One of the major open issues in neural network research includes a Network Designing Problem (NDP): find a polynomial-time procedure that produces minimal structures (the minimum intermediate size, thresholds and synapse weights) of multilayer threshold feed-forward networks so that they can yield outputs consistent with given sample sets of input-output data. The NDP includes as a sub-problem a Network Training Problem (NTP) where the intermediate size is given. The NTP has been studied mainly by use of iterative algorithms of network training. This paper, making use of both rate distortion theory in information theory and linear algebra, solves the NDP mathematically rigorously. On the basis of this mathematical solution, it furthermore develops a mathematical solution Procedure to the NDP that computes the minimal structure straightforwardly from the sample set. The Procedure precisely attains the minimum intermediate size, although its computational time complexity can be of non-polynomial order at worst cases. The paper also refers to a polynomial-time shortcut to the Procedure for practical use that can reach an approximately minimum intermediate size with its error measurable. The shortcut, when the intermediate size is pre-specified, reduces to a promising alternative as well to current network training algorithms to the NTP.

This paper describes an MOS charge-mode version of a T-Model neural-based PCM encoder. The neural-based PCM encoding networks are designed, simulated and implemented using MOS charge-mode circuits. Simulation results are given for both the T-Model and the Hopfield model CMOS charge-mode PCM encoders, and demonstrate the T-Model neural-based one performs the PCM encoding perfectly, while the Hopfield one fails to.

The number of solutions of a nonlinear equation x = sgn(Wx) is discussed. The equation is derived for the determination of equilibrium points of a kind of Hopfield neural networks. We impose some conditions on W. The conditions correspond to the case where a Hopfield neural network has n neurons arranged on a ring, each neuron has connections only from k preceding neurons and the magnitude of k connections decrease as the distance between two neurons increases. We show that the maximum number of solutions for the above case is extremely few and is independent of the number of neurons, n, if k is less than or equal to 4. We also show that the number of solutions generally increases exponentially with n by considering the case where k = n-1.

Cellular Neural Networks (CNNs) are nonlinear dynamic array processors with mainly local interconnections. In most of the applications, the local interconnection pattern, called cloning template, is translation invariant. In this paper, an optimal ring-coding method for rotation invariant description of given set of objects, is introduced. The design methodology of the templates based on the ring-codes and the synthesis of CNN analogic algorithms to detect standing and moving objects in a rotationally invariant way, discussed in detail. It is shown that the algorithms can be implemented using the CNN Universal Machine, the recently invented analogic visual microprocessor. The estimated time performance and the parallel detecting capability is emphasized, the limitations are also thoroughly investigated.

In this paper, a description of the Hopfield/Tank model used for the telecommunication network routing problem is given at first. And then through the "static" (i.e. the eigenvalue and the eigenspace of the connection matrix) and the dynamic analysis of the model, it has been found that the model has the faster rate to converge to the optimal or sub-optimal solutions from an initial value. Therefore the quality of the final solutions can be guaranteed. The influence of the initial value to the final solutions is also concerned in this paper. The simulation results are given at the end of this paper.

We propose an algorithm for image decomposition based on Hadamard functions, realized by answer-in-weights neural network, which has simple architecture and is explored with steepest decent method. This scheme saves memory consumption and it converges fast. Simulations with least mean square (LMS) and absolute mean (AM) errors on a 128128 image converge within 30 training epochs.

This paper presents an improved neural network for channel assignment problems in cellular mobile communication systems in the new co-channel interference model. Sengoku et al. first proposed the neural network for the same problem, which can find solutions only in small size cellular systems with up to 40 cells in our simulations. For the practical use in the next generation's cellular systems, the performance of our improved neural network is verified by large size cellular systems with up to 500 cells. The newly defined energy function and the motion equation with two heuristics in our neural network achieve the goal of finding optimum or near-optimum solutions in a nearly constant time.

In case of object recognition using 3-D configuration data, the scale and poses of the object are important factors. If they are not known, we can not compare the object with the models in the database. Hence we propose a strategy for object recognition independently of its scale and poses, which is based on Hopfield neural network. And we also propose a strategy for estimation of the camera motion to reconstruct 3-D configuration of the object. In this strategy, the camera motion is estimated only with the sequential images taken by a moving camera. Consequently, the 3-D configuration of the object is reconstructed only with the sequential images. And we adopt the multiple regression analysis for estimation of the camera motion parameters so as to reduce the errors of them.

A novel method is presented for designing discrete coeffcient FIR linear phase filters using Hopfield neural networks. The proposed method is based on the minimization of the energy function of Hopfield neural networks. In the proposed method, the optimal solution for each filter gain factor is first searched for, then the optimal filter gain factor is selected. Therefore, a good solution in the specified criterion can be obtained. The feature of the proposed method is that it can be used to design FIR linear phase filters with different criterions simultaneously. A design example is presented to demonstrate The effectiveness of the proposed method.

This paper described discrete time Cellular Neural Networks (DT-CNN) with two types of neuron circuits for image coding from an analog format to a digital format and their VLSI implementations. The image coding methods proposed in this paper have been investigated for a purpose of transmission of a coded image and restoration again without a large loss of an original image information. Each neuron circuti of a network receives one pixel of an input image, and processes it with binary outputs data fed from neighboring neuron circuits. Parallel dynamics quantization methods have been adopted for image coding methods. They are performed in networks to decide an output binary value of each neuron circuit according to output values of neighboring neuron circuits. Delayed binary outputs of neuron circuits in a neighborhood are directly connected to inputs of a current active neuron circuit. Next state of a network is computed form a current state at some neuron circuits in any time interval. Models of two types of neuron circuits and networks are presented and simulated to confirm an ability of proposed methods. Also, physical layout designs of coding chips have been done to show their possibility of VLSI realizations.

The dynamics of an artificial neural network derived from a biological system, and its two applications to engineering problems are examined. The model has a multi-layer structure simulating the primary and secondary components in the olfactory system. The basic element in each layer is an oscillator which simulates the interactions between excitatory and inhibitory local neuron populations. Chaotic dynamics emerges from interactions within and between the layers, which are connected to each other by feedforward and feedback lines with distributed delays. A set of electroencephalogram (EEG) obtained from mammalian olfactory system yields aperiodic oscillation with 1/f characteristics in its FFT power spectrum. The EEG also reveals abrupt state transitions between a basal and an activated state. The activated state with each inhalation consists of a burst of oscillation at a common time-varying instantaneous frequency that is spatially amplitude-modulated (AM). The spatial pattern of the activated state seems to represent the class of the input ot the system, which simulates the input from sensory receptors. The KIII model of the olfactory system yields sustained aperiodic oscillation with "1/f" spectrum by adjustment of its parameters. Input in the form of a spatially distributed step funciton induces a state transition to an activated state. This property gives the model its utility in pattern classification. Four different methods (SD, RMS, PCA and FFT) were applied to extract AM patterns of the common output wave forms of the model. The pattern classification capability of the model was evaluated, and synchronization of the output wave form was shown to be crucial in PCA and FFT methods. This synchronization has also been suggested to have an important role in biological systems related to the information extraction by spatiotemporal integration of the output of a transmitting area of cortex by a receiving area.

A novel algorithm, as an advanced Hybrid Channel Assignment strategy, for channel assignment problem in a cellular system is proposed. A difference from the conventional Hybrid Channel Assignment method is that flexible fixed channel allocations which are variable through the channel assignment can be performed in order to cope with varying traffic. This strategy utilizes the Channel Rearrangement technique using the artificial neural network algorithm in order to enhance channel occupancy on the fixed channels. The strategy is applied to two simulation models which are the spatial homogeneous and inhomogeneous systems in traffic. The simulation results show that the strategy can effectively improve blocking probability in comparison with pure dynamic channel assignment strategy only with the Channel Rearrangement.

This paper describes new modeling methods combining neural network and hidden Markov model applicable to modeling a time series such as speech signal. The idea assumes that the sequence is nonstationary and is a nonlinear autoregressive process whose parameters are controlled by a hidden Markov chain. One is the model where a non-linear predictor composed of a multi-layered neural network is defined at each state, another is the model where a multi-layered neural network is defined so that the path from the input layer to the output layer is divided into path-groups each of which corresponds to the state of the Markov chain. The latter is an extended model of the former. The parameter estimation methods for these models are shown, and other previously proposed models--one called Neural Prediction Model and another called Linear Predictive HMM--are shown to be special cases of the NPHMM proposed here. The experimental result affirms the justification of these proposed models.

The operation scheduling is an important subtask in the automatic synthesis of digital systems. Many greedy heuristics have been proposed for the operation scheduling, but they cannot find the globally best schedule. In this paper we present an algorithm to construct near optimal schedules. The algorithm combines characteristics of simulated annealing and neural networks. The neural network used in our scheduling algorithm is similar to that proposed by Hellstrom et al. However, while the problems of Refs. [11] and [12] have a single type of constraint, the problem considered in this paper has three types of constraints. As the result, the energy function of the proposed neural network is given by the weighted sum of three energy functions. To minimize the weighted sum of two or more energy functions, conventional methods try to find a good set of weights using a try and error method. Our algorithm takes a different approach than these methods. Results of the experiments show that the proposed algorithm can be used as an alternative heuristic for solving the operation scheduling problem. In addition, the proposed algorithm can exploit the inherent parallelism of the neural network.

We reported that a competitive learning neural network had the ability of self-organization in the classification of questionnaire survey data. In this letter, its self-organized learning was evaluated by means of mutual information. Mutual information may be useful to find efficently the network which can give optimal classification.

We present a superconducting neural network which functions as an RS flip-flop. We employ a coupled-SQUID as a neuron, which is a combination of a single-junction SQUID and a double-junction SQUID. A resistor is used as a fixed synapse. The network consists of two neurons and two synapses. The operation of the network is simulated under the junction current density of 100 kA/cm2. The result shows that the network is operated as an RS flip-flop with clock speed capability up to 50 GHz.

Calls using different media which require different transfer quality will arrive at ATM networks. Therefore it is important to develop a method for allocating network resources efficiently to individual calls by judging admission of calls. Various call admission control schemes have been already proposed, and these schemes assume that users specify values of traffic descriptors when they originate calls. However, it is sometimes difficult for users to specify these values accurately. This paper proposes a new ATM call admission control scheme based on cell transfer state monitoring which does not require that users specify values of traffic descriptors in detail when they originate calls. In this proposed scheme, the acceptance or rejection of calls is judged by comparing the monitored cell transfer state value with a threshold prepared in advance. This threshold must be adjusted according to changes in the characteristics of traffic applied to ATM networks. This is one of the most serious problems in the control scheme based on the monitoring of cell transfer state. Herein, this paper proposes neural network application to the control scheme in order to resolve this problem and improve performance. In principle, the threshold can be adjusted automatically by the self-learning function of the neural network, and the control can be maintained appropriately even if the characteristics of traffic applied to ATM networks change drastically. In this paper, the effectiveness of the application of a neural network is clarified by showing the configuration of this proposed control scheme with the neural network, a method for deciding various parameter values needed to implement this control scheme, and finally the results of a performance evaluation of the control scheme. Inputs required by the neural network are also discussed.

Switched-capacitor chaotic neurons fabricated in a full-custom integrated circuit are used to investigate the behavior of 2- and 3-neuron chaotic neural networks. Various sets of parameters are used to visualize the dynamical responses of the networks. Hysteresis of the network is also demonstrated. Lyapunov exponents are approximated from the measured data to characterize the state of each neuron. The effect of the finite length of data and the rounding effect of data acquisition system to the computation of Lyapunov exponents are briefly discussed.

We propose an asymmetric neural network which can solve inequality-constrained combinatorial optimization problems that are difficult to solve using symmetric neural networks. In this article, a knapsack problem that is one of such the problem is solved using the proposed network. Additionally, we study condition for obtaining a valid solution. In computer simulations, we show that the condition is correct and that the proposed network produces better solutions than the simple greedy algorithm.

This paper discusses a CMOS differential-difference amplifier circuit suitable for low voltage operation. A new multiple weighted input transconductor circuit structure is suggested to be use in DDA implementation. The proposed DDA can be employed in several analog/digital systems to improve their parameters. Selected examples of the proposed transconductor/DDA applications are also discussed.