Nanosignal processing: Stochastic resonance in carbon nanotubes that detect subthreshold signals

Experiments confirm that small amounts of noise help a nanotube transistor detect noisy subthreshold electrical signals. Gaussian, uniform, and impulsive (Cauchy) noise produced this feedforward stochastic-resonance effect by increasing both the nanotube system’s mutual information and its input-output correlation. The noise corrupted a synchronous Bernoulli or random digital sequence that fed into the thresholdlike nanotube transistor and produced a Bernoulli sequence. Both Shannon’s mutual information and correlation measured the performance gain by comparing the input and output sequences. This nanotube SR effect was robust: it persisted even when infinite-variance Cauchy noise corrupted the signal stream. Such noise-enhanced signal processing at the nanolevel promises applications to signal detection in wideband communication systems and biological and artificial neural networks. Noise can help carbon nanotube transistors detect subthreshold electrical signals by increasing the transistor’s input output mutual information or correlation. Several researchers have demonstrated the stochastic resonance (SR) effect for various types of threshold units or neurons. 1-6 Experiments