Search

We discuss CNNs in which the states are defined by the electrical phase of a dynamic physical process, such as electron tunneling in ultra-small junctions or integrate-and-fire processes in nanoscale structures or molecules. Such processes produce impulsive "neuron-like" waveforms which can be coupled to nearest neighbors in a 1D, 2D, or 3D array. Input data can be represented by the distribution of dc bias level, initial charge, or coupling strength within the array. Information processing can be realized through the nonlinear dynamics produced by interactions between elements, which give rise to an evolution of complex patterns in the phase-state. In this paper, we discuss information processing for a model physical system based on Coulomb blockade in a 2D array of ultra-small tunnel junctions.

Using sonograms and phase retrieval we experimentally investigate the pattern dependence of the amplitude and phase dynamics, the linewidth enhancement factor, and the chirp of a data-modulated 1.55 /spl mu/m vertical-cavity surface-emitting laser (VCSEL).

Optical chaos communications met intensive research efforts during the last 10 years [1,2], resulting in 2004 in a field experiment in the metropolitan optical network of Athens [3]. Current efforts are now concentrating on device oriented architectures, in order to improve reliability, efficiency, and security. Among the few competitive approaches developed in the literature, the electro-optic one showed first its strong potential in terms of bit rate and decoding quality [4]. However, it was suffering from a few drawbacks in terms of technological complexity, due to the intensity superposition condition required by the mixing between the binary intensity message and the chaotic optical intensity carrier.

Summary form only given. Experimental and theoretical studies of intensity and phase dynamics of spatially coupled lasers have produced interesting results. The investigation of these particular nonlinear oscillators in various well-controlled spatial arrangements gives closer insight into the relations between symmetry and dynamics. The experimental systems we consider consist of two or three coupled lasers in one semi-monolithic Nd:YVO/sub 4/ microchip resonator, generated by pump beams delivered through a 2D arrangement of single-mode optical fibers.

The first harmonic of the cantilever deflection in the tapping-mode operation of an atomic force microscope (AFM) is analyzed using asymptotic methods for weakly nonlinear oscillators. The resulting amplitude and phase dynamical equations are obtained which characterize the transient behavior of tapping-mode dynamics. The steady state behavior is analyzed by examining the fixed points of the amplitude phase dynamics and a simple stability criterion is obtained. Further with a simple tip-sample interaction model, the experimentally observed discontinuous jumps in the amplitude versus tip-sample separation plots are explained and the regions of the interaction regime probed by the tip are investigated.

This paper measures sub-10 ps dynamics and magnitudes of gain and phase changes in InAs/GaAs quantum dots semiconductor optical amplifiers (SOA), Also this paper investigates the devices' suitability for different signal processing schemes. It is found that the pump-induced gain saturation dominates over nonlinear phase shifts, and is is inferred that QD SOAs are especially useful in signal processing schemes that exploit gain effects only.

We use femtosecond electron diffraction to capture the disordering of the solid and the emergence of liquid structure during strongly-driven laser melting. We discuss the impact of femtosecond pulse propagation on bright electron source design.

Already several years ago [3], it was suggested that a laser with optical injection can respond to external perturbations by emitting spikes if the perturbation overcomes a certain threshold, all the spikes being identical to each other, in complete analogy with the behaviour of the simple theta-model neuron [1]. The underlying explanation of this phenomenon lies in the isomorphism between a model of neuron [1] and that of an over-damped oscillator with periodic forcing (often called the Adler equation [2]). The control of these spikes was only achieved recently [5], while their observation goes back to [4]. Although there is an excellent agreement between these observations and the predictions of the reduction to pure phase dynamics of a laser model [3], it was soon noticed that whenever the pure phase dynamics reduction ceases to be valid, more complex dynamical phenomena can take place, leading to multipulse excitability [6]. This has led to the investigation of the behaviour of a semiconductor laser locked to an external forcing in response to external perturbations beyond the simple case of the Adler-like mode. We observe resonator features of the neuron-like system and multi-spike responses. Beyond improving the understanding of the neuron-like excitable dynamics of the laser with injected signal, our aim is to leverage the multipulse excitability and the resonator features of the optical neuron analogue for optical data processing and to provide possible insight about complex solitons interactions in forced oscillatory media [7, 8].

We used a method based on the modelling of phase dynamics to quantify the coupling strength between EEG channels. A distinction between the characteristics of coupling strength for the cases of limb movement and background activity is observed.

This paper introduces an extension of the Internal Model Control algorithm for efficient disturbance rejection. The approach is based on ideas from model based predictive control and diophantine equation derivation. As an illustration of the power of the extension, an example from the process industry is borrowed, namely a drum boiler. The process is challenging for control since it has an integrator and non-minimum phase dynamics. The performance of the proposed extension is compared against nominal IMC design and PID. The simulation results suggest that the proposed algorithm outperforms the other implementations in terms of effective disturbance rejections.

In this paper, we propose a framework for generating coordinated periodic movements of robotic systems with external inputs. We developed an adaptive pattern generator model that is composed of a two-factor observation model with a style parameter and phase dynamics with a phase variable. The style parameter controls the spatial patterns of the generated trajectories, and the phase variable controls its temporal profiles. To validate the effectiveness of our proposed method, we applied it to a simulated humanoid model to perform biped walking behaviors coordinated with observed walking patterns and the environment. The robot successfully performed stable biped walking behaviors even when the style of the observed walking pattern and the period were suddenly changed.

In this paper a control configuration for inverse-response processes is presented. It results in a Smith-type predictor scheme that aims to put the non-minimum phase dynamics out of the feedback loop. The design is carried out analytically by solving an H ∞ weighted optimization problem assuming a second order stable process with one positive zero. The performance of the proposed control configuration is compared by simulation with two different approaches to show its applicability.

This work concerns output trajectory tracking for the McDonnell Dougglas YAV68B Harrier VSTOL aircraft. This system can perform three distinct flight modes, and is modelled with a Lagrangian nonlinear hybrid non-minimum phase dynamics. Output tracking is solved using classical input/output linearization scheme associated to an internal dynamics stable inversion for each flight mode dynamics. The performance of the global switched control law is discussed and validated through simulation tests.

A phased-array two-dimensional CPLL array is introduced. The phase dynamics are analyzed and an approximate solution are obtained using the method of continuum phase. By detuning the free-running frequencies of edge elements, linear phase distribution can be achieved among the array. An experimental CPLL array circuit is also shown and tested, proceedings.