An All-Digital VLSI ANN

  • Brian White
  • Mohamed I. Elmasry

Abstract

In recent years, there has been a great deal of research activity in artificial neural networks (ANN) in the area of simulation and hardware implementations [1, 2, 3, 4, 5, 6]. Because of the special features offered by ANNs, such as the capability to learn from examples, adaptation, parallelism, fault tolerance and noise resistance, they have been applied to a number of real-world problems including image and speech processing [3,6, 7, 8, 9]. To enhance the impact of ANNs and broaden the area of applications, it is imperative that ANNs benefit from the state-of-the-art VLSI and ULSI implementation technologies. Because these technologies are basically a digital implementation medium, ANNs must be adapted to an all-digital implementation approach. To illustrate this thesis, the research work reported in this paper offers a practical example of adapting an ANN model to an all-digital VLSI implementation.

Keywords

Nitron 

Preview

Unable to display preview. Download preview PDF.

Unable to display preview. Download preview PDF.

References

  1. [1]
    Rumelhart, D.E. and McClelland, J.L., Parallel Distributed Processing Volume 1: Foundations, Cambridge: The MIT Press, 1986.Google Scholar
  2. [2]
    Lippmann, R.P., “An introduction to computing with neural nets,” IEEE ASSP Magazine, vol. 4, pp. 4–22, Apr. 1987.CrossRefGoogle Scholar
  3. [3.
    ] Hecht-Nielsen, R., “Neurocomputing: picking the human brain,” IEEE Spectrum, vol. 25, pp. 36–41, Mar. 1988.CrossRefGoogle Scholar
  4. [4]
    IEEE Trans. Neural Networks, 1990.Google Scholar
  5. [5]
    Atlas, L.E. and Suzuki, Y., “Digital systems for artificial neural networks,” IEEE Circuits and Devices Magazine, pp. 20–24, Nov. 1989.Google Scholar
  6. [6]
    Treleaven, P., et al., “VLSI architectures for neural networks,” IEEE MICRO Magazine, pp. 8–27, Dec. 1989.Google Scholar
  7. [7]
    Proc. IEEE IJCNN, San Diego, 1990.Google Scholar
  8. [8]
    Howard, R.E., et al., “Optical character recognition: a technology driver for neural networks,” Proc. IEEE ISCAS, 1990, pp. 2433–2436.Google Scholar
  9. [9]
    Jackel, L.D., et al, “Hardware requirements for neural-net optical character recognition,” Proc. IEEE IJCNN, San Diego, 1990, vol. II, pp. 855–861.Google Scholar
  10. [10]
    Hammerstrom, D., “A VLSI architecture for high-performance, low-cost, on-chip learning,” Proc. IEEE IJCNN, San Diego, 1990, vol. II, pp. 537–544.Google Scholar
  11. [11]
    Tomlinson, M.S., et al., “A digital neural network architecture for VLSI,” Proc. IEEE IJCNN, San Diego, 1990, vol. II, pp. 545–550.Google Scholar
  12. [12]
    Gamal, A El., et al, “An architecture for electrically configurable gate arrays,” IEEE Journal of Solid-State Circuits, vol. 24, no. 2, pp. 394–398, Apr. 1989.CrossRefGoogle Scholar
  13. [13]
    Fukushima, K., “Neocognitron: a self-organizing neural network model for a mechanism of pattern recognition unaffected by shift in position,” Biol. Cybernetics, vol. 36, pp. 193–202, 1980.MathSciNetMATHCrossRefGoogle Scholar
  14. [14]
    Fukushima, K. and Miyake, S., “Neocognitron: a new algorithm for pattern recognition tolerant of deformations and shifts in position,” Pattern Recognition, vol. 15, no. 6, pp. 455–469, 1982.CrossRefGoogle Scholar
  15. [15]
    Fukushima, K., “Analysis of the process of visual pattern recognition by the neocognitron,” Neural Networks, vol. 2, pp. 413–420, 1989.CrossRefGoogle Scholar
  16. [16]
    Hammerstrom, D. and Means, E. “System design for a second generation neurocomputer,” Proc. IEEE IJCNN, Jan, 1990, vol. II, pp. 80–83.Google Scholar
  17. [17]
    Brown, H.K., et al., “Orthogonal extraction training algorithm,” Proc. IEEE IJCNN, Jan, 1990, vol. I, pp. 537–540.Google Scholar
  18. [18]
    Wilson, C.L., et al, “Self-organizing neural network character recognition on a massively parallel computer,” Proc. IEEE IJCNN, San Diego, 1990, vol. II, pp. 325–329.Google Scholar
  19. [19]
    Fukushima, K., Miyake, S. and Ito, T., “Neocognitron: a neural network model for a mechanism of visual pattern recognition,” IEEE Trans. Systems, Man, and Cybernetics, vol. SMC-13, no. 5, pp. 826–834, Sep. 1983.CrossRefGoogle Scholar
  20. [20]
    Lee, Y., et al, “Hangul recognition using neocognitron,” Proc. IEEE IJCNN, Jan, 1990, vol. I, pp. 416–419.Google Scholar
  21. [21]
    Wang, S. and Pan, C. “A neural network approach for Chinese character recognition,” Proc. IEEE IJCNN, San Diego, 1990, vol. II, pp. 917–923.Google Scholar
  22. [22]
    Barnard, E. and Casasent, D. “Shift invariance and the neocognitron,” Neural Networks, vol. 3, pp. 403–410, 1990.CrossRefGoogle Scholar
  23. [23]
    Johnson, K., et al, “Feature extraction in the neocognitron,” Proc IEEE IJCNN, 1988, vol. II, pp. 117–126.Google Scholar
  24. [24]
    White, B.A. and Elmasry, M.I., “Digi-Neocognitron: a Neocognitron neural network model for VLSI”, Report No. UW/VLSI 91–01, VLSI Group, University of Waterloo, Waterloo, Ontario, Jan, 1991.Google Scholar
  25. [25]
    White, B.A. and Elmasry, M.I. “VLSI architecture of the DigiNeocognitron neural network,” to be published.Google Scholar
  26. [26]
    Lim, Y.C., et al, “VLSI circuits for decomposing binary integers into signed power-of-two terms,” Proc. IEEE ISCAS, 1990, pp. 2304–2307.Google Scholar
  27. [27]
    Marchesi, M., et al, “Multi-layer perceptrons with discrete weights,” Proc. IEEE IJCNN, San Diego, 1990, vol. II, pp. 623–630.Google Scholar
  28. [28]
    Xu, Q., et al, “A fault tolerance analysis of a neocognitron model,” Proc. IEEE IJCNN, Jan, 1990, vol. II, pp. 559–562.Google Scholar
  29. [29]
    Weste, N.H.E. and Eshraghian, K., Principles of CMOS VLSI Design: A Systems Perspective, p. 344. Reading: Addison-Wesley Publishing Company, 1985.Google Scholar
  30. [30]
    ACTEL Corporation, “ACT 1 Family Gate Arrays Design Reference Manual”, 1989.Google Scholar
  31. [31]
    Texas Instruments, “TGC100 Series 1-micron CMOS Gate Arrays”, May, 1990.Google Scholar

Copyright information

© Springer Science+Business Media New York 1994

Authors and Affiliations

  • Brian White
  • Mohamed I. Elmasry

There are no affiliations available

Personalised recommendations