A new nonlinear time-domain op-amp macromodel using thresholdfunctions and digitally controlled network elements

A general-purpose nonlinear macromodel for the time-domain simulation of integrated circuit operational amplifiers (op amps), either bipolar or MOS, is presented. Three main differences exist between the macromodel and those previously reported in the literature for the time domain. First, all the op-amp nonlinearities are simulated using threshold elements and digital components, thus making them well suited for a mixed electrical/logical simulator. Secondly, the macromodel exhibits a superior performance in those cases where the op amp is driven by a large signal. Finally, the macromodel is advantageous in terms of CPU time. Several examples are included illustrating all of these advantages. The main application of this macromodel is for the accurate simulation of the analog part of a combined analog/digital integrated circuit     

Loss-free complex impedance network elements

The concept of loss-free complex impedance network elements (i.e., elements with active and reactive impedance components and yet loss-free) is introduced. Synthesis of such elements by means of switched-mode power converters with appropriate control has been demonstrated to be possible. Some possible power processing-related applications of these elements are indicated, such as improved matching between ac sources and loads and VAR compensation. It has further been demonstrated that loss-free complex impedance elements are suitable for modeling many power processing systems for the purpose of analysis as well as for design purposes.     

A transport network layer based on optical network elements

An approach to the realization of a broadband, flexible, multiwavelength transport network employing an optical network layer. The design methodology for a network demonstrator is presented, and the transmission, switching, line, and management/supervisory subsystems and components are described     

Synthesis of current-controlled nonlinear impedances

An integrable circuit technique is employed to obtain a class of current-controlled nonlinear impedances which is complementary to the recently proposed voltage-controlled class of nonlinear impedances.     

Realisation of basic 2-terminal network elements

The concept of basic 2-terminal network elements, which include resistors, inductors, capacitors and memristors as special cases, is introduced. It is shown that every basic 2-terminal network element can be realised by linear capacitors, linear inductors, linear algebraic controlled sources and one nonlinear resistor.     

Research on fuzzy neural network algorithms for nonlinear network traffic predicting

Neural networks provide massive parallelism,robust-ness ,and approxi mate reasoning,which are i mportantfor dealing with uncertain,inexact ,and ambiguous data,withill-defined problems and sparse data sets[1].It hasbeen proved that a neural network system …     

Synthesis of hysteresis-free nonlinear sharp-cutoff filters

This letter describes a technique for the synthesis of lowpass sharp-cutoff filters employing a nonlinear element. The nonlinearity is preselected, and the linear portion of the filter is synthetised around it. Such a procedure yields a very simple technique for synthetising the above class of nonlinear filters.     

On the cross-correlation function of a nonlinear network

It is shown that the shape of the cross-correlation function of a nonlinear network consisting of the cascade connection of a linear network, a memoryless nonlinearity (NL) and a second linear network is independent, in the Gaussian case, of the presence of NL only when NL is not even. An explicit expression for the cross-correlation function of such a system is then given.     

The movement from monoliths to component-based network elements

To be competitive in a rapidly growing market requires rapid upgrades to the performance and functionality of the network. One way to manage rapid upgrades of the network with minimum risk is to deploy equipment using a modular system architecture. Modularity allows a network operator to mix and match best of breed components to achieve the desired system rather than rely on vendors to implement specific technology before making crucial business decisions. This article begins with an overview of the current global movement toward standards that support network elements with modular system architecture. The story begins with university initiatives and the forming of OpenSig and IEEE P1520 more than five years ago, continuing with related and complementary initiatives by the Parlay Group, Softswitch Consortium, Multiservice Switching Forum, and several IETF working groups. Next, special attention is given to the component-based architecture of the Multiservice Switching Forum released in summer 2000. The trend of building network equipment from components with distinctly different functional specialties is described in three examples: media gateways, IP routers, and virtual IP routers. It is envisioned that component-based network infrastructure will spawn new markets for entrepreneurial developers, spurring competition and accelerating the creation of innovative solutions for all facets of global communications. The article concludes with a smorgasbord of new market opportunities     

A recurrent neural network for nonlinear convex optimization subject to nonlinear inequality constraints

This paper presents a novel recurrent neural network for solving nonlinear convex programming problems subject to nonlinear inequality constraints. Under the condition that the objective function is convex and all constraint functions are strictly convex or that the objective function is strictly convex and the constraint function is convex, the proposed neural network is proved to be stable in the sense of Lyapunov and globally convergent to an exact optimal solution. Compared with the existing neural networks for solving such nonlinear optimization problems, the proposed neural network has two major advantages. One is that it can solve convex programming problems with general convex inequality constraints. Another is that it does not require a Lipschitz condition on the objective function and constraint function. Simulation results are given to illustrate further the global convergence and performance of the proposed neural network for constrained nonlinear optimization.     

Cellular nonlinear network based on semiconductor tunneling nanostructure

We propose and analyze a cellular nonlinear network (CNN) based on a semiconductor nanostructure consisting of multiple layers of two semiconductors along with an incorporated quantum dot layer. An elementary logic cell of the proposed CNN consists of two resonant tunneling diodes connected in series through a quantum dot. The cell may be realized with multiple layers of two semiconductor materials with an embedded dot layer in between. The local interconnections of nanocells are achieved via tunneling between the neighboring quantum dots. Cells may be biased by the common column contacts, and only edge cells have individual I/O ports. Using approximate tunneling characteristics, we simulated network dynamics and found procedures leading to useful logic functionality. In order to illustrate network capabilities for image processing, we present examples of filtering, erosion, dilation, and edge detection carried out on a test image on a 400/spl times/269 cell template. The realization of a number of logic functions in one module is possible due to the incorporation of nonlinear (tunneling) elements for cell interconnections. The proposed CNN architecture for nanostructures demonstrates powerful computing potential that will be beneficial for many practical applications.     

Neural network architecture for solving nonlinear equation systems

A general neural network architecture is derived which can achieve ultrahigh-speed computation in solving large nonlinear equation systems. The computing time required for a solution is independent of the dimension of the equation system which is solved. The approach to solving equation systems by the method reported is very different from that of implementing in software a numerical analysis procedure and solution algorithm.<>     

Approximating nonlinear fading-memory operators using neural network models

A fading-memory system is defined as a system whose response map has unique asymptotic properties over some set of inputs. It is shown that any discrete-time fading-memory system can be uniformly approximated arbitrarily closely over a compact set of input sequences by uniformly approximating either its external or internal representation sufficiently closely. In other words, the problem of uniformly approximating a fading-memory system reduces to the problem of uniformly approximating continuous real-valued functions on compact sets. The perceptron is shown to realize a set of continuous real-valued functions that is uniformly dense on compacta in the set of all continuous functions. Using the perceptron to uniformly approximate the external and internal representations of a fading-memory system results, respectively, in simple nonlinear finite-memory and infinite-memory system models.     

Generalized neural network for nonsmooth nonlinear programming problems

In 1988 Kennedy and Chua introduced the dynamical canonical nonlinear programming circuit (NPC) to solve in real time nonlinear programming problems where the objective function and the constraints are smooth (twice continuously differentiable) functions. In this paper, a generalized circuit is introduced (G-NPC), which is aimed at solving in real time a much wider class of nonsmooth nonlinear programming problems where the objective function and the constraints are assumed to satisfy only the weak condition of being regular functions. G-NPC, which derives from a natural extension of NPC, has a neural-like architecture and also features the presence of constraint neurons modeled by ideal diodes with infinite slope in the conducting region. By using the Clarke's generalized gradient of the involved functions, G-NPC is shown to obey a gradient system of differential inclusions, and its dynamical behavior and optimization capabilities, both for convex and nonconvex problems, are rigorously analyzed in the framework of nonsmooth analysis and the theory of differential inclusions. In the special important case of linear and quadratic programming problems, salient dynamical features of G-NPC, namely the presence of sliding modes , trajectory convergence in finite time, and the ability to compute the exact optimal solution of the problem being modeled, are uncovered and explained in the developed analytical framework.     

The implementation of nonlinear dynamical systems with wavelet network

A dynamic wavelet network circuit implementation for modelling the nonlinear dynamical networks has been proposed in this study. The dynamical wavelet network includes static wavelet network with Mexican hat wavelet function, the voltage-controlled switches and capacitors. The circuit simulations have been done in Spice for the period-1 limit cycle, the spiral and double scroll attractors of the Chua's circuit.     

Neural network approximators in modelling dynamic nonlinear systems

Neural networks are often used to model nonlinear dynamic systems. The role of neural networks in the structure used in this paper is to approximate static nonlinear relations. To complete the nonlinear dynamic model a dynamic subsystem can add to the static nonlinearity in various forms. Various modelling procedures, however, exhibit a wide range of errors in estimating or predicting the model output. The paper discusses the sensitivity of modelling with respect both to the discretization and structural considerations in sampled data systems. As a tool for the sensitivity analysis equivalent input-output structures are introduced. The aim of the paper is to explain the relationships between the modelling error and the approximation error and to show how to reduce sensitivity by choosing proper system structure.