Novel Memtransistor-Based Lif Neuron with Tunable Ionic Dynamics for Spiking Neural Networks

To realize energy-efficient intelligent systems for spiking neural networks, we propose a novel compact artificial neuron based on ion-modulated memtransistors, which can realize the leaky, integrate-and-fire (LIF) functions by utilizing the ionic dynamics. Through the spontaneous decay dynamics of ions, the proposed neurons can avoid external capacitors or any other reset circuits. Moreover, we explore the influences of neuron firing under different stimulation parameters and tunable decay time constants under different base voltages; by adding this extra freedom of modulation, the network performances can be boosted.