We report on a monolithic semiconductor device that computes the coincidence of time delayed electrical impulses and selectively regenerates those impulses that meet within 10 mu s of each other. We obtain pulse regeneration using the negative feedback of a tunnel device and demonstrate reliable pulse amplification using an external circuit. The device emulates biological neurons which detect correlations between spike trains in real time and are able to modify synaptic weights through spike timing dependent plasticity.
|Number of pages||4|
|Journal||Physica Status Solidi A: Applications and Materials Science|
|Publication status||Published - 2007|
|Event||3rd International Conference on Micro-Nanoelectronics, Nanotechnology and MEMs - Athens, Greece|
Duration: 18 Nov 2007 → 21 Nov 2007
Samardak, A. S., Taylor, S., Nogaret, A., Hollier, G., Austin, J., Farrer, I., & Ritchie, D. A. (2007). Coincidence detection and spike regeneration in artificial neurons. Physica Status Solidi A: Applications and Materials Science, 205(11), 2651-2654. https://doi.org/10.1002/pssa.200780156