Spectrum efficiency of M-ary PSK land mobile radio is discussed theoretically, where the efficiency ηT (erl/Hz-m2) is defined as the spatial traffic density per system bandwidth. Effective techniques to increase ηT are reviewed. Effects of multi-level modulation, diversity reception, forward error correction (FEC) and detection scheme upon ηT are clarified. It is shown that the combination of quadri-phase-shift-keying with coherent detection (QCPSK) and the FEC at a certain coding rate makes ηT maximum. For the optimum system, ηT and the system capacity are evaluated under a practical condition.

In this paper we present a new relationship between two differently derived FIR low-pass differentiation algorithms: The time-domain based Lanczos algorithm, and the frequency domain Minimum Mean Square Error(MMSE) algorithm. We then discuss the theoretical significance and the practical results of this relationship.

This paper describes charging and deflection characteristics of a binary multi-nozzle ink jet printer. In a binary printer, ink drops are selectively deflected by applying a charge to them, while uncharged drops proceed in a straight line to form the part of characters. The printer used in our experiment has eight nozzles, 55µm in diameter and on 0.5 mm centers. Charging electrode slot width is 0.3 mm. Drop frequency is 40 kHz. Drop frequency is divided into printing frequency by 16 so that a moderate printing speed is obtained. The charge electrode array is positioned with respect to a nozzle array so that the drop formation occurs within a certain region inside the charge electrode slots. Then no crosstalk between the adjoining electrodes has been observed and the induced charge by the previously charged drop can be neglected. The deflection condithion, for defect-free prints, is such that the charged drops are collected into a catcher whereas the charged drops do not merge with the uncharged drops. The charging voltage and the separation between the undeflected ink streams and the deflection plated are important parameters. Experimentally a sufficient amount of deflection has been obtained by the image force of the charged drops without the use of a high deflection voltage.

Refractive-index dispersion mesuarements were performed using the minimum deviation method for bulk specimens at wavelengths from 0.4047 to 2.0581µm. The data were utilized to calculate an optimum index profile,αop, and zero-material-dispersion wavelength,λ0, for the optical fiber. For GeO2-SiO2 glass core fibers,αop values are 1.93 at 1.30µm, and 1.88 at 1.55µm. There are independent of GeO2 concentration. The wavelength λ0 shifts from 1.272 to 1.376µm, proportionally to the GeO2 concentration from 0-19.3 mol%. The data also show that the addition of P2O5 slightly increases material dispersion, and the value of αop for the P2O5-SiO2 core glass decreases monotonically from 2.16 at 0.4 µm to 1.91 at 2.0µm.

The noise properties of directly modulated laser diode influenced by reflected waves are given theoretically. Firstly, the quantum noise enhancement effect due to direct modulation is found theoretically in term of product between the quantum shot noise and the modulation current. Secondly, the high-frequency intensity noise due to reflected waves (often so-called self-coupling effect) is clarified to be due to resonance of the quantum shot noise to the external cavity. According to our theory, it is found that the noise of the narrow stripe laser diode with the stripe width comparable to the diffusion length of carrier is about 10 dB smaller than that of the wide stripe laser. And also, the reverse isolation loss of an optical isolator is estimated theoretically to be 30 dB to reduce the relative high-frequency intensity noise by 40 dB.

Switching threshold characteristics of the shaped Josephson junctions for tree-decoder use were investigated theoretically and experimentally. Side lobe suppression was successfully achieved by a constricted full window junction structure, whose optimum constriction could be predicted by Fourier integral of junction width function. Side lobe suppression ratio as large as 16 was obtaind for constriction ratio K/L1/3, and junction length L/λ03.3.

A transient analysis of the electromagnetic field by our new numerical method is shown to be useful for understanding the mechanism by which the electromagnetic field at the stationary state in the waveguide having the H-corner are brought about.

A store and fordward speech interpolation telephonic concentrator system is introduced. This paper presents first the system, describes its main parts and components and using typical parameters the probable system operation is outlined. Following, a study on capacity and delay of transmission of each voice packet is done by supposing that a M/D/1 model can be applied to the system. As voice has special characteristics and packets don't arrive randomly as supposed in the M/D/1 model, an analysis using a computer simulation is done. The source generating speech is closely matched with human voice by using a model with good resolution specially for small pauses of talkers during the active generation of speech. Using this source, simulations representing 120 or 180 sec. of actual time are done. As its results are different of those predicted by the M/D/1 model, the capacity of the system is forecasted again through examples. To increase even more the capacity without impairing the frozen out fraction of the speech or increasing the delay of each packed, a different method to decrease congestion in those short moments of high arrival rate of packets is intended. This method consist in to transmit from each voice sample only 7 bits during the high congestion moments, creating new capacity by shortening packets. Results of parameters obtained by simulation and probable capacity of the system are again shown.

A unified approach to the derivation of canonical single amplifier circuits capable of realizing second-order filters with finite transmission zeros is proposed. Application of the proposed approach leads to the known circuits, as well as some new ones. Design formulae for the new circuits are presented, and their salient features are pointed out. Finally, design examples are also presented.

A necessary condition is given for a closed set of k-valued logical functions to be essentially minimal. It is stated in terms of the associated 1-variable function of a generator of the set concerned.

This paper presents some methods for recognition of handprinted Arabid scripts. Arabic scripts are generally structured of curves and hence our recognition method is based on the identification of the component curves in the script. In the paper, first we propose a method for recognition of single handprinted Arabid characters, using the concept of contour tracing and identification of the curves in the character. The identification of a curve is performed through determination of its local directions. Next, we develope the method to the recognition of cursively handprinted scripts, by usage of the identification of the basic shapes. The basic shapes are identified through using the orientation of the curves and the relationships between the feature points in the script. The results of computer simulations show average recognition rates of more than 95% and 90% for singly written and cursively written characters, respectively. The advantage of our method is its powerful recognition of highly distorted shapes, particularly the shapes which are generated through rotation of the original pattern. Although the present method has been developed for the recognition of Arabic scripts, it can be applied equally well for recognizing many other singly and cursively written scripts.

As a means of improving the rate of convergence of the conventional echo canceller using the learning identification method, the authors have previously proposed a linear predictive algorithm. This algorithm shows better convergence than the learning identification method. However, in this algorithm, as well as in the learning identification method, a compromise is necessary between a relatively large step gain required for fast convergence and the relatively small step gain needed for noise insensitivity in the presence of noise. In this paper a new algorithm based on the linear predicitive algorithm is proposed, in which the step gain is determined as a function of the estimated values of noise and the parameters-error of the echo path model in order to improve both the rate of convergence and the noise insensitivity simultaneously. The efficiency of the proposed algorithm is examined by computer simulations. It has been shown that the proposed algorithm gives about twice the rate of convergence and about 10 dB lower parameters-error in the stationary state in comparison with the learning identification method. Besides, it has been proved that this algorithm guarantees non-divergence of the echo path model even during the period of double-talking" without any control device such as a double-talking detector.

In the case of rigid disk files, improvements in the recording medium and the head are required to increase bit density. This letter reports a high recording density of 2600 bit/mm (65000 bit/i), achieved by using high coercive force (100 Oe) sputtered γ-Fe2O3 thin film media and a narrow gap (2g0.15µm)head.

A nonlinear electric circuit model of the whole length basilar membrane (BM) of the human is constructed. It is of transmission analogue type. Nonlinearity is introduced into its loss factors, using resistors and Ge diodes. By observing the membrane velocity of the model, we examine: (i) response to combination tones (CT) with two primary tones of frequency fL, fH and level LL, LH (fLfH), (ii) 2fLfH CT cancellation characteristics, (iii) two-tone wuppression and (iv) other responses. Suppressed tone has frequency f1(characteristic frequency (CF) of given section) and level L1, and suppressor tone has frequency f2 and level L2. Distribution of 2fLfH CT component shows that at some sections it cominates the primaries in amplitude, while 2fHfL component is very much smaller all along the model. The 2fLfH CT distribution patterns in the model, when it is cancelled, seem to indicate good correspondence with the psychophysical cancellation. Three CT cansellation characteristics (LLLH varied, LL varied, fH varied) show good agreement with psychophysical experiments, including the nonmonotonic behavior of the CT. It seems to be possible to have more agreement by using sharper tuning property and stronger nonlinearity. For frequency f1f2, average value of BM velocity shows two-tone suppression, but for f2f1, it shows a little suppression. It will be worthwhile to consider other variables, such as BM displacement or acceleration, as model output.

A double-layer resist-film technique to fabricate the undercut profiles in optical and electron beam lithography is described. The process consists of successive spin-coating of two different resist films, developing the top resist layer, and etching the bottom layer in oxygen plasma. Oxygen plasma etching experiments have been accomplished to fabricate the desired undercut profiles. Diazo resists have been chosen as the top resist layer and PBS as the bottom resist layer. With the optimization of resists and plasma etching conditions, undercut profiles suitable for lift-off metallization are obtained with better pattern width control and better undercut control.

Using an electrically controllable (EC) Echelette grating which consists of numerous thin glass-plates on a unimorph transducer, the wavelength selection and intracavity modulation are demonstrated at CO2 10.6 µm laser, together with the basic characteristics of the grating and theoretical treatments. Under intracavity setting of the grating in three-mirror configuration of a resonator, the initially oscillated CO2 line, P22, is typically moved into P18 line with 100V applied to the grating. On the intensity modulation, about 30% of modulation is also obtained with only 10 V.

A depolarization phenomenon at 34.5 GHz due to wet snow was observed in a horizantal terrestrial path of 1.3 km. Data during snowfall show fairly large degradations of XPD in comparison with data during rain and theoretical values based upon the model of deformed raindrop, under the condition of the same copolar attenuation. Much larger deformation of snowflake than that of raindrop must be assumed to explain such inclination by the model of depolarization due to an oblate spheroidal snowflake. Effective differential propagation constants of snow, which are derived from measurements of XPD and phase difference between cross- and co-polar components, show that depolarization due to wet snow at 34.5 GHz comes from not only differential phase shift but also differential attenuation. A certain kind of anomalous depolarization on Earth-space paths may be explained by the model of depolarization due to wet snow.

The frequency of an AlGaAs DH laser has been stabilized with respect to a Fabry-Perot interferometer by controlling the injection current. The frequency stability of 2.11012 was obtained at τ90 ms, which was better than the free-funning stability by three orders of magnitude.

Fourier reconstruction has orginally been one of the methods for three-dimensional structure reconstruction in medical applications such as the computed tomography (CT). The technique is found to be applicable to designing circularly symmetric FIR digital filters. The algorithm is simple, straightforward and practical. The approach basically consists of designing a one-dimensional (1-D) FIR prototype, rotating its frequency response in the two-dimensional (2-D) frequency plane, inverse Fourier transforming and then windowing the result. An unlimited number of profections at different orientations of the obtained impulse response are essentially identical with the 1-D FIR prototype. The even size 2-D FIR filters can also be designed. The impulse response has eightfold (octagonal) symmetry. The frequency response is shown to have nearly circular symmetry that is evaluated with rms and maximum absolute errors.

An extraction method of density contour lines of gray level pictures is described. Two kinds of quads composed of four image pixels and non-image pixels are defined for making the density contour lines simple closed paths. From the resultant density contour lines of simple closed paths, a tree graph is made for executing the recognition of a house included in a monochromatic aerial photo of an urban area.