A method whereby information encoded in spiking activity or current amplitude of a population of neurons may be transferred to a second population of neurons or simulated neurons.
Neuromorphic engineering refers to the design of artificial neural systems based on those of biological nervous systems. Research in neuromorphic engineering has led to the development of neuromorphic computers with millions of neurons (and soon many more) capable of computing with very high power efficiency.
Researchers at the University of California, Davis have developed a method for dynamically routing information between populations of neurons in neuromorphic systems. This method may be used to form the basis for on-chip, controllable parallel computation in which information in the form of spikes or current amplitudes may be dynamically routed through sets of neural circuits with both fixed and plastic synaptic connectivity.
|United States Of America||Issued Patent||9,916,530||03/13/2018||2015-136|