A hardware-efficient ergodic sequential logic neuron network for brain prosthetic FPGA

This study presents a novel ergodic sequential logic neuron network. It is shown that the presented network can reproduce nonlinear input-output characteristics of a biologically plausible mathematical neuron network model. In addition, it is revealed that a post synaptic unit, which realizes nonlinear dynamics of excitatory post-synaptic potential, plays an important role to increase the reproduction ability of the presented network.