1,721,144 research outputs found
Analytical model for ideal generic memristor circuits based on the theory of Volterra
No full textThe use of memristors in future electronics is under deep investigation nowadays. The fields where these devices may find application are various. In fact there exist distinct types of memristors. Depending upon materials and operating conditions, different physical phenomena may emerge in these two-terminal devices, and the resulting dynamics may exhibit different characteristics. Progress in the investigation of the capabilities of memristors for integrated circuit applications may not leave aside the establishment of solid foundations on the circuit theoretic properties of these devices. Given the nonlinearity characterizing memristors, standard tools for the analysis and synthesis of circuits based upon them may be inappropriate. In this work we present rigorous studies elucidating the applicability of the Volterra series paradigm to model the nonlinear dynamics of a class of circuits with ideal generic memristors. The proposed nonlinear system theoretic approach provides closed-form analytical expressions for currents through all branches and voltages at any node in each circuit within the class. This study extends previous works where application of the Volterra theory was limited to circuits with ideal memristors
Exploring the Dynamics of Real-World Memristors on the Basis of Circuit Theoretic Model Predictions
No full textThe memristor represents the key circuit element for the development of the constitutive blocks of future non-volatile memory architectures and neuromorphic systems. However, resistance switching memories offer a plethora of further opportunities for the electronics of the future. By virtue of the compatibility between the well-established CMOS technology and the fabrication process of most memristors, the exploitation of the peculiar dynamic behaviour of resistance switching memories, which, in general, differ depending upon their material composition, may allow the development of new circuits, which, processing information in unconventional forms, may extend and/or complement the functionalities of state-of-the-art electronic systems. Further, the attractive capability of real-world non-volatile memristors to store and process information in the same physical nanoscale location open the fascinating opportunity to improve the low throughput of Von Neumann computing machines, due to the limited bandwidth of the bus transferring data between the memory and the central processing unit. Finally, the extreme sensitivity of their electrical behaviour to small changes in their initial condition/input and the intrinsic stochastic variability in their switching dynamics may be harnessed to develop innovative bio-signal sensors as well as new cryptographic circuits and systems. The derivation of accurate mathematical models for the electrical behaviour of real-world memristor nano-devices, and their later circuit- and system-theoretic investigation aimed at drawing a comprehensive picture of their peculiar nonlinear dynamic behaviour under the set of inputs and initial conditions expected of the application of interest are fundamental steps towards their conscious future use in integrated circuit design. With this in mind, the present paper adopts a powerful theoretic tool known as Dynamic Route Map to analyse some of the most reliable physics-based models of real-world resistance switching memories to reveal how a particular dynamic phenomenon, known as fading memory, and recently discovered in a tantalum oxide non-volatile memristor device fabricated at Hewlett Packard Labs, is ubiquitous at nanoscale. The physical mechanisms behind the emergence of history-erase effects in non-volatile memristor nano-devices is explained thoroughly for both the DC and the AC periodic excitation scenarios
Implementation of the XOR gate with two memristive neurons
No full textNeural networks enable the solution of various
complex problems, by building convoluted structures from simple
building blocks. In the past decade that more and more complex
neural networks were introduced and resulted higher accuracies
on commonly investigated benchmark datasets. As this trend
clearly demonstrates, the complexity of networks is typically
improved by increasing the number of neurons and layers in
their architecture, but higher complexity can also be achieved by
enriching the dynamics of the cells.
In this work we demonstrate that a simple memristor cellular
neural network containing two cells is able to solve the XOR
problem, which is not feasible for traditional neural networks
with only two cells. We train the parameters of this dynamical
system employing modern machine learning methods such as
gradient descent optimization. Our case study demonstrates how
the employment of complex circuit dynamics can extend the range
of solvable problems with a given number of neurons
The first ever real bistable memristors – Part I: theoretical insights on local fading memory
No full textIt has been recently shown that a current-controlled
extended memristor may exhibit bistable steady-state behavior
under dc as well as ac periodic stimuli. This brief employs standard
techniques from the nonlinear dynamics theory as well as
circuit and system theoretic concepts to explain the origin of the
asymptotic bistable behavior, which is the signature of a local
fading memory capability. Part II derives the first real memristor
featuring similar complex dynamics
Deep Memristive Cellular Neural Networks for Image Classification
No full textWe present simulation results of a deep cellular
neural network leveraging memristive dynamics to classify
images from standard datasets. We have investigated the use of
both volatile (NbO2-Mott) and non-volatile (TaOx) memristive
devices as output nonlinearity in neural networks. We simulated
deep neural networks using these devices and compared
their image classification accuracies on commonly investigated
datasets to traditional convolutional and cellular architectures
of similar complexity. Our results reveal that the exploitation
of memristive dynamics in cellular structures can increase
classification accuracy by more than 2.5 percent as compared
to the traditional convolutional implementations
Volterra model of a class of two-memristor circuits
No full textThe Volterra series approach represents a powerful
theoretical tool to model the dynamical behaviour of nonlinear
systems. In principle this paradigm characterizes completely the
dynamics of a system upon specification of an infinite set of
kernels. In practical implementations only a finite set of kernels
may be considered. The accuracy level of the Volterra seriesbased
modelling depends on the number of elements in this set.
The system output is then expanded as a finite algebraic sum of
multi-dimensional convolution integrals involving system input
and kernels. Memristors are nonlinear dynamical devices which
promise to improve and/or extend the functionalities of state-ofthe-
art electronic systems. However, classical circuit and system
theory techniques for modelling linear devices are not suitable for
them. The Volterra series paradigm has been recently adopted
to model a class of one-memristor circuits. In this work the
technique is extended to the analytical representation of twomemristor
circuits. The agreement between Volterra series-based
predictions and numerical integration results on an exemplary
circuit from the class provides evidence for the accuracy of the
modelling paradigm
Pattern Formation in an M-CNN Structure Utilizing a Locally Active NbOx Memristor
No full textIn this work, we present an application of the local activity theory by
demonstrating the emergence of complex patterns in aMemristor Cellular Nonlinear
Network (M-CNN) structure. The proposed M-CNN structure consists of identical
memristive cells, which are resistively coupled to each other in a two-dimensional
(2-D) grid form. Each cell contains a locally active NbOx memristor, a DC voltage
source, a bias resistor, and a capacitor. Firstly, the locally active memristor together
with its AC equivalent circuit is introduced. Secondly, the stability analysis of the
single cell is performed. Then, the opportune parameter space, associated with local
activity, edge-of-chaos, and sharp-edge-of-chaos domains, is determined in terms of
cell characteristics, namely, the DC operating point, the capacitor, and the coupling
resistor. Precisely, all the derivations are performed parametrically and a simplified
generic memristor model is employed to enhance the simulation speed. Simulation
results successfully show that complexity can be observed in resistively coupled
M-CNNs utilizing locally active memristors
The First Ever Real Bistable Memristors - Part II: Design and Analysis of a Local Fading Memory System
No full textPart I has provided theoretical insights on the
concept of local fading memory and analyzed a purely mathematical
memristor model that, under dc and ac periodic stimuli,
experiences memory loss in each of the basins of attraction of two
locally stable state-space attractors. This brief designs the first
ever real memristor with bistable stationary dc and ac behavior. A
rigorous theoretical analysis unveils the key mechanisms behind
the emergence of nonunique asymptotic dynamics in this novel
electronic circuit, falling into the class of extended memristors
Mathematical Investigation of Static Pattern Formation with a Locally Active Memristor Model
No full textWe present the mathematical investigation of static pattern formation in a Memristor Cellular Nonlinear Network (M-CNN), in consideration of the theory of local activity. The M-CNN has a planar grid form composed of identical memristive cells, which are purely resistively coupled to each other. The single cell contains a DC voltage source, a bias resistor, and a locally active memristor in parallel with a capacitor. The memristor model employed has a simple generic form which helps to reduce the simulation time, and has a functional AC equivalent circuit which facilitates further calculations. We adopt a circuit theoretical approach for the stability analysis of the single cell and a 3-cell ring configuration, as well as the examination of local activity, edge-of-chaos, and sharp-edge-of-chaos domains, which helps us to interpret the results in a better way. The emergence of static patterns is successfully confirmed by simulating the proposed resistively coupled M-CNN utilizing locally active memristors
Theory and Technology of Memristive Devices
No full textThis manuscript provides a comprehensive overview of the theory of memristors, including a detailed discussion of their nanotechnologies. A synergetic cooperation between theoreticians and experimenters is a fundamental aspect to foster progress in this multidisciplinary research field, which, as outlines in the last part of the paper, may open up various exciting opportunities for the electronics of the future
- …
