Advertisement

Current Mode Circuits for Programmable WTA Neural Network

  • Krzysztof Wawryn
  • Bogdan Strzeszewski
Article

Abstract

In this article prototype AB class programmable synaptic connection and WTAcircuits that set foundations for low power VLSI neural networks and other applications areproposed. The analysis of the circuits is given. A qualitative comparison of currentconsumption is made between standard A class and proposed AB class circuits. Layouts ofthe AB class transconductance programmable synaptic connection and 2-WTA circuits havebeen designed and then the prototype CMOS circuits have been manufactured and measured.Measured characteristics have been compared to simulated ones.

artificial neural network neuron cell winner take all algorithm current mode technique 

Preview

Unable to display preview. Download preview PDF.

Unable to display preview. Download preview PDF.

References

  1. 1.
    Mead, C. and Ismail, M., Analog Implementation of Neural Systems. Kluwer Academic Publishers, 1989.Google Scholar
  2. 2.
    Zurada, J., Introduction to Artificial Neural Systems. West Publishing Company, 1992.Google Scholar
  3. 3.
    He, Y. and Sanchez-Sinencio, E., "Min-net winner-take-all CMOS implementation." Electron. Lett. 29, pp. 1237–1239, 1993.CrossRefGoogle Scholar
  4. 4.
    Choi, J. and Sheu, B., "A high-precision VLSI winner-take-all circuit for self-organising neural networks." IEEE J. Solid-State Circuits 28, pp. 576–584, 1993.CrossRefGoogle Scholar
  5. 5.
    Sasaki, M., et al., "Fuzzy multiple input maximum and minimum circuit in current mode and their analyses using boundeddifference equations." IEEE Trans. Computers 39, pp. 443–450, 1993.Google Scholar
  6. 6.
    Pedroni, V. A., "Inhibitory mechanism analysis of complexity O(N) MOS winner-take-all networks." IEEE Trans. Circuits Syst. Part I, CAS-42, pp. 172–175, 1995.CrossRefGoogle Scholar
  7. 7.
    Smedley, S., Taylor, J. and Wilby, M., "A scalable high-speed current-mode winner-take-all network for VLSI neural applications." IEEE Trans. Circuits and Systems Part I, CAS-42, pp. 289–291, 1995.CrossRefGoogle Scholar
  8. 8.
    Wawryn, K. and Strzeszewski, B., "Programmable low power VLSI current mode neuron," in Proc. Int. Conf. Electronics, Circuits and Systems Rodos 1, pp. 406–409, 1996.Google Scholar
  9. 9.
    Kohonen, T., "An introduction to neural computing." Neural Networks 1, pp. 3–16, 1988.CrossRefGoogle Scholar
  10. 10.
    Wawryn, K. and Strzeszewski, B. "Low power VLSI neuron cells for artificial neural networks," in Proc. Int. Symp. Circuits and Systems, Atlanta 3, pp. 372–375, 1996.Google Scholar
  11. 11.
    11. Wawryn, K. and Strzeszewski, B. "Low power current mode approach to winner take all circuit." in Proc. Int. Symp. Circuits and Systems Monterey 1, pp. 393–396, 1998.Google Scholar
  12. 12.
    Holler, M., Tam, S., Castro, H. and Benson, R., "An electrically trainable artificial neural network (ETANN) with 1024 'float gate' synapses," in Proc. Int. Joint Conf. Neural Networks 2, pp. 191–196, 1989.CrossRefGoogle Scholar
  13. 13.
    Lee, B. W., Sheu, B. J. and Yang, H., "Analog floating gate synapses for general purpose VLSI neural computation." IEEE Trans. Circuits and Syst. CAS-38, pp. 654–658, 1991.CrossRefGoogle Scholar
  14. 14.
    Schneider, C. and Card, H., "Analog CMOS synaptic learning circuits adapted from invertebrate biology." IEEE Trans. Circuits and Syst. CAS-38, pp. 1430–1438, 1991.CrossRefGoogle Scholar
  15. 15.
    Betta, G. F. D., Graffi, S., Kovacs, Zs. M. and Masetti, G., "CMOS implementation of an analogically programmable cellular neural network." IEEE Trans. Circuits and Syst. Part II, CAS-40, pp. 206–215, 1993.CrossRefGoogle Scholar
  16. 16.
    Wawryn K., "AB class current mode multipliers for programmable neural networks." Electron. Lett. 32(20), pp. 1902–1904, 1996.CrossRefGoogle Scholar
  17. 17.
    Chua, L. O. and Yang, L. "Cellular neural networks: Theory." IEEE Trans. Circuits and Systems CAS-35, pp. 1257–1274, 1988.CrossRefMathSciNetGoogle Scholar
  18. 18.
    Rodriguez-Vazquez, A., Espejo, S., Dominguez-Castro, R., Huertas, J. L. and Sanchez-Sinencio, E., "Current-mode techniques for the implementation of continuous-and discrete-time cellular neural networks." IEEE Trans. Circuits and Systems CAS-40, pp. 132–146, 1993.Google Scholar
  19. 19.
    Sedra, A., S. and Roberts, G. W., "Current conveyor theory and practice." in Analogue IC Design: The Current Mode Approach, C. Toumazou, F. J. Lidgey and D. G. Haigh, eds., Peter Peregrinus Ltd., 1990.Google Scholar
  20. 20.
    Demosthenenous, A., Akbari-Dilmaghani, R., Smedley, S. and Taylor, J., "Enhanced modular CMOS current-mode winnertake-all network," in Proc. Int. Conf. Elctronics, Circuits and Systems, Rodos 1, pp. 402–405, 1996.Google Scholar

Copyright information

© Kluwer Academic Publishers 2001

Authors and Affiliations

  • Krzysztof Wawryn
    • 1
  • Bogdan Strzeszewski
    • 1
  1. 1.Department of ElectronicsTechnical University of KoszalinKoszalinPoland

Personalised recommendations