We report on an experimental hysteresis in the Hopfield networks and examine the effect of the hysteresis on some important characteristics of the Hopfield networks. The detail mathematic description of the hysteresis phenomenon in the Hopfield networks is given. It suggests that the hysteresis results from fully–connected interconnection of the Hopfield networks and the hysteresis tends to makes the Hopfield networks difficult to reach the global minimum. This paper presents a T–Model network approach to overcoming the hysteresis phenomenon by employing a half–connected interconnection. As a result, there is no hysteresis phenomenon found in the T–Model networks. Theoretical analysis of the T–Model networks is also given. The hysteresis phenomenon in the Hopfield and the T–Model networks is illustrated through experiments and simulations. The experiments agree with the theoretical analysis very well.

We discuss a processor scheduling problem for parallel logic simulation of combinational circuits. In the processor scheduling problem, to be discussed in this paper, for logic simulation using time–first method, the time needed for each gate evaluation is not given beforehand, and is not constant. This feature distinguishes the processor scheduling problem from typical task scheduling problems. First, we devise newly Algorithm MET to solve the processor scheduling problem. The key idea of Algorithm MET is to determine processor scheduling incrementally and dynamically. Then, experimental evaluations using well–known twelve benchmark combinational circuits show the usefulness of Algorithm, MET, compared with conventional static algorithms. We believe that this is a first step to implement parallel logic simulation of combinational circuits.

This paper discusses the characteristics of the nonlinearly amplified spread–spectrum (SS) signals. We evaluate the symbol error–rate performance with the conventional receiver, changing the length of the spreading sequence. In addition, we also propose the receiver with MLSE. The configuration of the MLSE for the nonlinearly amplified signals is generally complicated; however we show that the complexity of the MLSE receiver can be reduced, as the number of required reference sequences in the receiver for an SS signal is small. As the result, it is shown that the error rate performance of the nonlinearly amplified SS signal can be improved by this proposed receiver and that the degradation caused by the nonlinear amplification can be made negligibly small with a sufficiently long spreading sequence.

This paper describes that a neural network, which consists of neurons with piecewise–linear sigmoid characteristics, is able to approximate any piecewise–linear map with origin symmetry. The neural network can generate "deterministic diffusion" originating from its diffusive trajectory.

In this paper we study the bifurcation phenomena of quasi–periodic states of a model of the human circadian rhythm, which is described by a system of coupled van der Pol equations with a periodic external forcing term. In the system a periodic or quasi–periodic solution corresponds to a synchronized or desynchronized state of the circadian rhythm, respectively. By using a stroboscopic mapping, called a Poincar mapping, the periodic or quasi–periodic solution is reduced to a fixed point or an invariant closed curve (ab. ICC). Hence we can discuss the bifurcations for the periodic and quasi–periodic solutions by considering that of the fixed point and ICC of the mapping. At first, the geometrical behavior of the 3 generic bifurcations, i.e., tangent, Hopf and period doublig bifurcations, of the periodic solutions is given, Then, we use a qualitative approach to bring out the similar behavior for the bifurcations of the periodic and quasi–periodic solutions in the phase space and in the Poincarsection respectively. At last, we show bifurcation diagrams concerning both periodic and quasi–periodic solutions, in different parameter planes. For the ICC, we concentrate our attention on the period doubling cascade route to chaos, the folding of the parameter plane, the windows in the chaos and the occurrence of the type I intermittency.

A new algorithm, which is incorporated into the waveform relaxation analysis, for efficiently simulating the transient response of single lossy transmission lines or lossy coupled multiconductor transmission lines, terminated with arbitrary networks will be presented. This method exploits the inherent delay present in a transmission line for achieving simulation efficiency equivalent to obtaining converged waveforms with a single iteration by the conventional iterative waveform relaxation approach. To this end we propose 'line delay window partitioning' algorithm in which the simulation interval is divided into sequential windows of duration equal to the transmission line delay. This window scheme enables the computation of the reflected voltage waveforms accurately, ahead of simulation, in each window. It should be noted that the present window partitioning scheme is different from the existing window techniques which are aimed at exploiting the non–uniform convergence in different windows. In contrast, the present window technique is equivalent to achieving uniform convergence in all the windows with a single iteration. In addition our method eliminates the need to simulate the transmission line delay by the application of Branin's classical method of characteristics. Further, we describe a simple and efficient method to compute the attenuated waveforms using a particular form of lumped element model of attenuation function. Simulation examples of both single and coupled lines terminated with linear and nonlinear elements will be presented. Comparison indicates that the present method is several times faster than the previous waveform relaxation method and its accuracy is verified by the circuit simulator PSpice.

A technique for pulse code modulation (PCM) encoding using a T-Model neural network is described. Performance evaluation on both the T-Model and the Hopfield model neural-based PCM encoders is carried out with PSpice simulations. The PSpice simulations also show that the T-Model neural-based PCM encoder computes to a global minimum much more effectively and more quickly than the Hopfield one.

This paper describes a simple system of measuring the spatial distributions of spectral intensities with AgI-421 nm and AgI-546 nm among many optical spectrums emitted from an arc discharge between separating Ag contacts. In order to detect the intensities of two optical spectrums, the prototype equipment has two sets assembled with a CCD color linear image sensor, a lens and optical filters, which are arranged on rectangularity. The intensities of two spectrums can be recorded with 2 ms time-resolution within a long arc duration on a digital memory. The recorded digital signals are processed by using a personal computer in order to reconstruct two spatial distributions of spectral intensities in a cross section of arc column with the Algebraic Reconstruction Technique.

Recognizable series is a model of a sequential machine. A recognizable series S is represented by a triple (λ,µ,γ), called a linear representation of S, where λ is a row vector of dimension n specifying the initial state, γ is a column vector of dimension n specifying the output at a state, and µ is a morphism from input words to nn matrices specifying the state transition. The output for an input word w is defined as λ(µw) γ, called the coefficient of w in S, and written as (S,w). We present an algorithm which constructs a reduced linear representation of an unknown recognizable series S, with coefficients in a commutative field, using coefficient queries and equivalence queries. The answer to a coefficient query, with a word w, is the coefficient (S, w) of w in S. When one asks an equivalence query with a linear representation (λ,µ,γ), if (λ,µ,γ) is a linear representation of S, yes is returned, and otherwise a word c such that λ (µc) γ(S, c) and the coefficient (S, c) are returned: Such a word c is called a counterexample for the query. For each execution step of the algorithm, the execution time consumed from the initial step to the current step is O(mN 4M), where N is the dimension of a reduced linear representation of S, M is the maximum time consumed by a single fundamental operation (addition, subtraction, multiplication or division), and m is the maximum length of counterexamples as answers to equivalence queries returned until that step.

In this paper, the relationship between contact materials and sticking characteristics, and stability of contact resistance to obtain excellent contacts for telecommunication relays, is studied. The contact switching current for telecommunication relay is low. Moreover, contact force and opening force in these relay are respectively several mN. Nine kinds of contact materials are selected as a experimental factor. They are Ag, Ag-Ni (Ni: 0.03 to 20%), Ag-Cu 10%, Ag-Pd 60% and Pd-Ru 10%, and are overlaid with gold except Pd-Ru 10%. In this study, contact life tests on a commercial ultra-miniature telecommunication relay by mounting above-mentioned contacts are conducted. The sticking and the contact resistance are monitored at each switching operation in the contact life test. After the life test, the contact surfaces are observed, and the depth of crater, the height of pip and projected concave area are measured, then the relationship between the sticking morphologies and the composition of each material are studied. As the result of this study, the contact sticking of telecommunication rely is assumed to be the result of mechanical locking, and the effects of the Ni content in the Ag-Ni contacts is clarified. Moreover, it is confirmed that the effects of opening force on the sticking characteristics are remarkable.

This paper presents a support method for specifying communication systems. Generally, a set of requirements for a target system is partial and ambiguous to construct the whole system, namely it lacks certain necessary descriptions for the target system. To attack this problem, our method enables a designer to obtain such necessary descriptions from specifications stored in a knowledge base, namely by reusing specifications, and helps the designer to specify the target system completely. In our support method, we adopt labelled transition systems (LTSs) which are state transition graphs and are shared as a common notion by most FDTs. Therefore, our method is the common approach to FDTs. We propose a new idea about similarity berween LTSs, and propose an algorithm to suggest similar LTSs to the designer.

On the surface of contacts which are exposed to the atmosphere, the reaction with gases in the atmosphere produces contaminant films including oxides. The contact reliability is degraded by the contaminant films. Humidity in the atmospheric environment also influences on the surface of contacts. However, influence of humidity on the surface has not been clarified. In the present paper, influence of humidity on the Cu surface and the oxides (CuO + Cu2O) on it were studied with respect to the thickness of the oxide film and contact resistance characteristics both for static and for sliding contacts. The thickness was measured by ellipsometric analysis. Topographic image affected by humidification was also observed by scanning tunneling microscope (STM). In the atmospheric environment, the clean surface of Cu was found to oxidize with fluctuations of the thickness for lapse of exposure time due to the fluctuations of the humidity. It was also found that the thickness of the oxide film decreases immediately after the humidification, and increases under dehumidification. Changes in contact resistance affected by humidity was corresponding to the change in the film thickness. Immediately after humidification contact resistance decreased, and increased with dehumidification both for static and for sliding contacts. For the mechanism of the influence of humidity on the oxide, chemical reduction of hydrogen generated by decomposition of the absorbed water molecule (H2O) was derived. The clean Cu surface was oxidized by oxygen due to absorbed water molecule and atmosphere.

This paper presents an effective application of Net-theory for all the stages of the communication protocol development process. Net-theory provides a basic mathematical model and tool for development of communication protocol. The special usability of Net-theory is that 1) visual representation of the system's stadic/dynamic structure, so that users may easily understand the represented contents, 2) formal specifications based on mathematical basis of Net-theory admit automatic verification, implementation and conformance testing. We have seen that Net-theory which has the above usability can provide a systematic and advanced paradigm for effective communication protocol development.

It is desirable to design an ATM switch that is nonblocking at the connection level by using simple connection admission control (CAC) schemes. To accomplish this goal, it is necessary to consider the relationships between CAC, cell-level quality-of-services (QOS), and the structure of multistage switches as well as switch modules. In this paper, we formulate a framework to design a multistage nonblocking ATM switch. We show that if a switch has the property of the Sufficiency of Knowledge of External Loads (SKEL), i.e., the property that its cell-level performance is robust to the distribution of incoming traffic among all inputs, then the switch is also nonblocking at the connection-level by using a simplified CAC that guarantees QOS of a connection by controlling the aggregate loads on outputs. Furthermore, we show that a Clos three-stage network using SKEL switch modules and Multipath Self-Routing (MPSR) also has the SKEL property and is a nonblocking switching network that needs CAC only at its outputs. We also demonstrate a design of multistage nonblocking ATM switches with Knockout switch modules.

Experimental optical gain characteristics of an erbium-doped fiber amplifier have not been explained well by conventional laser schemes in the case of two-channel amplification. Modified simple laser schemes including cross relaxation among degenerate levels were valid for the explanation of the optical gain dependence on input signal power and on the erbium-doped fiber length.

A radial line slot antenna (RLSA) is a slotted waveguide planar array for the direct broadcast from satellite (DBS) subscriber antennas. A single-layered RLSA (SL-RLSA) is excited by a radially outward traveling wave. The antenna efficiency of more than 85% has already been realized. These antennas are designed on the assumption of perfectly rotationally symmetrical traveling wave excitation; the slot design is based upon the analysis of a slot pair on the rectangular waveguide model with periodic boundary walls. However, the slots perturb the inner field and the actual antenna operation is not perfectly symmetrical. This causes the efficiency reduction especially for very small size antenna. This paper presents a fundamental analysis of the inner field of the radial waveguide. It is impossible to analyze all the slot pairs in the aperture as it is and only the slots in the inner few turns are considered since these provide dominant perturbation. The calculated results are verified by the experiments and reasonable agreement is demonstrated. Some design policies are suggested for enhancing the rotational symmetry.

A new optical path cross-connect system architecture (OPXC) based on delivery and coupling matrix switches is described. This OPXC provides the maximum compatibility for a wavelength path (WP) network and a virtual wavelength path (VWP) network. In other words, the proposed architecture easily evolves from WP-OPXC to VWP-OPXC. This salient feature can not been achieved with conventional OPXCs. Another attractive feature of this OPXC is its high modularity for OPXC capacity expansion.

Among the problems that face ITS designers, the problem of measuring the student knowledge state after concept learning in order to initially adapt a skill acquisition session according to a student's own necessities is a hard one. Typical approaches are the use of some sort of test to assess the student knowledge and choose an initial set of parameters for a session, or use, regardless the particular necessities of a student, a pre-defined set of initial parameters. We consider the fromer to be disrupting for learning and the latter too simple to deal with the broad possibilities that are faced. It is known that students show different behaviors during concept learning depending on the experience, background and actual understanding (the way a student is understanding a concept) during concept learning. Our approach here is to classify the different behaviors through fuzzy proposition and link them with a student model through fuzzy rules to use in an expert system, and with it, select the most suitable problem-solving strategy for each particular student in order to clear his misunderstandings and facilitate the learning of problem-solving skills. The use of probabilistic reasoning (i.e. Bayesian statistics) instead of fuzzy logic is not suitable for the present situation because of the rigidity and precision of the rules that do not allow a proper manipulation of the vagueness involved in the student behavior. We apply this idea to a circuit analysis ITS where the concept learning session is carried out on a Hypertext environment and the skill acquisition session on an interactive problem-solving environment. By tracing the student use of the Hypertext environment we can know the student behavior and use it as a premise in the fuzzy inference.

As a generalization of the tree automaton, tree automata with various types of memory are introduced and their relation to context-free grammars with memory is studied. Relations between computation trees of tree automata with memory and derivation trees of context-free grammars with memory are established, and as a consequence, the languages generated by context-free grammars with memory are characterized in terms of the sets of trees recognizable by tree automata with memory. Also various types of traversal of labeled trees recognizable by tree automata with memory are considered.

Several distances between trees have been proposed. However, most of the reports on distances have dealt with rooted and ordered trees. This paper proposes two distances between unrooted and cyclically ordered trees (CO-trees) and their computing methods. A CO-tree is a tree embedded in a plane. These distances are defined based on Tai's mapping (TM) and a strongly structure preserving mapping (SSPM) between CO-trees. The time complexities to compute the distances between two CO-trees Ta and Tb are OT (N 2aN 2b) for the distance based on a TM and OT(mambNaNb) for that on an SSPM, respectively, where ma(mb) and Na(Nb) are the largest degree of a vertex and the number of vertices of Ta(Tb), respectively. The space complexities of both methods are Os(NaNb). Those distances can be applied to the clustering of CO-trees.