Transition phenomena associated with synchronization are investigated between van der Pol and hard-mode oscillators with slightly different frequencies interacting through linear coupling. Three types transitions between asynchronous and synchronous states appear, and power law divergence of time describing the transition is observed near the transition point.

Tensile strength of heavy metal fluoride glass fibers was investigated. The effect of proof testing on the fiber strength was evaluated theoretically and experimentally. It was found that the strength distribution of the fluoride fiber after proof testing can be predicted, based on the slow-crack growth model. This result enables to extend the cable design for silica fibers to that for fluoride fibers to assure long-term reliability.

The maximum power density of FM carrier modulated by a multi-channel telephony signal (Gaussian noise) with the approximated CCIR pre-emphasis characteristics is calculated by the summation of convolutions up to the 300th order. The results are shown in a figure as a function of m (rms modulation index) and e (ratio of minimum to maximum baseband frequencies) and are compared with approximations.

Power reflectivity, extinction ratios and scattering loss of optical switches with intersecting waveguides are analyzed and their electrode-length dependence is investingated. The electrode is found to be several hundred microns long if the refractive index variation is about 1%. It is, therefore, anticipated that small size optical switches are realizable.

A chemical depth profile of anodic oxide on GaSb was investigated by using XPS method. The constant current anodic oxidation was able to grow thick oxide layers of several thousand angstroms. And the oxide thickness was controlled well by the terminating voltage. The oxidized gallium and the oxidized antimony were distributed uniformly throughout the oxide layer. And there was less pile-up of the elemental antimony at the oxide-semiconductor interface.

The problems of robust tracking with stability for uncertain plants are considered. Uncertrain plants of one type are systems whose transfer matrices are only known to belong to a convex hull of prescribed transfer matrices P₁, P₂,, P_N . Such uncertain systems include some dynamical interval systems as examples. A necessary and sufficient condition for the solvability is represented in terms of extreme points P₁, P₂,, P_N of the convex hull. The solvability condition is testable and leads to a systematic procedure for designing a robust feedback control system.

A set A is said to be invertibly paddable if there are two polynomial time computable functions pad and decode such that () xA if and only if pad (x, y) A, and () decode (pad (x, y)) y. We consider three number theoretical problems that are used in certain cryptosystems (decision of quadratic residuosity, computation of discrete logarithm and computation of Euler's totient function), and show that the sets that represent these problems are invertibly paddable. These results imply that, if these sets are not in P, then they have complexity cores C such that neither C not the complement of C are sparse.

Along with the advances in network digitalization and the development of new service capabilities, international telecommunication networks have been evolving toward ISDNs. The introduction of computer-controlled network operation techniques such as dynamic routing in the ISDN era will facilitate the highly efficient use of network resources with high reliabilities. In this paper, an international ISDN traffic profile model is introduced, and then a simple design method for international network using the dynamic routing of the pre-planned time-varying type is presented. A constraint concerning traffic survivability for robust network design is taken into account in the proposed method. The improvement of resource utilization and reliability performance attained by the introduction of the dynamic routing scheme is numerically evaluated based upon the design results for simplified models of international ISDNs.

This paper describes the post-detection diversity improvement of a binary DPSK system in the fast Rician fading environment encountered in mobile satellite communication system. Firstly, the error probability is presented by a closed form which contains no special functions, then the optimum weighting factors in diversity branches to minimize the bit error rate (BER) are investigated. To minimize the BER, the maximum likelihood ratio detection principle is used for the detector outputs with exponential-type probability density function. Some unknown channel parameters contained in the optimum weighting factors are estimated by the channel estimators in diversity branches. Finally, the theoretical BER performances of three kinds of mobile satellite communication systems (satellite/aircraft, satellite/land mobile and satellite/ship) are presented.

Multiple site diversity systems are foreseen for earth to satellite paths operating at frequencies above 10 GHz in localities with high rain-induced attenuation. In the present paper, a model for the prediction of the site diversity performance based on the two-dimensional gamma distribution is proposed. The model is oriented for application to earth-space paths located in Japan and other locations with similar climatic conditions. Numerical results are compared with a limited set of available experimental data in Japan. Some useful conclusions are deduced.

A wideband loop antenna system has been tested using the sounding rocket S-310-18 to confirm its performance in detecting electromagnetic waves for frequencies from ELF to MF band in the lower ionosphere. The loop antenna system detected the one-hop whistlers with a high Signal-to-Noise ratio in 0.1-12 kHz, the spectra of which are used to study the propagation characteristics in the ionosphere. The VLF signal transmitted from the NDT station (Yosami, 17.40.05 kHz) was continuously received during the flight of rocket. The altitude dependence of the wave intensity and polarization are obtained. The polarization change of NDT signal from linear (free space mode) to circular (whistler mode) was clearly observed at a certain altitude in the ionosphere. The three Omega waves from Tsushima were detected and their height patterns of wave intensity are derived. A middle frequency radio wave (873 kHz) from Kumamoto Broadcasting Station was for the first time detected by the wideband loop antenna. It was found that the wave intensity decreases abruptly in the lower ionosphere due to absorption and reflection. This phenomenon was found to occur during both ascent and descent. Thus, it has been shown that the wideband loop antenna system tested in this rocket experiment has potential merit when applied in a space vehicle.

Based on the MSN algorithm, an adaptive algorithm capable of controlling antenna patterns by phase weight only is derived. Then, an algorithm for phase nulling convenient in case not needing adaptive weight control is derived. This algorithm with null directions only being the parameter is convenient for pattern synthesis and pattern improvement. Finally, two examples are presented to demonstrate usefulness and validity of the technique.

We have already measured ground clutter using an X-band and L-band radars. Here, ground clutter was first measured using a millimeter wave radar with the frequency of 34.86 GHz, which is located on the campus of the University of Electro-Communications. The pulsewidth of the radar was 30 ns. Thus the spatial resolution was as small as 4.5 m. We have found that the clutter amplitude distribution obeys a Weibull distribution with shape parameter of c0.497 to 0.675 at depression angles of 0.8 to 1.9 deg., when reflectors are ordinary terrain and such structures as landing strips at airport and buildings. To improve target detectability in such Weibull-distributed ground clutter, Weibull constant false alarm rate (CFAR) system will be required.

Some approximations accepted in the relaxation time approximation for GaAs have been discussed using the particle simulation. The kinetic energy of electron drift motion is not negligibly small in the submicron device. The relaxation times of energy and momentum evaluated from the steady state data on the energy and momentum fairly differ from those in the real relaxations. The valley population is not a simple function of the energy only. The relaxation time of energy depends on not only the energy, but also the valley population.

High density and high speed nonvolatile random access memory is described. Using the conventional floating gate EEPROM process, a dynamic RAM cell has been merged into an EEPROM cell. Data stored on the DRAM cell can be backed up by the EEPROM cell. The data transfer between the DRAM and the EEPROM is executed simultaneously on all memory cells on a same word line. An experimental 16 kbit memory has been manufactured by 1.5 µm design rule CMOS process. The cell size is 17 µm17 µm and the chip size is 57.2 mm3.75 mm. The address access time of 100 ns and the page read access time of 20 ns have been achieved. A nonvolatile CAM (Content Addressable Memory) cell will also be proposed.

In this paper we consider efficient algorithms in a semi-dynamic situation, that is, input data are entered one by one while some of them may be deleted on the way and then finally we are required to solve a given problem. We evaluate such algorithms by a three tuple (T_i, T_d, T_p) where T_i and T_d denote the time needed for insertion and deletion of data, respectively, and T_p is the time for solving a given problem. We present efficient algorithms for several different problems including the convex hull problem, the visible-pair enumeration problem and the visibility region problem.

This paper deals with a popular puzzle known as Hi-Q. The puzzle is generalized: the board is extended to the size nn, an initial position of the puzzle is given, and a place is given on which only one token is finally placed. The complexity of the generalized Hi-Q is proved NP-complete.