Precisely-timed synchronization among spiking neural circuits on analog VLSIs

A neural network exhibiting precisely–timed synchronization under a noisy environment with depressing synapses is proposed [1]. We explain how the neural network was constructed by using silicon neural circuits [2] and depressing synapse circuits [3], which we have already developed, and evaluate the precision of timing among silicon neurons. Moreover, a simple analog spike timing dependent plasticity STDP circuit was designed for constructing a neural network that exhibited robust synchronization under a noisy environment. We confirmed that it operated as required.