potential utility for the creation of physically intelligent machines capable of evolution and learning. In addition to the exciting behaviors described in Sect. 5.6,
complex networks of coupled nonlinear elements such as this commonly manifest
non-trivial spatiotemporal evolution through dynamic system reconfigurations. Such
reconfigurations, readily described by critical dynamics, enable enhanced maintenance of system correlations and more effective signal propagation. Indicators of
criticality typically include power-law scaling of 1/f fluctuations and temporal
metastability. Analysis of the power spectral density of network conductivity in
the activated state revealed 1/f power law scaling over five orders of magnitude with
f % 1.4 (data not shown). Temporal metastability was observed during sub-threshold
pulsed voltage stimulation, analogous to methods employed in neuroscience to
probe cortical cultures. Under typical conditions, the current response fluctuated
over a wide range of metastable conductance states associated with discrete network
configurations (Fig. 11c–f), as classified by residence times in a given state ranging
from milliseconds (within a single stimulation pulse) to several seconds (across
hundreds of pulses). Comparing the probability of state duration with its time
duration indicated a power law distribution with %1.8 (Fig. 11e–f), indicating a
Fig. 11 Electrical characteristics of complex nanoelectroionic networks. (a) Experimental I-V
curve demonstrating pinched hysteresis; RON ¼ 8 KΩ, ROFF > 10 MΩ. (b) Ultrasensitive IR
image of a distributed device conductance under external bias at 300 K. (c, e) Representative
network current response to a 2 V pulse showing switching between discrete, metastable conductance states. (d, f) Temporal correlation of metastable states observed during pulsed stimulation
demonstrated power law scaling for probability, P(D), of duration. Power law scaling existed for
residence time both (d) within a single 10 ms pulse and (f) over 2.5 s during extended periods of
pulsed stimulation (Stieg Adv. Mat 2012)
Atomic Switch Networks for Neuroarchitectonics: Past, Present, Future
229
complex networks of coupled nonlinear elements such as this commonly manifest
non-trivial spatiotemporal evolution through dynamic system reconfigurations. Such
reconfigurations, readily described by critical dynamics, enable enhanced maintenance of system correlations and more effective signal propagation. Indicators of
criticality typically include power-law scaling of 1/f fluctuations and temporal
metastability. Analysis of the power spectral density of network conductivity in
the activated state revealed 1/f power law scaling over five orders of magnitude with
f % 1.4 (data not shown). Temporal metastability was observed during sub-threshold
pulsed voltage stimulation, analogous to methods employed in neuroscience to
probe cortical cultures. Under typical conditions, the current response fluctuated
over a wide range of metastable conductance states associated with discrete network
configurations (Fig. 11c–f), as classified by residence times in a given state ranging
from milliseconds (within a single stimulation pulse) to several seconds (across
hundreds of pulses). Comparing the probability of state duration with its time
duration indicated a power law distribution with %1.8 (Fig. 11e–f), indicating a
Fig. 11 Electrical characteristics of complex nanoelectroionic networks. (a) Experimental I-V
curve demonstrating pinched hysteresis; RON ¼ 8 KΩ, ROFF > 10 MΩ. (b) Ultrasensitive IR
image of a distributed device conductance under external bias at 300 K. (c, e) Representative
network current response to a 2 V pulse showing switching between discrete, metastable conductance states. (d, f) Temporal correlation of metastable states observed during pulsed stimulation
demonstrated power law scaling for probability, P(D), of duration. Power law scaling existed for
residence time both (d) within a single 10 ms pulse and (f) over 2.5 s during extended periods of
pulsed stimulation (Stieg Adv. Mat 2012)
Atomic Switch Networks for Neuroarchitectonics: Past, Present, Future
229