Abstract
A CMOS Gaussian/Triangular Basis functions computation circuit suitable for analog neural networks is proposed. The circuit can be configured to realize any of the two functions. The circuit can approximate these functions with relative root-mean-square error less than 1%. It is shown that the center, width, and peak amplitude of the dc transfer characteristic can be independently controlled. SPICE simulation results using 0.18 μ m CMOS process model parameters of TSMC18 technology are included.
This is a preview of subscription content, log in via an institution to check access.
Similar content being viewed by others
References
P.D. Wasserman, Advanced Methods in Neural Computing, New York: Van Nostrand Reinhold, 1993.
C. Mead, Analog VLSI and Neural Systems, Addison-Wesley, Reading, PA: 1989.
A.F. Murray, “Silicon implementations of neural networks,” IEE Proceeding of Radar and Signal Processing, vol. 138, no. 1, pp. 3–12, Feb. 1991.
B.J.Sheu and J. Choi, Neural Information Processing and VLSI, Kluwer Academic Publicshers, Boston, 1994.
A. F. Murray and Robin Woodburn, “The Prospects for Analogue Neural VLSI,” International Journal of Neural Systems, vol. 8, no. 5 & 6, pp. 559–579, 1997.
V.F. Koosh and R.M. Goodman, “Analog VLSI neural network with digital perturbative learning,” ISCAS II : Analog and Digital Signal Processing, vol. 49, no. 5, pp. 359–368, 2002.
E. Sanchez-Sinencio, J. Ramirez-Angulo, B. Linares-Barranco, and A. Rodriguez-Vazquez, “Operational transconductance amplifier-based nonlinear function synthesis,” IEEE Jorn. of Solid-State Circuits, vol. 24, pp. 1576–1586, 1989.
J. Ramirez-Angulo, E. Sanchez-Sinencio, and A. Rodriguez-Vazquez, “A Piecewise-linear function approximation using current mode circuits,” ISCAS'92, vol. 4, pp. 2025–2028, 1992.
A. Rodriguez-Vazquez and M. Delgado-Restituto, “CMOS design of chaotic oscillators using state variables: A monolithic Chua's circuit,” ISCAS'93, vol. 40, no. 10, pp. 596–613, 1993.
N. Manaresi, R. Rovatti, E. Franchi, and G. Baccarani, “A current-mode Piecewise-linear function approximation circuit based on Fuzzy-logic,” ISCAS'98, vol. 3, pp. 127–130, 1998.
T. Serrano-Gotarredona, “Current-mode fully-programmable piece-wise-linear block for neuro-fuzzy applications,” Elec. Letters, vol. 38, no. 20, pp. 1165–1166, 2002.
M. Balsi, “Current-mode programmable piecewise-linear neural synapses,” Int. J. of Circ. Theory and Appl. vol. 31, pp. 265–275, 2003.
T. Delbück, “‘Bump’ circuits for computing similarity and dissimilarity of analog voltage,” in Proc. Int. Neural Network Society, Seattle, WA, 1991.
S.S. Watkins and P.M. Chau, “A radial basis function neurocomputer implemented with analog VLSI circuits,” in Int. Joint Conf. Neural Networks, vol. 2, pp. 607–612, 1992.
S. Churcher, A.F. Murray, and H.M. Reekie, “Programmable analogue VLSI for radial basis function networks,” Electron. Lett., vol. 29, no. 18, pp. 1603–1605, Sept. 1994.
J. Choi, B.J. Sheu, and J.C.-F. Chang, “A Gaussian synapse circuit for analog neural networks,” IEEE Trans. VLSI Syst., vol. 2, pp. 129–133, March. 1994.
S.T. Lee and K.T. Lau, “An Analog Gaussian Synapse For Artificial Neural Networks,” Proceedings of the 38th Midwest Symposium on Circuits and Systems, vol. 1. pp. 77–80, Aug. 1995.
R. Kuppuswamy, L. Theogarajan, and L.A. Akers, “A programmable Gaussian node,” IEEE International Conference on Neural Networks, vol. 2, pp. 880–885, June 1996.
S.Y. Lin, R.-J. Huang, and T.-D. Chiueh “A tunable gaussian/square function computation circuit for analog neural networks,” IEEE Transactions on Circuits and Systems-II, vol. 45, no. 3, pp. 441–446, Mar. 1998.
M. Samet, M. Masmoudi, F. Ghozzi, Y. B. Ayed, and A.M. Alimi, “A Beta neuron in CMOS subthreshold mode,” ICM '98. Proceedings of the Tenth International Conference on Microelectronics, pp. 290–293, Dec. 1998.
Munir Samet, “A Beta basis function neural network in CMOS subthreshold mode,” Int. J. Electronics, vol. 88, no. 6, pp. 645–657, 2001.
D.S. Masmoudi, A.T. Dieng, and M. Masmoudi, “A Subthreshold Mode Programmable Implementation of the Gaussian Function for RBF Neural Networks Applications,” in Proceedings of the 2002 IEEE Internatinal Symposium on Intelligent Control, pp. 454–459, Oct. 2002.
S. Satyanarayana, Y.P. Tsividis, and H.P. Graf, A reconfigurable VLSI neural network, Journal of Solid-State Circuits, vol. 27, pp. 67–81, 1992.
J. Platt, “A resource-allocating neural network for function interpolation,” Neural Computation, vol. 3, pp. 213–225, 1991.
P.R. Gray, J. Hurst, S.H. Lewis, R.G. Meyer, Analysis and Design of Analog Integrated Circuits, 4th edition, John wiley & Sons, New York 2001.
R.J. Baker, H.W. Li, and D.E. Boyce, “CMOS circuit design, Layout and simulation,” IEEE Press, New York, 1998.
Author information
Authors and Affiliations
Additional information
Muhammad Taher Abuelma'Atti was born in Cairo, Egypt, in 1942. He received the B.Sc. degree in Electrical Engineering in 1963 from the University of Cairo, Cairo, Egypt, the Ph.D. degree in 1979 and the Doctor of Science degree in 1999 both from the University of Bradford, Bradford, England.
From 1963 to 1967, he worked at the Military Technical College in Cairo as a Teaching Assistant. He was with the Iron and Steel Company in Helwan, Cairo, from 1967 to 1973 as a Senior Electrical Engineer. From 1973 to 1976 he was with the College of Engineering, University of Riyadh, Saudi Arabia, as a Teaching Assistant. From 1980 to 1981, he worked with the Faculty of Engineering, University of Khartoum, Sudan, as an Assistant Professor, and from 1981 to 1982 he was with the College of Engineering, King Saud University, Riyadh, Saudi Arabia, as an Assistant Professor. In 1982 he joined the College of Engineering, University of Bahrain and in 1987 he became an Associate Professor. In 1991 he joined the College of Engineering Sciences, King Fahd University of Petroleum and Minerals, Dhahran, Saudi Arabia, where he became a Full Professor in January 1995.
Dr. Abuelma'Atti is the recipient of the 1994/1995 Excellence in Teaching Award and the 1995/1996 and 2000/2001 Excellence in Research Award. Both at King Fahd University of Petroleum and Minerals. Dr. Abuelma'Atti is a contributor to Encyclopedia of RF and Microwave Engineering, Kai Chang, Editor, (New York: John Wiley, 2005), Survey of Instrumentation and Measurement, S.A. Dyer, Editor, (New York: John Wiley, 2001), The Encyclopedia of Electrical and Electronic Engineering, J.G. Webster, Editor, (New York:John Wiley, 1999), and Selected Papers on Analog Fiber-Optic Links, E.I. Ackerman, C.H. Cox III and N.A. Riza, Editors, SPIE Milestone Series, (Washington: SPIE Optical Engineering Press, 1998).
His research interests include problems related to analysis and design of nonlinear electronic circuits and systems, analog integrated circuits and active networks design. He is the author or co-author of over 500 journal articles and technical presentations.
Abdullah Bakri Shwehneh was born in Aleppo Syria, in 1973. He received the B. Sc. degree in electrical engineering in 1998 from Sumy State University, Sumy, Ukraine. In 2001, he received the Postgraduate Diploma in Automatic Control from Aleppo State University, Aleppo, Syria.
In 2001, he joined the “Electronic Brain Company for Computer and Electronics” as an Electronic & Computer Engineer and since 2002, he is with King Fahd University of Petroleum and Minerals (KFUPM), Dhahran, Saudi Arabia, as a Research Assistant. In June 2005 he obtained his Master of Science Degree in Analog Electronics from KFUPM. At present he is a Ph.D. student at KFUPM. His main interests are in Nonlinear circuits, VLSI Analog Design and Neural Networks hardware implementation.
Rights and permissions
About this article
Cite this article
Abuelma'Ati, M.T., Shwehneh, A. A Reconfigurable Gaussian/Triangular Basis Functions Computation Circuit. Analog Integr Circ Sig Process 47, 53–64 (2006). https://doi.org/10.1007/s10470-006-2775-2
Received:
Revised:
Accepted:
Issue Date:
DOI: https://doi.org/10.1007/s10470-006-2775-2