A Beta basis function neural network in CMOS subthreshold mode

Abstract

Beta basis function neural networks (BBFNNs) are powerful systems for learning and universal approximation. In this paper, we present a hardware implementation of the Beta neuron using the CMOS subthreshold mode. We describe the low power± low voltage analogue Beta neuron circuit. Three main modules are used to realize the electronic Beta function: a logarithmic currentto-voltage converter, a multiplier and an exponential voltage-to-current converter. Simulation results show the validity of our neural hardware implementation. The parameters of the electronic Beta function are controlled independently by current sources. This analogue implementation could be used easily to realize analogue BBFNNs.

A Reconfigurable Gaussian/Triangular Basis Functions Computation Circuit

“…In this circuit a special attention is paid to improve symmetrical characteristics of the function. Other current-mode CMOS-based RBF circuits, with transistors operating in weak inversion mode, were developed [20][21][22]. The main disadvantage of these circuits is their large area consumption which makes them not attractive for VLSI design.…”