In this paper, we present the theoretical development to stabilize a class of uncertain time-delay system. The system under consideration is described in state space model containing distributed delay, uncertain parameters and disturbance. The main idea is to transform the system state into an equivalent one, which is easier to analyze its behavior and stability. Then, a computational method of robust controller design is presented in two parts. The first part is based on solving a Riccati equation arising in the optimal control theory. In the second part, the finite dimensional Lyapunov min-max approach is employed to cope with the uncertainties. Finally, we show how the resulting control law ensures asymptotic stability of the overall system.

Feature selection is an integral part of any pattern recognition system. Removal of redundant features improves the efficiency of a classifier as well as cut down the cost of future feature extraction. Recently neural network classifiers have become extremely popular compared to their counterparts from statistical theory. Some works on the use of artificial neural network as a feature selector have already been reported. In this work a simple feature selection algorithm has been proposed in which a fractal neural network, a modified version of multilayer perceptron, has been used as a feature selector. Experiments have been done with IRIS and SONAR data set by simulation. Results suggest that the algorithm with the fractal network architecture works well for removal of redundant informations as tested by classification rate. The fractal neural network takes lesser training time than the conventional multilayer perceptron for its lower connectivity while its performance is comparable to the multilayer perceptron. The ease of hardware implementation is also an attractive point in designing feature selector with fractal neural network.

This paper presents the generator polynomial matrices and the upper bound on the constraint length of punctured convolutional codes (PCCs), respectively. By virtue of these properties, we provide the puncturing realizations of the good known nonsystematic and systematic high rate CCs.

This paper treats the systematic design of chaos generators which are capable of generating continuous-time signals with prescribed probability density function and power density spectra. For a specific signal model a statistical analysis is performed such that the inverse problem, i. e. the calculation of the model parameters from prescribed signal characteristics, can be solved. Finally from the obtained model parameters and the model structure the signal generating system is constructed. The approach is illustrated by several examples.

Though considerable effort has recently been devoted to hardware realization of one-dimensional chaotic systems, the influence of implementation inaccuracies is often underestimated and limited to non-idealities in the non-linear map. Here we investigate the consequences of sample-and-hold errors. Two degrees of freedom in the design space are considered: the choice of the map and the sample-and-hold architecture. Current-mode systems based on Bernoulli Shift, on Tent Map and on Tailed Tent Map are taken into account and coupled with an order-one model of sample-and-hold to ascertain error causes and suggest implementation improvements.

We propose a hysteresis neural network system solving NP-Hard optimization problems, the N-Queens Problem. The continuous system with binary outputs searches a solution of the problem without energy function. The output vector corresponds to a complete solution when the output vector becomes stable. That is, this system does never become stable without satisfying the constraints of the problem. Though it is very hard to remove limit cycle completely from this system, we can propose a new method to reduce the possibility of limit cycle by controlling time constants.

This paper deals with a method of estimating the parameters and the order of a linear system using differential digital filters and the resultant. From the observed signals of the input and output of an objective system, we extract the differential signals from the zero order to an appropriate high order with the same phase characteristics, using several digital filters. On the assumption that the system order is known, we estimate the parameters of the transfer function and evaluate the estimation error bounds. We propose a criterion function generated by the product of the highest order coefficients and the resultant of the numerator and denominator of the estimated transfer function. Applying this criterion function, we can estimate the order of the objective system. The threshold corresponding to this criterion function is evaluated from the deviation in the frequency characteristics of the used differential filters and the error bound of the estimated parameters. In order to demonstrate the propriety of the proposed method, some numerical simulations are presented.

We propose and analyze a traffic model of a cellular radio communication network with an arbitrary cell connection and arbitrary probabilistic movement of mobiles between the cells. Our analytic model consists of birth-and-death processes for individual cells connected by the numerical adjustment of hand-off rates. This approximation is validated by simulation. We evaluate the probabilities of the immediate loss, the completion, and the forced termination during hand-off for an arbitrary call in the network. Our numerical examples reveal the cases in which the increase in the generation rate of new calls results in the increase in the loss probability without affecting much the probability of forced termination in a limited service area.

The concept of a schedulable region (SR) was introduced to characterize the capacity of a multiplexer and provide a separation between call-level and cell-level phenomena. In this paper, we present a framework and algorithm for real-time estimation of the schedulable region. A major problem associated with online estimation is that the objects of measurement are not fixed in the presence of call arrivals and departures. The invariance property is exploited to carry out measurements in the presence of call arrivals and departures. By virtue of it, the equivalent bandwidth could be defined on the condition of the number of each traffic class call in progress. Another important thing we consider here is that the search algorithm to estimate the effective bandwidth should be chosen depending on the arrival statistics and QOS constraints. The algorithms presented here have been implemented on an ATM switch.

This paper presents two competitive learning methods with the objective of avoiding the initial dependency of weight (reference) vectors. The first is termed the refractory and competitive learning algorithm. The algorithm has a refractory period: Once the cell has fired, a winner unit corresponding to the cell is not selected until a certain amount of time has passed. Thus, a specific unit does not become a winner in the early stage of processing. The second is termed the creative and competitive learning algorithm. The algorithm is presented as follows: First, only one output unit is prepared at the initial stage, and a weight vector according to the unit is updated under the competitive learning. Next, output units are created sequentially to a prespecified number based on the criterion of the partition error, and competitive learning is carried out until the ternimation condition is satisfied. Finally, we discuss algorithms which have little dependence on the initial values and compare them with the proposed algorithms. Experimental results are presented in order to show that the proposed methods are effective in the case of average distortion.

We investigate bifurcations of burst oscillations with rectangular waveform observed in a modified Bonhöffer-van der Pol equation, which is considered as a circuit model for neurons of a feeding rhythm generator. In particular, we clarify a mechanism of properties in a one-parameter graph on the period of oscillations, showing a staircase with hysteresis jumps, by studying a successive bifurcation process including a chain of homoclinic bifurcations. The occurrence of homoclinic bifurcations is confirmed by using the linking number of limit cycles related with the stable manifold through an equilibrium.

This paper shows the design of multi-stage fuzzy inference system with smaller number of rules based upon the optimization of rules by using the genetic algorithm. Since the number of rules of fuzzy inference system increases exponentially in proportion to the number of input variables powered by the number of membership function, it is preferred to divide the inference system into several stages (multi-stage fuzzy inference system) and decrease the number of rules compared to the single stage system. In each stage of inference only a portion of input variables are used as the input, and the output of the stage is treated as an input to the next stage. If we use the simplified inference scheme and assume the shape of membership function is given, the same backpropagation algorithm is available to optimize the weight of each rule as is usually used in the single stage inference system. On the other hand, the shape of the membership function is optimized by using the GA (genetic algorithm) where the characteristics of the membership function is represented as a set of string to which the crossover and mutation operation is applied. By combining the backpropagation algorithm and the GA, we have a comprehensive optimization scheme of learning for the multi-stage fuzzy inference system. The inference system is applied to the automatic bond rating based upon the financial ratios obtained from the financial statement by using the prescribed evaluation of rating published by the rating institution. As a result, we have similar performance of the multi-stage fuzzy inference system as the single stage system with remarkably smaller number of rules.

This paper develops an algorithm based on the Modular Approach to solve singly constrained separable discrete optimization problems (Nonlinear Knapsack Problems). The Modular Approach uses fathoming and integration techniques repeatedly. The fathoming reduces the decision space of variables. The integration reduces the number of variables in the problem by combining several variables into one variable. Computational experiments for "hard" test problems with up to 1000 variables are provided. Each variable has up to 1000 integer values.

Recently there have been several attempts to construct a Markov information source based on chaotic dynamics of the PLM (piecewise-linear-monotonic) onto maps. Study, however, soon informs us that Kalman's 1956 embedding of a Markov chain is to be highly appreciated. In this paper Kalman's procedure for embedding a prescribed Markov chain into chaotic dynamics of the PLM onto map is revisited and improved by using the PLM onto map with the minimum number of subintervals.

A novel analog-computation system using quantum-dot spin glass is proposed. Analog computation is a processing method that solves a mathematical problem by applying an analogy of a physical system to the problem. A 2D array of quantum dots is constructed by mixing two-dot (antiferromagnetic interaction) and three-dot (ferromagnetic interaction) systems. The simulation results show that the array shows spin-glass-like behavior. We then mapped two combinatorial optimization problems onto the quantum-dot spin glasses, and found their optimal solutions. The results demonstrate that quantum-dot spin glass can perform analog computation and solve a complex mathematical problem.

We developd a method of implementing a multiple-valued Hopfield network on electronic circuits by using single-electron circuit technology. The single-electron circuit shows quantized behavior in its operation because of the discrete tunnel transport of electrons. It can therefore be successfully used for implementing neuron operation of the multiple-valued Hopfield network. The authors developed a single-electron neuron circuit that can produce the staircase transfer function required for the multiple-valued neuron. A method for constructing a multiple-valued Hopfield network by combining the neuron circuits was also developed. A sample network was designed that solves an example of the quadratic integer-programming problem. And a computer simulation demonstrated that the sample network can converge to its optimal state that represents the correct solution to the problem.

This paper proposes applications of a code-division multiplexing technique to VLSI systems free from interconnection problems. We employ a pseudo-random orthogonal m-sequence carrier as a multiplexable information carrier to achieve efficient data transmission. Using orthogonal property of m-sequences, we can multiplex several computational activities into a single circuit, and execute in parallel using multiplexed data transmission with reduced interconnection. Also, randomness of m-sequences offers the high tolerance to interference (jamming), and suppression of dynamic range of signals while maintaining a sufficient signal-to-noise ratio (SNR). We demonstrate application examples of multiplex computing circuits, neural networks, and spread-spectrum image processing to show the advantages.

An analog vector quantization processor has been designed based on the neuron-MOS (νMOS) technology. In order to achieve a high integrating density, template information is merged into the matching cell (the absolute value circuitry) using the νMOS ROM technology. A new-architecture νMOS winner-take-all (WTA) circuit is employed for fully-parallel search for the minimum-distance vector. The WTA performs multi-resolution winner search with an automatic feedback gain control. A test chip having 256 16-element fixed template vectors has been built in a 1.5-µm double-polysilicon CMOS technology with the chip size of 7.2 mm 7.2 mm, and the basic operation of the circuits has been demonstrated.

In this paper we propose a heterogeneous agents model that represents speculative dynamics by using the synergetic approach. We consider the markets for three securities (a stock, a bond, and a foreign currency). Each market consists of two typical types of investors: fundamentalists and bandwagon traders. We show the characteristic patterns of speculative prices (speculative bubbles and speculative chaos) which are generated by trading between the fundamentalists and bandwagon traders.

We consider a circuit composed of linear capacitors, nonlinear resistors, and dc voltage sources and show the possibility that the total energy dissipated at resistors in the above circuit is smaller than the energy stored at capacitors. Linear passive circuits cannot possess such a property.