Noise Effects on Performance of Signal Detection in an Analog VLSI Resonate-And Fire Neuron

In this paper, we present analog VLSI implementation of a resonate-and-fire neuron (RFN) model, and then consider noise effects on its performance of signal detection. The RFN circuit is a silicon spiking neuron that has second-order membrane dynamics and exhibits dynamic behavior, such as fast subthreshold oscillation, coincidence detection, frequency preference, and post-inhibitory rebound. Due to such dynamic behavior, the RFN circuit acts as a band-pass filter and a coincidence detector at circuit level. Through SPICE simulations, we will demonstrate noise effects on the coincidence detection and the frequency preference of the RFN circuit.