Performance of LIF neuron circuit based on threshold switching memristors

Abstract

Building physical systems that mimic biological functions is crucial for enhancing the efficiency and scalability of neuromorphic computing. The Leaky Integrate-and-Fire (LIF) neuron model has gained attention owing to its simple architecture, low power consumption, high temporal resolution, and real-time processing. In this study, the spiking behaviors of LIF neuron circuits based on two types of threshold switching memristor devices (TSM)—Pt/Al₂O₃/HfO₂/Ag (5 nm)/Pt and Pt/Al₂O₃/HfO₂/Ag-NIs(5 nm)/Pt—were systematically investigated and compared under various voltage inputs and circuit elements. Simulations were also performed to explain the patterns of spiking behaviors observed in the experiments. The LIF neuron circuit based on the TSM device embedded with Ag nano-islands (Ag-NIs) demonstrates unique spike response characteristics in contrast to those without Ag-NIs. Specifically, the spike amplitude increases with increasing input voltage amplitude while the output spike frequency remains stable. This study highlights the significant influence of TSM materials on the performance of LIF neuron circuits and paves the way for the design of more efficient and reliable neuromorphic computing architectures.