Two new algorithms are presented for determining the elements of the matrix function Al of a digital system of which the matrix of dynamics has nonzero elements only in its superdiagonal (ones) and in its last row (a particular combination of the coefficients of the transfer function denominator). Advantages and disadvantages of these algorithms are discussed. It is shown that when using the matrix function Al only for the calculation of the system response, the second algorithm presented here seems to be particulary useful due to the simplicity of its application and due to the fact that it allows us to omit the calculations of the eigenvalues of the matrix of dynamics, which seem to have been necessary for most of the known methods. It also reduces the number of computer operations which must be performed from n3(n-2) using direct matrix multiplication, to n(n-2). The second algorithm can be efficiently used for the determination of the elements of limited power of the matrix of dynamics. Both of the algorithms presented in this paper are particularly useful when applied to the matrix of dynamics of a digital filter, of which the transfer function is presented in the direct form of realization, since, the matrix of dynamics of such a digital system is expressed directly in a proper form.

The structural design of a glazed alumina substrate for use in a hybrid IC repeater in the CS-36M submarine cable system is studied. The thickness of the alumina substrate, the thickness of the glaze, and the thermal expansion coefficient of the glass used for the glaze, are theoretically determine for the large (110 mm in diameter) substrate in terms of the load stress when the repeater is installed, the thermal strain when the glaze is formed, and displacement due to lateral pressure. It is found that the desirable thickness of the alumina substrate is approximately 2 mm, the desirable glaze thickness is approximately 65 µm, and the desirable thermal expansion coefficient is (7.77.5)10-6-1. A glazed alumina substrate is fabricated on the basis of these structural design parameters, and it is subjected to the mechanical strength test and form test prescribed for the CS-36M submarine cable system. The results prove that the substrate has sufficient mechanical strength and that there is no deformation problem. The glazed alumina substrate developed in this study is applied in a repeater used in the CS-36M submarine cable system and it contributes to the very high reliability of the repeater.

Iterative patterns can be described by the component pattern of iteration, the shape of which may usually involve the way of its connection for the iteration. Based on a parallel translation group for each iterative pattern, it is shown that the shape of component pattern can be a rectangle for any iterative pattern, which is called a tile and is related to the basis of parallel translation group. And it is shown that the component pattern can be identified by a tessera" which is defined by a non-degenerate colored tile. Thus the tessera may be said to be a compact description of iterative patterns. Some properties of tesseras and tiles are also discussed. The number of tesseras with a tile is evaluated, which means that the number of distinct iterative patterns for each shape of component patterns is enumerated.

For an incident plane wave, an exact solution of the moment equation is found in a special case where the correlation function B(r,z) of the fluctuating part of the relative dielectric constant can be approximately expressed as B(r,z)a0(z)+a2(z)r2+a4(z)r4, r2x2+y2; and the perturbation solution is obtained in the general case B(r,z)a0(z)+a2(z)r2+a4(z)r4+. The two solutions indicate that the r4 term in B(r, z) produces the log-normal distribution and the complete correlation of the intensity and that the higher term than r4 yields both declines of them. The perturbation solution shows that the condition for validity of the special case is B(0, 0)(L/l)31, (L/l)kl, and rm-rnl, m, n1v, where k is the wave number in free space, l is the correlation length of the medium fluctuation, L is the range between the transmitting place (z0) and the receiving plane (zL), and rm, rn denote the observation positions on the receiving plane.

It is known that psychotropic compounds give change to evoked response of human. A system for the Auditory Brain Stem Response (ABR) was experimentally developed, and ABR of subjects in general anesthesia were measured for the purpose of assessing the depth of anesthesia.

The transfer function of optical fibre is now more frequently being specified in terms of frequency response. While measuring the amplitude over frequencies up to several GHz is quite straightforward, determining the phase response presents difficulties. This paper examines the nature of the practical problems associated with measuring the differential phase across a long fiber and shows how they can be overcome. A measurement system is described which performs the required measurements easily and quickly with a minimum of samples. Some results are given for 900 m of fibre over the range 25 to 1250 MHz and using these the capability of the system is compared with a time domain measurement. The frequency domain technique appears to be the superior method.

The finite-element solutions of the discontinuity problems in a dielectric slab waveguide bounded by perfect electric or magnetic conductors are examined numerically and it is confirmed that the finite-element formulation for the bounded configuration facilitates solving the discontinuity problems in an open surface waveguide by placing the bounds appropriately away from the guiding surface.

We extend the position angle in the two-conductor transmission line to the position angle matrix in the multiconductor lines. The position angle matrix is applied to analysis of cascaded transmission lines, so that the relationship of the voltage vectors (or the current vectors, or the impedance matrices) between two different points on the line can be expressed in the compact form by using the position angle matrices.

Cardiac output can be measured from the rate of absorption of soluble inert gas across the lung using an electric model of human body suggested by W. Mapleson. Step response of the model becomes the sum of exponentials. Cardiac output is estimated from time constants and amplitude coefficients of the response. System for the measurement of the cardiac output which is based on this theory will be introduced below.

Modal noise in a quasi-single mode fiber propagating a few LP mode is studied experimentally by simulating lossy connectors and imperfectly coupled detectors. The signal-to-noise ratio of the so called two-mode fiber is found to be high as in the usual multimode fiber.

This paper describes an open question in a mathematical formulation of the guided complex wave supported by a slightly random surface. In case of the TM wave propagating over a randomly reactive plane surface, a formal wave solution is obtained and is shown to have an equivalent network representation. It is pointed out, however, that such a formal solution has no physical significance, because the effective surface impedance in the transverse resonance condition determining the complex propagation constant is not an analytic function on the complex wavenumber plane and because the formal solution gives a diverging variance of the guided complex wave. It is concluded that it is still an open problem to have a correct mathematical formulation for the guided wave over a random surface. Finally suggestions are given for obtaining a correct mathematical formulation.

Molded demountable single fiber connectors have been constructed, which contain precision balls within the plug moldings to locate the fiver and to align a pair of plugs. Transmission loss characteristics were examined and fiber placement precision investigated. The samples tested have exhibited less than 0.2 dB losses with index matching and less than 2 µm fiber displacements on an average for ordinary multimode fibers. For practical application, arbitrary diameter plugs using two sorts of balls and premolding plug ferrules were also investigated. The results exhibited high precision and low losses corresponding to the design condition. The molding procedure is simple and needs no complicated alignment mechanisms. These connectors are low in cost and possible to assemble in the field.

Realizability criteria for infinite dimensional, time-varying, continuous, nonlinear dynamical systems are studies. We obtain a necessary and sufficient condition for given input/output behaviors to be realized by the above systems in which the state spaces are Hilbert spaces. Moreover, we consider realization by dynamical systems which are governed by dissipative, time-varying, nonlinear differential equations and obtain a realizability criterion for this case. The theory of nonlinear evolution operators in Hilbert spaces plays an important role in out study.

In this paper, an unified analysis presented for evaluating the statistical properties of a general class of the quadratic detectors. The probability density function (pdf) for the output of the quadratic detectors is derived by the characteristic function method which requires the contour integration in a complex plane for the Laplace inversion operation, without enforcing to solve directly the eigenvalues of the integral equation as used by Kac, which is only possible for a special case of Gaussian input. We treat a general situation in which the tow inputs to the detector are narrowband signal-plus-interference-plus-noise. A general and closed-form expression of the output pdf of the quadratic detector is derived. We show that the unified analysis results can be applied not only to stationary Gaussian noise channels but also to Rayleigh fading channels, and that the results previously derived can easily be obtained using our unified results.

A characterization of circuits and cut-sets of a 2-connected series-parallel graph is given using the representation tree. Cut-sets correspond to subtrees of the representation tree called P-subtree and circuits are characterized in the dual form.

The distance transformation of a digital picture (DT) is one of the most important techniques in image processing. In this paper, we propose a new variant called the max-type distance transformation for binary pictures (max-type DT). This transformation can be compared to the propagation of a wave, which propagates with a constant velocity to the pixels whose connectivity numbers are less than two. The connectivity number is calculated on the binary pattern obtained by regarding the pixels which the wave has not reached yet as 1-pixels and all the others as 0-pixels. The transformed value of a pixel is defined as the time when the wave reaches the pixel. After the transformation, the pixels on the core line of an input picture are given larger distance values increasing along the core line. Therefore, the max-type DT is effectively used in thinning and structure analysis of binary pictures. In the structure analysis using the max-type DT, both distance information and topological properties of binary pictures are used together. As an example, experimental results using hand-printed characters are shown. The max-type DT can also be applied to extract shape features by measuring the length of the core line such as the major axis of an ellipse.

Stimulated Raman scattering was investigated in optical fibers, pumped by a mode-locked and Q-switched Nd:YAG laser or a CW Nd:YAG laser. The critical powers for the first five Stokes, at which each Stokes begins to build up, decreased with an increase in the length of fiber and became constant values above 200 meter long in the case of a mode-locked and Q-switched pump. The first and the second Stokes were observed with 0.6 W and 2.1 W of CW pump power input, respectively, in a 35 km single-span fiber. Experiments showed that the pulse wave separation effect played an important role in stimulated Raman scattering characteristics of optical fibers.

A newly devised beam-steerable antenna which is capable of steering its radiation beam without moving a main reflector is presented. This is the offset type spherical reflector antenna with two specific subreflectors. With the application of mathematical programming, a new technique to determine the shapes of the subreflectors is proposed in order to compoensate for the spherical aberration and the asymmetrical distortion of the aperture distribution, which deteriorates the electric characteristics. An experimental antenna has demonstrated its excellent performance and supported the adequacy and availability of the designing procedure. The results suggest the feasibility of the beam-steerable antenna for satellite communication use.

For full solid-state key telephone systems, high breakover voltage IC switches, which can connect to the central office lines directly and form a large scale matrix within a small volume, has been desired. Firstly, this paper compares several kinds of IC switches, which are suitable for formation of speech path matrix and shows that the photo-coupled PNPN crosspoint switch, connected in bridge format, is superier to the others because of low transmission loss, small matrix volume, reliability against lightning, self-latching speech path during power suspension and NTT technical requirements satisfaction. Secondly, the paper clarifies required electrical conditions for the PNPN element and the crosspoint circuit. They are breakover voltage above 300 V, dv/dt firing torelance above 160 V/0.1 µs, on-resistance within 9 Ω, less than 0.97 V forward voltage drop and restricting the reverse leak current which flows in the optically triggered PNPN element. Next, the IC switch structure is described, along with available technologies to satisfy the electrical requirements, such as canal isolation, highly efficient photo coupling structure and light shielding of anode junction. Actual IC switch characteristics satisfy all required values.

An efficient finite-element program utilizing approximate analytical solutions is described for the analysis of topographic waveguides for acoustic surface waves. In this method, the only ridge region is divided into triangular elements, and therefore it is possible to use computer memory more economically. Calculations have been made for the rectangular ridge waveguide, the dovetail ridge waveguide, the truncated knife edge waveguide and the wedge waveguide. The results obtained agree well with the earlier theoretical and experimental results.