Direct Signal Encoding With Analog Resonate-and-Fire Neurons

Sensors are an essential element in a wide range of applications. As the number sensors increases, so does amount data collected with them. This raises challenge efficiently processing this data. Spiking Neural Networks (SNNs) represents promising approach to solve problem through event-based, parallelized processing. For SNNs be genuinely efficient, some fundamental challenges arise, like converting analog signals spike events. An emerging possibility is use Resonate-and-Fire (R&F) neurons, capable reacting specific frequency components input signals. In work, we present possible implementation for R&F neuron and show practical encoding into spiking domain using actual measurements. The coding method allows sensor directly applied efficient future, can potentially enable direct integration sensors.