A low-power CMOS spiking neuron design with frequency adaptation in 180 nm technology at 300 mV supply
Analog Integrated Circuits and Signal Processing · 2026-11 · 129 (2)
DOI: 10.1007/s10470-026-02670-7 ↗Kurz gefasst
Fachartikel in der Zeitschrift Analog Integrated Circuits and Signal Processing, online erschienen am 28.10.2026. Verfasst von Mehdi Haji Rahimi und 1 weiteren Autor:innen. Die Arbeit nennt 23 Literaturangaben.
Abstract (Originalsprache)
This paper presents a low-power conductance-based Leaky Integrate-and-Fire (LIF) spiking neuron in TSMC 180-nm CMOS technology operating at 300 mV supply. The circuit uses a Differential Pair Integrator (DPI), an enhanced positive feedback comparator that boosts spike amplitude by 67%, and a frequency adaptation mechanism with a slow variable capacitor. Post-layout simulations with parasitic extraction in Cadence Virtuoso yield a stable firing frequency of 255 ± 2 Hz, 1.0 V spike amplitude, and 9.3 pJ/spike energy consumption. The design achieves 22.5% better energy efficiency and strong PVT robustness, making it suitable for large-scale, low-cost neuromorphic systems in edge computing, IoT, and robotics.
Zitieren
APA
Rahimi, M. H., & Borjlu, S. R. (2026). A low-power CMOS spiking neuron design with frequency adaptation in 180 nm technology at 300 mV supply. Analog Integrated Circuits and Signal Processing, 129(2). https://doi.org/10.1007/s10470-026-02670-7
Vancouver
Rahimi MH, Borjlu SR. A low-power CMOS spiking neuron design with frequency adaptation in 180 nm technology at 300 mV supply. Analog Integrated Circuits and Signal Processing. 2026;129(2). doi:10.1007/s10470-026-02670-7
BibTeX
@article{Rahimi2026,
author = {Rahimi, Mehdi Haji and Borjlu, Shaban Rezaei},
title = {A low-power CMOS spiking neuron design with frequency adaptation in 180 nm technology at 300 mV supply},
journal = {Analog Integrated Circuits and Signal Processing},
year = {2026},
volume = {129},
number = {2},
doi = {10.1007/s10470-026-02670-7}
}
EUVIMED