Abstract
This paper presents an intriguing relation between neural networks having as weights the Fibonacci numbers and the Addition of (two) binary numbers. The practical application of interest is that such ‘Fibonacci’ networks are VLSI-optimal with respect to the area of the circuit. We shortly present the state-of-the-art, and detail a class of multilayer solutions for Addition. For this class we will prove constructively that the weights of the threshold gates implementing the Boolean functions are the Fibonacci numbers. As the weights are the smallest integers (by construction), the area of the VLSI circuit—estimated as the sum of the digits needed to represent the weights—is minimised. Therefore this class of solutions is VLSI-optimal. Conclusions and open questions are ending the paper.
Preview
Unable to display preview. Download preview PDF.
References
Alon, N., & Bruck, J. (1991). Explicit Construction of Depth-2 Majority Circuits for Comparison and Addition. IBM Technical Report RJ 8300 (75661). San Jose, CA: IBM Almaden Research Center.
Alon, N., & Bruck, J. (1994). Depth-2 Threshold Logic Circuits for Logic and Arithmetic Functions. Patent US 5357528.
Beiu, V., Peperstraete, J.A., Vandewalle, J., & Lauwereins, R., (1994a). Area-Time Performances of Some Neural Computations. In P. Borne. T. Fukuda & S.G. Tzafestas (Eds.): Proc. IMACS Intl. Symp. on Signal Proc. Robotics and Neural Networks, Lille, France (pp. 664–668). Lille: GERF EC.
Beiu, V., Peperstraete, J.A., Vandewalle, J., & Lauwereins, R., (1994b). On the Circuit Complexity of Feedforward Neural Networks. In M. Marinaro & P.G. Morasso (eds.): Proc. Intl. Conf. on Artif. Neural Networks, Sorrento, Italy (pp. 521–524). Springer-Verlag.
Beiu, V., Peperstraete, J.A., Vandewalle, J., & Lauwereins, R., (1994c). Optimal Parallel Addition Means Constant Fan-In Threshold Gates. In Proc. Intl. Conf. on Technical Informatics, Timisoara (vol. 5, pp. 166–177). Timisoara: Technical University of Timisoara Press.
Beiu, V. (1996). Constant Fan-In Discrete Neural Networks Are VLSI-Optimal. In S. W. Ellacott, J.C. Mason, & I.J. Anderson (eds.): Mathematics of Neural Networks—Models, Algorithms and Applications (pp. 89–94). Kluwer Academic.
Beiu, V., & Taylor, J.G. (1996). On the circuit complexity of sigmoid feed-forward neural networks. Neural Networks, 9(7), 1155–1171.
Beiu, V. (1998). On the Circuit and VLSI Complexity of Threshold Gate Comparison. Neurocomputing, 19(1), 77–98.
Brent, R.P., & Kung, H.T. (1982). A Regular Layout for Parallel Adders. IEEE Trans. on Comp., 31(3), 260–264.
Cannas, S.A. (1995). Arithmetic Perceptrons. Neural Computation, 7(1), 173–181.
Chandra, A.K., Stockmeyer, L.J., & Vishkin, U. (1984). Constant Depth Reducibility. SIAM J. Comput., 13(2), 423–539.
Chang, P.K., Schlag, M.D.F., Thomborson, C.D., & Oklobdzija, V.G. (1992). Delay Optimization of Carry-Skip Adders and Block Carry-Lookahead Adders Using Multidimensional Programming. IEEE Trans. on Comp., 41(8), 920–930.
Cotofana, S., & Vassiliadis, S. (1997). Low Weight and Fan-In Neural Networks for Basic Arithmetic Operations. In Proc. IMACS World Congress on Sci. Comput., Modelling and Appl. Maths. (vol. 4, pp. 227–232).
Doran, R.W. (1988). Variants of an Improved Carry Look-Ahead Adder. IEEE Trans. on Comp., 37(9), 1110–1113.
Han, T., Carlson, D.A., & Levitan, S.P. (1987). VLSI Design of High-Speed, Low-Area Addition Circuitry. In Proc. Intl. Conf. on Circuit Design (pp. 418–422). IEEE Press.
Hu, S. (1965). Threshold Logic. Berkeley, Los Angeles: University of California Press.
Hwang, K. (1979). Computer Arithmetic: Principles, Architecture and Design. New York: John Wiley & Sons.
Kelliher, T.P., Owens, R.M., Irwin, M.J., & Hwang, T.-T. (1992). ELM A Fast Addition Algorithm Discovered by a Program. IEEE Trans. on Comp., 41(9), 1181–1184.
Khrapchenko, V.M. (1967). Asymptotic Estimation of Addition Time of a Parallel Adder. Problemy Kibernetiki, 19, 107–122 (in Russian); English translation (1970) in System Theory Research 19, 105–122.
Kogge, P.M., & Stone, H.S. (1973). A Parallel Algorithm for the Efficient Solution of a General Class of Recurrence Equations. IEEE Trans. on Comp., 22(8), 783–791.
Ladner, R.E., & Fischer, M.J. (1980). Parallel Prefix Computations. J. ACM, 27(4), 831–838.
Ling, H. (1981). High Speed Binary Adder. IBM J. Res. Develop., 25(2&3), 156–166.
Lynch, T., & Swartzlander, E.E., Jr. (1992). A Spanning Tree Carry Lookahead Adder. IEEE Trans. on Comp., 41(8), 931–939.
Montoye, R.K. (1981). Area-Time Efficient Addition in Charge Based Technology. Proc. Intl. Design Automation Conf. (pp. 862–872). IEEE Press.
Ngai, T.F., & Irwin, M.J. (1985). Regular Area-Efficient Carry-Lookahead Adders. Proc. Intl. Symp. on Comp. Arithmetic (pp. 9–15). ACM Press.
Ong, S., & Atkins, D.E. (1983). A Comparison of ALU Structures for VLSI Technology. Proc. Intl. Symp. on Comp. Arithmetic (pp. 10–16). ACM Press.
Parberry, I. (1994). Circuit Complexity and Neural Networks. Cambridge, MA: MIT Press.
Quach, N.T., & Flynn, M.J. (1992). High-Speed Addition in CMOS. IEEE Trans. on Comp., 41(12), 1612–1615.
Rhyne, T. (1984). Limitations on Carry Lookahead Networks. IEEE Trans. on Comp., 33(4), 373–374.
Siu, K.-Y., Roychowdhury, V., & Kailath, T. (1990). Computing with Almost Optimal Size Threshold Circuits. (Technical Report, Info. Sys. Lab., Stanford University).
Siu, K.-Y., & Bruck, J. (1990). Neural Computation of Arithmetic Functions. Proc. IEEE, 78(10), 1669–1675.
Siu, K.-Y., & Bruck, J. (1991). On the Power of Threshold Circuits with Small Weights. SIAM J. Disc. Maths., 4(3), 423–435.
Siu, K.-Y., Roychowdhury, V.P., & Kailath, T. (1991). Depth-Size Tradeoffs for Neural Computations. IEEE Trans. on Comp., 40(12), 1402–1412.
Sugla, B. (1985). Parallel Computation with Limited Resources. PhD Dissertation, Dept. ECE. Univ. of Massachusetts, Amherst, MA.
Vassiliadis, S., Cotofana, S., & Berteles, K. (1996). 2-1 Addition and Related Arithmetic Operations with Threshold Logic. IEEE Trans. on Comp., 45(9), 1062–1068.
Waser, S., & Flynn, M.J. (1982). Introduction to Arithmetic of Digital Systems. Holt, Rinehart and Winston.
Wegener, I. (1987). The Complexity of Boolean Functions. Chichester: Wiley-Teubner.
Wegener, I. (1996). Unbounded Fan-In Circuits. In L.L. Keener (ed.): Advances in the Theory of Computation and Computational Mathematics, 123–153. Ablex.
Wegener, I. (1993). Optimal Lower Bounds on the Depth of Polynomial Size Threshold Circuits for Some Arithmetic Functions. Information Proc.Lett., 46, 85–87.
Wei, B.W.Y., & Thompson, C.D. (1990). Area-Time Optimal Adder Design. IEEE Trans. on Comp., 39(5), 666–675.
Author information
Authors and Affiliations
Corresponding author
Editor information
Rights and permissions
Copyright information
© 1999 Springer-Verlag Berlin Heidelberg
About this paper
Cite this paper
Beiu, V. (1999). Neural addition and fibonacci numbers. In: Mira, J., Sánchez-Andrés, J.V. (eds) Engineering Applications of Bio-Inspired Artificial Neural Networks. IWANN 1999. Lecture Notes in Computer Science, vol 1607. Springer, Berlin, Heidelberg . https://doi.org/10.1007/BFb0100486
Download citation
DOI: https://doi.org/10.1007/BFb0100486
Published:
Publisher Name: Springer, Berlin, Heidelberg
Print ISBN: 978-3-540-66068-2
Online ISBN: 978-3-540-48772-2
eBook Packages: Springer Book Archive