Bio-Inspired Stochastic Neural Networks for Nanoelectronics

Abstract

Nanoelectronics and neural networks are an ideal combination. Fault tolerance found in neuraI networks can make reliable systems from nanoelectronics. However, conventional neuraI networks are not suitable due to the topological requirements of nanoelectronics. Therefore a new design of neuraI networks using biological principles, because of the similarity with nanoelectronics is necessary. Besides new topologies, such as lattice structures and locai bio-inspired learning mies based on correlation, for example Classical conditioning and Hebbian learning, this paper aIso describes a discrete state Markov process suitable for modeling the stochastics in nanoelectronics.

Nanocomputing in the Presence of Defects and Faults: A Survey

“…In [45], Rouw and Hoekstra identify two major challenges to nanoelectronic neural network fabrication. Neural networks are typically connection-rich, requiring each node to communicate across long distances.…”