Fast Determination of Similarity Between Two Vectors by Means of Analog CMOS Technique
Treść / Zawartość
In this paper, an analog approach to determining a resemblance between two multidimensional vectors is proposed. As the resemblance measure, Euclidean distance is used. The main advantage of the presented method is a very high speed of the Euclidean-distance-measure calculations. The achieved high speed results from the fact that most of arithmetic operations needed to realize the calculations are carried out in parallel. This concerns the required operations of squaring a difference of two corresponding components of the compared vectors. Operating in a transconductane mode (voltage difference squaring transconductors) and a current mode (output square-root extracting circuit), our CMOS circuit is power saving. Its low-power operation results from the fact that sub-circuits of our calculator responsible for the squaring operations (a great number of them in case of large multidimensional vectors) consume no power in the absence of input signals. This takes place when corresponding components of the compared vectors are both equal to zero. The circuit also consumes a reasonably low amount of energy when processing (comparing) a different from zero input data (corresponding vector components). A simplified description of the applied differential squaring transconductors as well as the output current-mode square-root extraction circuit is given and a problem of good cooperation between them is discussed and proper solutions indicated. SPICE simulation results are shown to be in a good agreement with the theory presented.
Bibliogr. 19 poz., wykr.
- Faculty of Telecommunication and Electrical Engineering, University of Technology and Life Sciences, Kaliskiego 7, 85-796 Bydgoszcz, Poland, firstname.lastname@example.org
-  S. Ahalt, A. Krishnamurthy, P. Chen, and D. Melton, “Competitive learning algorithms for vector quantization,” Neural Networks, vol. 3, pp. 131-134, 1990.
-  G. Cauwenberghs, M. A. Bayoumi, and E. Sanchez-Sinencio, Learning on silicon: Adaptive VLSI Neural Systems. Kluwer Academic Publishers, 1999.
-  S.-L. Chen, H.-Y. Lee, Y.-W. Chu, C.-A. Chen, C.-C. Lin, and C.-H. Luo, “A variable control system for wireless body sensor network,” in Proc. IEEE International Symposium on Circuits and Systems (ISCAS), May 2008, paper 18-21, pp. 2034-2037.
-  Y. Chen and F. Bastani, “ANN with two-dendrite neurons and its weight initialization,” in Proc. International Joint Conference on Neural Networks (IJCNN), Baltimore, USA, 1992, pp. 139-146.
-  D. DeSieno, “Adding a conscience to competitive learning,” in Proc. IEEE Conference Neural Network, vol. 1, 1988, pp. 117-124.
-  R. Długosz, T. Talaśka, and R. Wojtyna, “New binary-tree-based Winner-Takes-All circuit for learning on silicon Kohonen’s networks,” in Proc. IEEE Int. Conf. Signals and Electronic Systems (ICSES), Łódź, Poland, 2006.
-  S. Fakhraie and K. C. Smith, VLSI-compatible implementations for artificial neural networks. Kluwer Academic Publishers, 1997.
-  L. Gatet, H.Tap-Béteille, and F. Bony, “Comparison between analog and digital neural network implementations for range-finding applications,” IEEE Trans. Neural Netw., vol. 20, no. 3, Mar. 2009.
-  M. Holler et al., “An electrically trainable artificial neural network (ETANN) with 10240 'floating gate' synapses,” in Proc. International Joint Conference on Neural Networks (IJCNN), Jun. 1989, pp. 191-196.
-  B. Linares-Barranco, E. Sanchez-Sinencio, A. Rodriguez-Vazquez, and J. L. Huertas, “A CMOS analog adaptive BAM with on-chip learning and weight refreshing,” IEEE Trans. Neural Netw., vol. 4, no. 3, pp. 445-455, 1993.
-  D. Macq, M. Verleysen, P. Jespers, and J.-D. Legat, “Analog implementation of a Kohonen map with on-chip learning,” IEEE Trans. Neural Netw., vol. 4, no. 3, pp. 456-461, 1993.
-  A. Rajah and M. K. Hani, “ASIC design of a Kohonen Neural Network microchip,” in Proc. IEEE International Conference on Semiconductor Electronics (ICSE), 2004, pp. 148-158.
-  T. Talaśka, R. Długosz, and W. Pedrycz, “Adaptive weight change mechanism for Kohonens’s Neural Network implemented in CMOS 0:18m technology,” in Proc. European Symposium on Artificial Neural Networks (ESANN), Bruges, Belgium, Jun. 2007, pp. 151-156.
-  R. Wojtyna, “Simple CMOS transconductance-mode differential squarer,” in Proc. IEEE Workshop Signal Processing’2005, Poznań, Poland, 2005, p. 171.
-  R. Wojtyna, “Current-mode analog square rooter for hardware neuroprocessing,” in Proc. IEEE Workshop Signal Processing’2006, Poznań, Poland, 2006.
-  R. Wojtyna, “CMOS transconductance-mode analog circuit for fast determining Euclidean distance,” Elektronika, no. 4, pp. 65-68, 2007.
-  R. Wojtyna, “Current-mode analog memory with extended storage time for hardware-implemented neural networks,” Elektronika, no. 3, pp. 34-38, 2009.
-  R. Wojtyna and T. Talaśka, “Improved power-saving synapse for adaptive neuroprocessing on silicon,” in Proc. IEEE Int. Conf. Signals and Electronic Systems (ICSES), Poznań, Poland, 2004, pp. 27-30.
-  T. Kohonen, Self-Organizing Maps. Berlin: Springer Verlag, 2001.