PARALLEL MULTIPLIER BASED ON SPIKING-TYPE NEURAL NETWORKS WITH UNIVERSAL UNARY TO BE IMPLEMENTED IN EMBEDDED DIGITAL SYSTEMS.

摘要

The present disclosure is related to a parallel multiplier system that processes positive integers. The multiplier model is based on spiking-type neural networks in order to process excessively large numbers with the minimal computation and hardware. First, the numbers that the multiplier processes are converted to a unary system, in which each unit is represented by a spike. Subsequently, the set of neurons work in parallel, where each neuron processes the spikes serially. Finally, the resulting spikes, which have been produced by neurons excitation, are sent to a decoder to convert them into a decimal representation.