A lightweight CMOS-based LIF circuit: modeling and spiking regulation.

Area of Science:

• Neuromorphic Engineering
• Integrated Circuit Design
• Computational Neuroscience

Background:

• Spike-based neuromorphic computing requires efficient hardware.
• Lightweight spiking circuits are crucial for large-scale implementations.
• Existing designs may lack efficiency or robustness.

Purpose of the Study:

• To present a novel, lightweight CMOS-based Leaky Integrate-and-Fire (LIF) circuit.
• To demonstrate the circuit's spiking activity and regulatory mechanisms.
• To analyze its mathematical model, functional characteristics, and robustness.

Main Methods:

• Design and simulation of a minimalist LIF circuit using CMOS technology (2 P-MOS, 3 N-MOS).
• Derivation of a mathematical model relating input current to output voltage.
• Analysis of spiking frequency versus input current intensity.
• Circuit simulations using Cadence Virtuoso.

Main Results:

• The LIF circuit accurately implements integration, threshold detection, discharge, and reset functions.
• Demonstrated robustness against temperature and process variations.
• Achieved a compact area of [Formula: see text] and low energy consumption of 0.096 pJ/spike (0.18 µm CMOS).

Conclusions:

• The developed LIF circuit offers a viable approach for compact and low-power neuromorphic hardware.
• Its efficiency and robustness support its use in large-scale neuromorphic systems.
• This work contributes to advancing spike-based computing hardware.