1. CSN
  2. All Volumes and Issues
  3. Volume 2, Number 1, 2020
  4. Simplified Parallel Sorting Discrete-Time Neural Network Model

Simplified Parallel Sorting Discrete-Time Neural Network Model

CSN.
2020;
Volume 2, Number 1
: pp. 94 - 101
https://doi.org/10.23939/csn2020.01.094
Authors:
  1. Pavlo Tymoshchuk
1
Lviv Polytechnic National University, Department of Computer Aided Design Systems

A model of parallel sorting neural network of discrete-time has been proposed. The model is described by system of difference equations and by step functions. The model is based on simplified neural circuit of discrete-time that identifies maximal/minimal values of input data and is described by difference equation and by step functions. A bound from above on a number of iterations required for reaching convergence of search process to steady state is determined. The model does not need a knowledge of change range of input data. In order to use the model a minimal difference between values of input data should be known. The network can process unknown input data with finite values, located in arbitrary unknown finite range. The network is characterized by moderate computational complexity and complexity of software implementation, any finite resolution of input data, speed,. Computing simulation results illustrating efficiency of the network are given.

parallel sorting
neural network
difference equation
computational complexity
hardware implementation
  1. Knuth D. E., The Art of Computer Programming, Sorting and Searching. Reading, MA: Addison-Wesley, 1973
  2. Akl S. G., Parallel Sorting Algorithms, Orlando, FL: Academic, 1985.
  3. Atkins M., “Sorting by Hopfield nets”, in Proc. Int. Joint Conf. Neural Netw., Washington, DC, USA, 1989, pp. 65–68.
  4. Takefuji Y. and Lee K.-S., “A super parallel sorting algorithm based on neural networks”, IEEE Trans. Circuits Syst., vol. CAS—37, 1990. no. 11, pp. 1425–1429.
  5. W. Chen and K. Hsieh, “A neural sorting network with O(1) time complexity”, in Proc. Int. Joint Conf. Neural Networks, vol. III, San Diego, CA, 1990, pp. 793–798.
  6. Kwon T. M. and Zervakis M., “A parallel sorting network without comparators: A neural network approach,” in Proc. Int. Joint Conf. Neural Networks, vol. I, Baltimore, MD, 1992, pp. 701–706.
  7. Tseng Y.-H. and Wu J.-L., “Solving sorting and related problems by quadratic perceptrons”, Electron. Lett., 1992. vol. 28, no. 10, pp. 906–908,.
  8. Wang J., “Analysis and design of an analog sorting network,” IEEE Trans. Neural Networks, 1995. vol. 6, no. 4, pp. 962–971, Jul.
  9. Kwon T. M. and Zervakis M., “KWTA networks and their applications”, Multidimensional Syst. and Signal Processing, 1995. vol. 6, no. 4, pp. 333–346, Oct.
  10. . Wang J., “Analysis and design of a k-winners-take-all model with a single state variable and the Heaviside step activation function”, IEEE Trans. Neural Networks. Sept. 2010. vol. 21, no. 9, pp. 1496–1506.
  11. Alnuweiri H. M. and Kumar V. K. P., “Optimal VLSI sorting with reduced number of processors”, IEEE Trans. Comput., 1991. vol. C-40, pp. 105–110.
  12. Rovetta S. and Zunino R., “Minimal-connectivity programmable circuit for analog sorting”, IEE Proc. Circuits, Devices Syst., vol. 146, no. 3, pp. 108–110, Aug. 1999.
  13. Cichocki A. and Unbehauen R., Neural Networks for Optimization and Signal Processing. New York, NY, USA: Wiley, 1993.
  14. Tymoshchuk P. V., “A discrete-time dynamic K-winners-take-all neural circuit”, Neurocomputing, vol. 72, 2009, pp. 3191–3202.