Additional generalisation scheme of active RC allpass filters

A generalisation scheme of the active RC allpass networks with one differential input operational amplifier and three RC 2-ports is presented. The 2-ports are unbalanced, but not earthed, and only the voltage transfer functions of these networks are used in this generalisation, thus making the analysis of the circuit and the circuit transfer functions considerably easier.     

Complex-variable distribution theory for Laplace and z transforms

The author proposes a generalisation of the theory of generalised functions, also known as the theory of distributions, by extending the theory to include generalised functions of a complex variable, both in the complex plane associated with continuous-time functions and that with discrete-time functions. The generalisation provides, among others, mathematical justifications of the properties of recently introduced generalised Dirac-delta impulses, using the principles of distribution theory. Properties of generalised functions of a complex variable are explored both in the Laplace domain associated with continuous-time functions and the z domain associated with discrete-time functions. Shifting of distributions, scaling, derivation, convolution with distributions and convolution with ordinary functions are evaluated in Laplace and z domains. Three-dimensional generalisations of sequences leading to generalised impulses, and of test functions in Laplace and z domains are presented. New expanded Laplace and z transforms are obtained using the proposed generalisation.     

Numerical simulation of the spectral response and the I(V) characteristics of solar cells

The spectral response of p-n silicon solar cells is calculated as a generalisation of results for monochromatic light illumination. The total short-circuit current is calculated from the spectral response and used for I(V)plottings.     

Architecture of multichannel DQDB network

The authors explain the basic architecture of the multichannel DQDB (distributed queue dual bus) network which is a generalisation of the single channel DQDB to the multichannel case by wavelength division multiplexing technology     

Structural reliability of centralised tree network

An efficient algorithm is introduced for assigning a terminal as a centre in a tree network that makes the network reliable under a given reliability index. Some generalisation for multiple-centre assignment is provided.     

Boundary and volume filtering of linear distributed-parameter systems

The optimal-filtering problem of linear distributed-parameter systems corrupted by boundary and/or volume stochastic disturbances is considered. The optimal filter is shown to be a generalisation of that previously derived for pure volume noise.     

Digital kernel perceptron

It is shown that a kernel-based perceptron can be efficiently implemented in digital hardware using very few components. Despite its simplicity, the experimental results on standard data sets show remarkable performance in terms of generalisation error     

Optimal nonlinear filtering with constraints

The impulse response of nonlinear systems with limitation of band has been considered. On the base of the minimum mean square error criterion the generalisation of the Wiener-Hopf's equation has been calculated.     

Emerging Nonlinear Optical Fibers: Revised Fundamentals, Fabrication and Access to Extreme Nonlinearity

We review the properties of conventional and emerging nonlinear fibers. More specifically, we describe advances in soft glass microstructured fibers, including fabrication, the demonstration of extreme nonlinearity and a generalisation of the underpinning nonlinear theory.      

Nonlinear time series modelling and prediction using Gaussian RBFnetwork with evolutionary structure optimisation

An evolutionary structure optimisation method for the Gaussian radial basis function network is presented for modelling and predicting nonlinear time series. The generalisation performance is significantly improved with a much smaller network, compared with that of the previous clustering and least square learning method     

Autoregressive modelling in vector spaces: An application to narrow-bandwidth spectral estimation

A generalisation of the Burg algorithm to vector spaces is described. This can be used to subsample time series that have narrow frequency bandwidth, thereby reducing the computation time and the complexity of the autoregressive model required. The technique is preferable to conventional subsampling.     

New method of obtaining reduced-order models for linear multivariable systems

A new approach for model reduction is presented which may be considered as the generalisation of the Padé-approximation technique. It is based on the generation of successive partial realisations of a high-order multivariable system in state-variable form, and is computationally straightforward.     

Replacing resamplers in filter banks without losing perfect reconstruction

The condition under which the resamplers in a filter bank can be replaced without losing perfect reconstruction is presented. This is the generalisation of the common knowledge that removing resamplers and/or inserting unimodular resamplers do not destroy the perfect reconstruction of filter banks.     

Multiple fault simulation in linear active circuits

An efficient algorithm is described for simulating the effects of multiple faults in linear active circuits. The process is also applicable to the multiparameter large-change sensitivity analysis of linear circuits. It is based on a generalisation of the branch-at-a-time circuit-analysis algorithm of Kron and Branin.     

Large-signal effects in an optical BSO amplifier

It is shown that the gain obtainable by coherent amplification using two-wave mixing in BSO crystals is limited by nonlinear effects. A model based on a generalisation of existing theories is presented, giving the dependence of gain coefficient on input beam ratio and leading to good agreement with experimental results.     

Solving two-spiral problem through input data encoding

The authors show that the two-spiral problem can be easily solved using a standard back propagation neural network by properly encoding the raw input data. The authors also examine and compare several data encoding schemes for use in neural networks to improve the training efficiency and generalisation property     

Leaky rays on optical fibres of arbitrary (circularly symmetric) index profiles

The local plane-wave decomposition approach used to analyse optical fibres of arbitrary refractive-index profiles has been extended to include the case of so called `leaky? rays. The result thus obtained for acceptance angle represents a generalisation of results derived previously for simple forms of the profile by geometrical-optics methods.     

Aggregation matrices for a class of low-order models for large-scale systems

A simple technique for aggregation is established for model reduction by projection. This leads to a direct formula for deriving the aggregation matrix. It is shown that aggregation is a generalisation of projection, and the work of Aoki is related to that of Mitra, Davison and Chidambara.     

General rule for the compensation of the sensitivity of a network to its active devices

A technique is described which uses additional amplifiers in existing RC active networks to reduce the effect on transfer functions of simultaneous similar changes in amplifier gains. This method is a generalisation of that suggested in an earlier paper, having the advantage of being applicable to a wider range of networks.     

General relations for nonperiodic bandlimited signals in three-dimensional electron flow

Relations for bandlimited signals in electron flow are presented that may be considered as a generalisation of power relations for sinusoidal signals. Inequalities defining limiting values for the power densities are obtained when signals exist in three frequency bands only which correspond to the frequencies in a parametric amplifier.