Frequency Adaptive Single-Transistor Neuron Based on Temporal Charge Trapping

Spiking frequency adaptation (SFA) plays an important role in neuromorphic systems, enhancing energy efficiency and resilience to unwanted noise. In this study, the SFA function is implemented in a single-transistor neuron utilizing a silicon–oxide–nitride–silicon (SONS) gate-stack. The experimental results confirm that the SFA function is successfully achieved through temporal hole trapping in the nitride layer adjacent to the channel. Unlike circuit-based approaches, this work presents a novel method for implementing SFA at the single-transistor level, significantly improving both area and energy efficiencies.