CSFNN synapse and neuron design using current mode analog circuitry


ERKMEN B. , Yidirm T.

11th International Conference on Knowledge-Based and Intelligent Information and Engineering Systems, KES 2007, and 17th Italian Workshop on Neural Networks, WIRN 2007, Vietri sul Mare, Italy, 12 - 14 September 2007, pp.17-25 identifier identifier

  • Publication Type: Conference Paper / Full Text
  • Volume:
  • Doi Number: 10.1007/978-3-540-74819-9_3
  • City: Vietri sul Mare
  • Country: Italy
  • Page Numbers: pp.17-25

Abstract

In this paper, a neuron and synapse circuitry of Conic Section Neural Network (CSFNN) is presented. The proposed circuit has been designed to compute the Radial Basis Function (RBF) and Multilayer Perceptron (MLP) propagation rules on a single hardware to form a CSFNN neuron. Decision boundaries, hyper plane (for MLP) and hyper sphere (for RBF), are special cases of Conic Section Neural Networks depending on the data distribution of a given applications. Current mode analog hardware has been designed and the simulations of the neuron and synapse circuitry have been realized using Cadence with AMIS 0.5μm CMOS transistor model parameters. Simulation results show that the outputs of the circuits are very accurately matched with ideal curve. Open and closed decision boundaries have also been obtained using designed circuitry to demonstrate functionality of designed CSFNN neuron. © Springer-Verlag Berlin Heidelberg 2007.