In instructional design, designers (or teachers) choose an optimum educational media (or educational system) adaptive to the learners' abilities. Although instructional design has been recognized as one of the important teachers' activities thus far, they left most of it to their individual subjectives. In order to perform the design in an objective way, a new mathematical method is proposed here to discriminate if the concerned educational system will be effective to learners on not. It is based on the use of linear discriminant analysis. However, when we apply it to the discrimination on education, the conventional method is not always appropriate to the field of education because many educational events are fuzzy, thus it requires refinement. From the point that the author regards the fuzziness proper to education, in order to aid teachers' instructional designing in an objective way, this paper presents a new logic which fixes linear discriminant function, the equations of error probability and Maharanobis' Generalized Distance considering the fuzziness.

Bit error rate (BER) performance of convolutional coding/soft decision Viterbi decoding is investigated theoretically. Firstly, equations are derived to calculate the tight upper bound of the BER performance for the multi-level soft decision with arbitrary threshold spacing, considering the quantized metric used in the Viterbi decoder. Experimental results on the BER performance of a rate 1/2 code with constraint length 7 are shown to demonstrate that the derived equations provide the precise BER performance. Then, the BER performance of various codes is calculated for 4- and 8-level soft decision as well as hard decision and ideal (infinite level) soft decision. The codes examined in this paper cover typical codes with 64 states of rate 1/4 through 3/4 and the rate 1/2 codes with constraint length 3 through 14. From the BER curves obtained for these codes, the coding gain is clarified as a function of coding rate and code constraint length. Main results obtained in this paper are as follows. (1) Among codes with 64 states, rate 1/3 code gives the maximum coding gain. (2) Coding gain of a rate 1/2 code increases as constraint length K becomes large. In the case of 8-level soft decision, for instance, the coding gain of the code with K14 reaches 7.1 dB at BER10^-6, which is close to the maximum coding gain obtained by an infinite constraint length code (estimated to be 7.9 dB).

The algorithm for the fast Walsh-Hadamard transform (FWT) derived from the Walsh-Paley function provides fast and simple processors which calculate the i-th Walsh transform F_i (i0, 1, , 2ⁿ1) of the natural-, dyadic-, and sequency-ordered Walsh-Hadamard transforms respectively according to control signals. This processor needs only 2・2ⁿ1 adder-subtractors (A/S) and is n times as fast as the one of the iterative type with n iterations. The characteristics of FWT with the term T resemble that of the assembly of digital filters. The digital filter representing FWT can be analyzed with z-transform and its transfer function is described as a simple function of z. The 3 dB bandwidth calculated by FWT is almost constant (about 0.85/T) for n2, 3, ...,10. All the Walsh transform F_i can be calculated by the processor with n・2ⁿ A/S, while the processor using delay lines needs only 2・2ⁿ1 A/S. The improved processor with delay lines deals with successive data and calculates all the Walsh transform or the Walsh spectrograms at the same time. These processors are effective for the recognition of speech and the analysis of signals.

In digital satellite communication systems, the high power amplifier (HPA) in the satellite repeater and earth station should operate in the vicinity of the saturation point, in order to effectively use output power. Recently, the adoption of constant envelope modulation schemes, like minimum-shift-keying (MSK) and MSK-type modulation (i.e. modified MSK), have been considered to avoid nonlinear harmful influence. This paper describes a new MSK-type modulation scheme, called Gaussian filtered and amplitude limited MSK (GLMSK). The idea behind GLMSK is to remove the envelope variations of the filtered MSK at the modulator by using a limiter with a small AM-PM conversion coefficient. GLMSK performance is calculated and estimated on some parameters, and the implemented GLMSK modem operation and performance are shown at a nominal data transmission rate of a 100 Mbit/s and an intermediate frequency of 1.7 GHz. Constant envelope GLMSK with a narrow spectral band can be achieved by using conventional high-speed MSK modem techniques.

The discrete Josephson junction transmission line which has N series Josephson junctions in each loop is discussed by computer simulation. It is found that a single quantum in this line can be stuffed in about one loop even if the inductance of each loop decreases to be negligibly small, i.e., the contribution of flux to a fluxoid is decreased. In this line the very small size fluxoid quanta with little flux can be employed as information bits. The resistive Josephson transmission lines whose loops have resistances in series are also discussed. In the resistive lines various operations are possible because of relaxation of the quantization conditions.

The new profile-control technique was investigated in regard to the modified CVD method to reduce OH-ion concentration in GeO₂-P₂O₅-doped silica graded-index optical fibers. The new profile-control technique, in which only exygen flow rate is varied for a parabolic radial index profile, was successfully attained by utilizing effects of exygen on properties of the deposited glass layer. Amounts of GeCl₄ gases, which do not react with oxygen, exist in the heated zone of a supporting tube and can effectively eliminate H₂O contaminations by the new profile-control technique. A GeO₂-P₂O₅-doped silica graded-index optical fiber was obtained, that simultaneously has a high transmission bandwidth of 1.5 GHz・km and OH-ion absorption loss of 0.9 dB/km at 1.39 µm.

A nonlinear model of the basilar membrane (BM) is examined for the two-tone suppression (TTS). The authors tested the BM displacement, velocity and acceleration as a possible candidate of the stimulus to the hair cell. Only in the BM acceleration TTS was observed, that has quite similar tendencies to physiological data.

A fault location test set applicable to a submarine cable has been developed. This is capable of estimating the distance between a fault and the repeater adjacent to it by utilizing the nonlinear modulation characteristic of the repeater which fluctuates periodically at a fixed frequency.

This letter presents experimentally that the power gain of axial-mode helical antenna can be improved by installing a relatively thin dielectric cylinder outside the helix, and also presented in some theoretical consideration to clarify the above effect approximately.

An improved numerical solution method for Bayesian estimation in nonlinear image restoration is presented. The constraint condition for the previous Bayesian method is also derived only in a simplest case. This method are shown to be more practical than the previous restoration methods.