The performance improvement of the partial response maximum-likelihood (PRML) system for (1, 7) run-length limited (RLL) code is studied. As a new PRML system, PR (1, 1, 0, 1, 1) system called modified E2PR4 (ME2PR4 ) followed by Viterbi detector for (1, 7) RLL code is proposed. At first, a determination method of the tap weights in transversal filter to equalize to PR (1, 1, 0, 1, 1) characteristic taking account of a noise correlation is described. And the equalization characteristics of the transversal filter are evaluated. Then, a Viterbi detector for ME2PR4 utilizing the constraint of run-length of (1, 7) RLL code is presented. Finally, the bit-error rate is obtained by computer simulation and the performance is compared with that of the conventional PRML systems called PR4, EPR4 and E2PR4 systems with Viterbi detector. The results show that among these systems our system exhibits the best performance and the SNR improvement increases with the increase in the linear density.

This paper describes some dynamical properties of higher order neural networks with decreasing energy functions. First, we will show that for any symmetric higher order neural network which permits only one element to transit at each step, there are only periodic sequences with the length 1. Further, it will be shown that for any higher order neural network, with decreasing energy functions, which permits all elements to transit at each step, there does not exist any periodic sequence with the length being over k + 1, where k is the order of the network. Lastly, we will give a characterization for higher order neural networks, with the order 2 and a decreasing energy function each, which permit plural elements to transit at each step and have periodic sequences only with the lengh 1.

The capability of generalization is the most desirable property of a learning system. It is well known that to achieve good generalization, the complexity of the system should match the intrinsic complexity of the problem to be learned. In this work, introduction of fractal connection structure in nonlinear learning systems like multilayer perceptrons as a means of improving its generalization capability in classification problems has been investigated via simulation on sonar data set in underwater target classification problem. It has been found that fractally connected net has better generalization capability compared to the fully connected net and a randomly connected net of same average connectivity for proper choice of fractal dimension which controlls the average connectivity of the net.

In this paper, Chay's bursting pancreatic β-cell model is updated to include a role for [Ca2+]ER, the luminal calcium concentration in the endoplasmic reticulum (ER). The model contains a calcium current which is activated by voltage and inactivated by [Ca2+]i. It also contains a cationic nonselective current (INS) that is activated by depletion of luminal Ca2+ in the ER. In this model, [Ca2+]ER oscillates slowly, and this slow dynamic drives electrical bursting and the [Ca2+]i oscillations. This model is capable of providing answers to some puzzling phenomena,which the previous models could not (e. g., why do single pancreatic β-cells burst with a low frequency while the cells in an islet burst with a much higher frequency ?). Verification of the model prediction that [Ca2+]ER is a primary oscillator that drives electrical bursting and [Ca2+]i oscillations in pancreatic β-cells awaits experimental testing. Experiments using fluorescent dyes such as mag-fura-2-AM could provide relevant information.

Conventional learning algorithms are considered to be a sort of estimation of the true recognition function from sample patterns. Such an estimation requires a good assumption on a prior distribution underlying behind learning data. On the other hand the human being sounds to be able to acquire a better result from an extremely small number of samples. This forces us to think that the human being might use a suitable prior (called presupposition here), which is an essential key to make recognition machines highly flexible. In the present paper we propose a framework for guessing the learner's presupposition used in his learning process based on his learning result. First it is pointed out that such a guess requires to assume what kind of estimation method the learner uses and that the problem of guessing the presupposition becomes in general ill-defined. With these in mind, the framework is given under the assumption that the learner utilizes the Bayesian estimation method, and a method how to determine the presupposition is demonstrated under two examples of constraints to both of a family of presuppositions and a set of recognition functions. Finally a simple example of learning with a presupposition is demonstrated to show that the guessed presupposition guarantees a better fitting to the samples and prevents a learning machine from falling into over learning.

It is shown that five optimal and one quasioptimal binary codes with respect to the Griesmer bound can be obtained from cyclic codes over GF(2fm). An [m(2em - 1), em, 2em-1m] code, a [3(22e - 1), 2e, 322e-1] code, a [2(22e - 1), 2, (22e+2 - 4)/3] code, a [3(22e - 1), 2, 22e+1 - 2] code, and a [3(22e - 1), 2(e+1), 322e-1 - 2] code are optimal and a [2(22e - 1), 2(e + 1), 22e - 2] code is quasi-optimal.

This paper proposes linear predictive transmission diversity for TDMA/TDD personal communication systems and evaluates the effects of fading correlation and unequal average signal power Rayleigh fading on these system. The average bit error rate (BER) performance is calculated by computer simulation and the BER of zero order prediction is theoretically analyzed. The performance degradation caused by the error from prediction, fading correlation, and unequal average signal power is found to be almost independent of each other.

In the paper a new type of Multilayer Perceptron, developed in Quaternion Algebra, is adopted to realize short-time prediction of chaotic time series. The new introduced neural structure, based on MLP and developed in the hypercomplex quaternion algebra (HMLP) allows accurate results with a decreased network complexity with respect to the real MLP. The short term prediction of various chaotic circuits and systems has been performed, with particular emphasys to the Chua's circuit, the Saito's circuit with hyperchaotic behaviour and the Lorenz system. The accuracy of the prediction is evaluated through a correlation index between the actual predicted terms of the time series. A comparison of the performance obtained with both the real MLP and the hypercomplex one is also reported.

Deterministic nonlinear prediction is applied to both artificial and real time series data in order to investigate orbital-instabilities, short-term predictabilities and long-term unpredictabilities, which are important characteristics of deterministic chaos. As an example of artificial data, bimodal maps of chaotic neuron models are approximated by radial basis function networks, and the approximation abilities are evaluated by applying deterministic nonlinear prediction, estimating Lyapunov exponents and reconstructing bifurcation diagrams of chaotic neuron models. The functional approximation is also applied to squid giant axon response as an example of real data. Two metnods, the standard and smoothing interpolation, are adopted to construct radial basis function networks; while the former is the conventional method that reproduces data points strictly, the latter considers both faithfulness and smoothness of interpolation which is suitable under existence of noise. In order to take a balance between faithfulness and smoothness of interpolation, cross validation is applied to obtain an optimal one. As a result, it is confirmed that by the smoothing interpolation prediction performances are very high and estimated Lyapunov exponents are very similar to actual ones, even though in the case of periodic responses. Moreover, it is confirmed that reconstructed bifurcation diagrams are very similar to the original ones.

A new type of guided-wave polarization splitter is proposed for the operation at optical frequencies. The basic structure of the device is a bifurcating waveguide fabricated in a uniaxial crystalline material such as LiNbO3. The splitting behavior of optical waves into two waves with mutually perpendicular directions of polarization by an optically anisotropic material is utilized in the branching section of the present polarization splitter. Once of the conspicuous features of the device is free of any electrical control via the electro-optic effects. Some numerical results obtained with the finite difference beam propagation method indicate that extinction ratios better than 20dB are possible of realization for both TE and TM modes.

We propose a theory of general frame multiresolution analysis (GFMRA) which generalizes both the theory of multiresolution analysis based on an affine orthonormal basis and the theory of frame multiresolution analysis based on an affine frame to a general frame. We also discuss the problem of perfectly representing a function by using a wavelet frame which is not limited to being of affine type. We call it a "generalized affine wavelet frame." We then characterize the GFMRA and provide the necessary and sufficient conditions for the existence of a generalized affine wavelet frame.

Synthesis tools are now extensively used in the VLSI circuit design process. They allow a much higher design productivity, but the designer often does not directly control the circuit structure. Thus, when circuits are dedicated to dependable applications, designers have difficulties in implementing manually the devices needed to obtain fault detection or tolerance capabilities. The ASYL-SdF System has been developed over the last few years in order to avoid this break in the design flow, and to facilitate the designer's work when dependability is targeted. This paper gives an overview of the resulting tool, its synthesis flow for fault detection and fault tolerance in Finite State Machines, its limitations and the current developments. Actual circuit implementation results are given in terms of area overheads, expected reliability and experimental fault detection coverage.

The electromagnetic force of evanescent field acting on dielectric slab is studied with the use of Maxwell stress tensor. The results show that dielectrics slab may receive always an attractive force when the incident wave is evanescent field while a pressure or an attractive force when the wave is propagating one. The magnitude of the attractive force by evanescent field is much larger than that of the propagating wave. And here some numerical examples are given.

This study clarifies the effects of network complexity and network map transformation on the ability of network managers to use graphic network displays. Maps of Japan and the United States with outlines of their respective prefectures or states were displayed on a CRT. Each map displayed a fictitious network of nodes and their interconnections. These networks were two-level hierarchical and non-meshed, meaning that each low-level node was connected to a single high-level node, but not all high-level nodes were linked together. The subjects, task was to identify a path between two low-level nodes. In each trial, two low-level nodes were highlighted, and the subject attempted to find the shortest path between these nodes. This was done by using a mouse to select intermediate nodes. Completing a path required a minimum of 4 node traversals. Three variables were manipulated. First, the number of nodes was defined as the total number of low-level nodes in a network (70, 150, or 200). The second variable was the level of transformation. Very densely populated areas of the maps were systematically transformed to reduce congestion. There were three levels of transformation. The final variable was the country map used, that is, the map of Japan and the map of the United States. Several behavioral measures were used. The most informativ. appeared to be the time required to complete a path (the response time), and how often subjects returned to previous portions of a path (back-ups). For both of these measures, the data pattern was essentially the same. Increasing the number of nodes hurts performance. This was particularly pronounced when the map of Japan was tested. However, as the level of transformation increased, this effect was substantially reduced or completely eliminated. The results are discussed in terms of engineering rules and guidelines for designing graphical network representations.

An efficient algorithm is proposed for finding all solutions of piecewise-linear resistive circuits The algorithm is based on the idea of "contraction" of the solution domain using a sign test. The proposed algorithm is efficient because many large super-regions containing no solution are eliminated in early steps.

Finite-difference time-domain (FDTD) analysis is performed to evaluate the distributions of specific absorption rate (SAR) in a human head during use of a handheld portable telephone. A heterogeneous head model has been assumed which is comprised of 273 108 cubic cells 2.5 mm on a side, with the electrical properties of anatomical equivalents. A handset model has been assumed to be a metal box with either a quarter-wavelength monopole or a half-wavelength dipole operating at 900 MHz or 1.5 GHz. The maximum local SARs in the head are evaluated under various exposure conditions. The dependence of the maximum local SARs on the difference in the structures or parameters of the model, i.e. the distance between the antenna and the head, the heterogeneity of the head, the antenna type, the volume of the smoothing region of the local SAR value, skin electrical constants, and the presence or absence of auricles, are examined. It is shown that the heterogeneity of the head barely affect the maximum local SAR when the telephone is located sufficiently close to the head. It is also shown that the electrical constants of skin which has lower conductivity provide the lower maximum local SAR in the head while the maximum local SAR within the brain is not significantly affected. The auricle which lies in closest proximity to the antenna is shown to have significant effect on the maximum local SAR. It is suggested that the presence of the auricle enhances the maximum local SAR by a factor that is 1.7-2.4 larger than the model without auricles.

A new type of noncausal stochastic model is proposed to represent a random image with double peak spectrum. The model based on the assumption that the double peak spectrum is expressed by a product of two spectra located at two symmetric positions in the 2D spatial frequency space. Estimation of model parameters is made by means of minimizing the "whiteness" which was proposed in authors' previous work. In a simulation for model estimation we make use of computer-generated random images with double peak spectrum. Comparing this with the estimation by a causal model, we demonstrate that the present method can better estimate not only the spectral peak location but also the spectral shape. The proposed model can be extend to an image model with multl-peak spectrum. However, Increase of parameters makes the model estimation more difficult We try a model with triple peak spectra since a real texture image usually possesses a spectral peak at the origin besides the two peaks. A result shows that the estimation of three spectral positions are good enough, but their spectral shapes are not necessarily satisfactory. It is expected that the estimation of multi-peaked spectral model can be made better by improving the process of minimizing the "whiteness."

In this paper, we propose a novel SPICE oriented steady-state analysis of nonlinear circuits based on the circuit partition technique. Namely, a given circuit is partitioned into the linear and nonlinear subnetworks by the application of the substitution theorem. Each subnetwork is solved using SPICE simulator by the different techniques of AC analysis and transient analysis, respectively, whose steady-state reponse is found by an iteration method. The novel points of our algorithm are as follows: Once the linear subnetworks are solved by AC analysis, each subnetwork is replaced by a simple equivalent RL or RC circuit at each frequency component. On the other hand, the reponse of nonlinear subnetworks are solved by transient analysis. If we assume that the sensitivity circuit is approximated at the DC operational point, the variational value will be also calculated from a simple RL ro RC circuit. Thus, our method is very simple and can be also applied to large scale circuits, effciently. To improve the convergency, we introduce a compensation technique which is usefully applied to stiff circuits containing components such as diodes and transistors.

This paper describes a waveform relaxationbased coupled lossy transmission line circuit simulator DESIRE3T+. First, the generalized method of characteristics (GMC) is reviewed, which replaces a lossy transmission line with an equivalent disjoint network. Next, the generalized line delay window (GLDW) partitioning technique is proposed, which accelerates the transient analysis of the circuits including transmission lines replaced by GMC model. Finally GMC model and GLDW technique are implemented in hte relaxation-based circuit simulator DESIRE3T+ which can analyze bipolar transistor circuits by using the dynamic decomposition technique, and the performance is estimated.

For very low bit-rate video coding such as under 64 kbps, it is unreasonable to encode and transmit all the information. Thus, it is very important to choose the "important" information and encode it efficiently. In this paper, we first propose an image separation-composition method to solve this problem. At the encoder, an image is separated into a low-frequency part and two (horizontal and vertical) edge parts, which are considered as "important" information for human visualization. The low-frequency part is encoded by using block DCT and linear quantization. And the edges are selected by their values and encoded by using Chain coding to remain the most of the important parts for human visualization. At the decoder, the image is reconstructed by first generating the high-frequency parts from the horizontal and vertical edge parts, respectively, and then applying the inverse wavelet transform to the low frequency part and high frequency parts. This composition algorithm has less computational complexity than the conventional analytic/synthetic algorithms because it is not based on iterating approach. Moreover, to reduce the temporal redundancy efficiently, we propose a hierarchical motion detection and a motion interpolation /extrapolation algorithm. We detect motion vectors and motion regions between two reconstructed images and then predict the motion vectors of the current image from the previous detected motion vectors and motion regions by using the interpolation/extrapolation both at the encoder and at the decoder. Therefore, it is unnecessary to transmit the motion vectors and motion regions. This algorithm reduces not only the temporal redundancy but also bit-rates for coding side information . Furthermore, because the motion detection is completely syntax independent, any type of motion detection can be used. We show some simulation results of the proposed video coding algorithm with the coding bit-rate down to 24 kbps and 10 kbps.