generates a local output. The local outputs from several submodules are added up to generate an overall network output. In each submodule, each input variable is used to address a memory location. Outputs from several memory blocks in a submodule are multiplied together to form a local submodule output. To implement the structure using analog circuits, each digital input needs to be converted into an analog one by a D/A converter. The product of the outputs from the memory blocks can be computed using a cascade multiplier. These products are then added up. In this study, the possible hardware realization is investigated. Simulations using PSpice for the D/A converter, the multiplier, and the adder have been performed. An example with 4 inputs and 3 submodules has been simulated to test and evaluate the overall design. Results show that the performance of the circuit is good.

Original languageEnglish
Pages167-172
Number of pages6
Publication statusPublished - 1999 Dec 1
EventProceedings of the 1999 Artificial Neural Networks in Engineering Conference (ANNIE '99) - St. Louis, MO, USA
Duration: 1999 Nov 71999 Nov 10

Other

OtherProceedings of the 1999 Artificial Neural Networks in Engineering Conference (ANNIE '99)
CitySt. Louis, MO, USA
Period99-11-0799-11-10

Fingerprint

Computer systems programming
Computer architecture
Digital to analog conversion
Neural networks
Hardware
Data storage equipment
Computer simulation
Adders
Analog circuits
Analog integrated circuits
Networks (circuits)

All Science Journal Classification (ASJC) codes

  • Software

Cite this

Wang, C-S., & Lin, C. S. (1999). Sum-of-product neural network and its hardware design. 167-172. Paper presented at Proceedings of the 1999 Artificial Neural Networks in Engineering Conference (ANNIE '99), St. Louis, MO, USA, .
Wang, Chau-Shing ; Lin, Chun Shin. / Sum-of-product neural network and its hardware design. Paper presented at Proceedings of the 1999 Artificial Neural Networks in Engineering Conference (ANNIE '99), St. Louis, MO, USA, .6 p.
@conference{b1647b4eabf946c183de733eb55f609a,
title = "Sum-of-product neural network and its hardware design",
abstract = "This study investigates the possible hardware realization of the sum-of-product neural networks (SOPNN). A SOPNN is a memory-based neural network, whose output has a sum-of-product form. The neural network structure consists of several submodules, of which each receives inputs and generates a local output. The local outputs from several submodules are added up to generate an overall network output. In each submodule, each input variable is used to address a memory location. Outputs from several memory blocks in a submodule are multiplied together to form a local submodule output. To implement the structure using analog circuits, each digital input needs to be converted into an analog one by a D/A converter. The product of the outputs from the memory blocks can be computed using a cascade multiplier. These products are then added up. In this study, the possible hardware realization is investigated. Simulations using PSpice for the D/A converter, the multiplier, and the adder have been performed. An example with 4 inputs and 3 submodules has been simulated to test and evaluate the overall design. Results show that the performance of the circuit is good.",
author = "Chau-Shing Wang and Lin, {Chun Shin}",
year = "1999",
month = "12",
day = "1",
language = "English",
pages = "167--172",
note = "Proceedings of the 1999 Artificial Neural Networks in Engineering Conference (ANNIE '99) ; Conference date: 07-11-1999 Through 10-11-1999",

}

Wang, C-S & Lin, CS 1999, 'Sum-of-product neural network and its hardware design' Paper presented at Proceedings of the 1999 Artificial Neural Networks in Engineering Conference (ANNIE '99), St. Louis, MO, USA, 99-11-07 - 99-11-10, pp. 167-172.

Sum-of-product neural network and its hardware design. / Wang, Chau-Shing; Lin, Chun Shin.

1999. 167-172 Paper presented at Proceedings of the 1999 Artificial Neural Networks in Engineering Conference (ANNIE '99), St. Louis, MO, USA, .

Research output: Contribution to conferencePaper

TY - CONF

T1 - Sum-of-product neural network and its hardware design

AU - Wang, Chau-Shing

AU - Lin, Chun Shin

PY - 1999/12/1

Y1 - 1999/12/1

N2 - This study investigates the possible hardware realization of the sum-of-product neural networks (SOPNN). A SOPNN is a memory-based neural network, whose output has a sum-of-product form. The neural network structure consists of several submodules, of which each receives inputs and generates a local output. The local outputs from several submodules are added up to generate an overall network output. In each submodule, each input variable is used to address a memory location. Outputs from several memory blocks in a submodule are multiplied together to form a local submodule output. To implement the structure using analog circuits, each digital input needs to be converted into an analog one by a D/A converter. The product of the outputs from the memory blocks can be computed using a cascade multiplier. These products are then added up. In this study, the possible hardware realization is investigated. Simulations using PSpice for the D/A converter, the multiplier, and the adder have been performed. An example with 4 inputs and 3 submodules has been simulated to test and evaluate the overall design. Results show that the performance of the circuit is good.

AB - This study investigates the possible hardware realization of the sum-of-product neural networks (SOPNN). A SOPNN is a memory-based neural network, whose output has a sum-of-product form. The neural network structure consists of several submodules, of which each receives inputs and generates a local output. The local outputs from several submodules are added up to generate an overall network output. In each submodule, each input variable is used to address a memory location. Outputs from several memory blocks in a submodule are multiplied together to form a local submodule output. To implement the structure using analog circuits, each digital input needs to be converted into an analog one by a D/A converter. The product of the outputs from the memory blocks can be computed using a cascade multiplier. These products are then added up. In this study, the possible hardware realization is investigated. Simulations using PSpice for the D/A converter, the multiplier, and the adder have been performed. An example with 4 inputs and 3 submodules has been simulated to test and evaluate the overall design. Results show that the performance of the circuit is good.

UR - http://www.scopus.com/inward/record.url?scp=0033317932&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=0033317932&partnerID=8YFLogxK

M3 - Paper

AN - SCOPUS:0033317932

SP - 167

EP - 172

ER -

Wang C-S, Lin CS. Sum-of-product neural network and its hardware design. 1999. Paper presented at Proceedings of the 1999 Artificial Neural Networks in Engineering Conference (ANNIE '99), St. Louis, MO, USA, .