A novel hardware-efficient spiking neuron model based on asynchronous cellular automaton dynamics exhibiting various nonlinear response curves

In this paper, a novel hardware-efficient spiking neuron model based on an asynchronous cellular automaton is proposed. It is shown that the proposed model can exhibit typical nonlinear response curves (so-called IF-curves) such as class 1 excitability without hysteresis, class 2 excitability without hysteresis, and class 2 excitability and class 1 spiking with hysteresis. The proposed model is implemented by an FPGA device and it is shown that the FPGA-implemented proposed model can exhibit the typical nonlinear response curves. Furthermore, it is shown that the proposed model can be implemented by much fewer circuit elements compared to some conventional neuron models that exhibit the typical nonlinear response curves.