Analog Circuit Implementation of a Cortical Neuron

Abstract

Cortical neurons play a predominant role in major functions like motor and sensory actions, cognition, perception, etc. The analysis, modeling of cortical neurons facilitates the implementation of faster and smarter neuromorphic architectures. This paper presents the implementation of an analog CMOS circuit that resembles the functionality of cortical neurons. This silicon neuron circuit comprises only 14 MOSFETS and is capable of providing various kinds of spiking patterns such as regular, fast-spiking, and bursting, just by varying bias voltages. This property enables the fabrication of many neurons onto a single chip and exhibits flexibility of emulating different neuronal behaviors with a little modification in bias voltages. This makes a circuit that can be used as a basic cell in the implementation of spiking neural networks, brain-inspired circuits, and cognitive robots, etc. The circuit is analyzed for its performance and the simulations are carried out using cadence virtuoso at 180nm technology node.