|
1. |
Simultaneous approximation for two kinds of system characteristics using descent method |
|
Electronics and Communications in Japan (Part I: Communications),
Volume 64,
Issue 6,
1981,
Page 1-9
Wimat Kiranon,
Toichi Machida,
Masao Yoneyama,
Preview
|
PDF (461KB)
|
|
摘要:
AbstractThe descent method is applied for simultaneous approximation for system characteristics with two different characteristic requirements. With given initial parameters, the descent directional vectors which derease two performance functions simultaneously are determined from the gradient vector of the performance function of each characteristic and the successive approximation is made from approximated parameters. In the algorithm, both performance functions always decrease at each step under given conditions. Next with newly defined weighting coefficients, both performance functions are decreased simultaneously at a given rate. The convergence limit of this simultaneous approximation is made clear. The effectiveness of the present method and examples of weighting coefficients are discussed.
ISSN:8756-6621
DOI:10.1002/ecja.4410640602
出版商:Wiley Subscription Services, Inc., A Wiley Company
年代:1981
数据来源: WILEY
|
2. |
Adaptive detection of unknown signal in colored noise of unknown parameters |
|
Electronics and Communications in Japan (Part I: Communications),
Volume 64,
Issue 6,
1981,
Page 10-19
Shigehiko Miki,
Kunihiko Yoshida,
Seiko Kaneku,
Preview
|
PDF (662KB)
|
|
摘要:
AbstractThis paper describes the adaptive detection of an unknown signal in colored noise. Given the statistical parameters of the observed noise, the Gremlin machine is well known as a learning scheme for the signal pattern. When this scheme is extended to unknown noise parameters, the correlation of observations based on noise poses a problem.Here, additive colored noise is modelled by a self‐recursive process of Gaussian type of past‐dependent width k and the dispersion of its random component is unknown. Its coefficient group is taken as one of the elements of a known finite collection. Further, k does not exceed the observation space width, etc. Based on these premises, first an optimal supervised learning scheme is derived using a Bayesian scheme. Next, the theoretical convergence is studied for this learning scheme and it is shown that, under certain conditions, the detection scheme converges to that for all known parameters. Further, for a special case, the theoretical error probability of a case with assumed estimated noise parameters is computed. The importance of study of correlation between observations is pointed out at the same time the case in which the true values of the noise parameters are not included in the known finite collection is also studied. Thus, the effectiveness of the present learning scheme is clearly shown. Finally, characteristics of the learning scheme are also studied by computer simulat
ISSN:8756-6621
DOI:10.1002/ecja.4410640603
出版商:Wiley Subscription Services, Inc., A Wiley Company
年代:1981
数据来源: WILEY
|
3. |
Linear and nonlinear analysis of electromyography time series |
|
Electronics and Communications in Japan (Part I: Communications),
Volume 64,
Issue 6,
1981,
Page 20-28
Naohiro Ishii,
Akira Iwata,
Nobuo Suzumura,
Yoshiki Yamazaki,
Preview
|
PDF (429KB)
|
|
摘要:
AbstractThe purpose of this paper is to extract the information of the posture control of an experimenter standing on a plate. In the experiment, the object is vibrated sinusoidally in the vertical direction on the plate and the electromyography (EMG) and force on the plate are measured. A method of estimating the delay time of the system, the system response of the position component and the velocity component of the input is developed. The pseudo‐white noise is constructed by changing the angular frequency of the input. It is shown that the EMG and the force are well represented by a nonlinear input‐output relat
ISSN:8756-6621
DOI:10.1002/ecja.4410640604
出版商:Wiley Subscription Services, Inc., A Wiley Company
年代:1981
数据来源: WILEY
|
4. |
Speech quality of adaptive delta modulation (ADM) codec with syllabic companding |
|
Electronics and Communications in Japan (Part I: Communications),
Volume 64,
Issue 6,
1981,
Page 29-38
Kenzo Itoh,
Nobuhiko Kitawaki,
Kazuhiko Kakehi,
Shinji Hayashi,
Preview
|
PDF (629KB)
|
|
摘要:
AbstractDelta modulation with syllabic companding is one of the voice digitalization systems for which experimental studies have been conducted to evaluate speech quality. First, appropriate circuit constants are provided by use of the objective evaluation measure S/NOfor the experiments and then studies are made on the features of the quantization noise in the system. Second, the cepstrum distance measure obtained from linear prediction analysis is introduced as an objective evaluation measure and its relationship to subjective evaluation is also studied. The following conclusions are made. (1) Lowering the sampling frequency to 32 kHz has little effect on articulation and can provide fair or better than fair quality if appropriate signal input levels and received speech volumes are maintained. (2) The effect of circuit noise at about 2.0 mV at the encoder input on voice articulation is deemed to be equivalent to having the same circuit noise at the decoder output. (3) The detectability threshold of transmission errors is about 2.4 × 10−4, which proves that the system is less prone to code errors than the log‐PCM system. In addition, the coincidence of the cepstrum distance measure with the subjective evaluation value is better than that of the segmental
ISSN:8756-6621
DOI:10.1002/ecja.4410640605
出版商:Wiley Subscription Services, Inc., A Wiley Company
年代:1981
数据来源: WILEY
|
5. |
Analysis of monopole antenna erected on circular conductor |
|
Electronics and Communications in Japan (Part I: Communications),
Volume 64,
Issue 6,
1981,
Page 39-46
Jiro Iwashige,
Preview
|
PDF (497KB)
|
|
摘要:
AbstractA vertical monopole antenna erected on the base of a circular cone is analyzed by the equivalent edge current (EEC) and the geometrical theory of diffraction (GTD). The cone may be converted into a disk, column or truncated cone. The current distribution and input impedance are calculated by the moment method. The multiple diffraction on the circular plane is considered in the calculation of radiation patterns by the EEC method (or GTD). The experimental results agree approximately with the calculation.
ISSN:8756-6621
DOI:10.1002/ecja.4410640606
出版商:Wiley Subscription Services, Inc., A Wiley Company
年代:1981
数据来源: WILEY
|
6. |
A receiver system with a pitch synchronized interpolator |
|
Electronics and Communications in Japan (Part I: Communications),
Volume 64,
Issue 6,
1981,
Page 47-57
Hideo Murakami,
Kazunari Irie,
Preview
|
PDF (749KB)
|
|
摘要:
AbstractIn mobile communications (e.g., vehicular telephony) instantaneous interruption is caused by rapid fading while passing through a downtown area where buildings are dense. The interference caused by this fading markedly deteriorates the quality of communication. To improve the quality of communication, the authors propose a method called pitch synchronized interpolation which interpolates the transient signal by employing the time intervals between speech signals. This paper deals with a receiver system which incorporates this pitch synchronized interpolator. In the pitch synchronized detection or interpolation process of this system, the received speech signal, after undergoing A/D conversion, is fed to a sequential delay memory. This delay memory speech signal is processed by a digital signal processor. Various methods and parameters used here are discussed (e.g., the pitch period detecting method, the frequency of execution of pitch period detection, coding methods, maximum interpolation time, the damping coefficient for damping the interpolation output level to zero when the interruption period exceeds the interpolation time and the design parameters for the speech specimen period). Further, the necessary condition for these parameters are clearly defined. Also, based on the studies of these parameters, a prototype model of pitch synchronized interpolator is designed and fabricated with a reduced pitch period detection cycle of 4 ms and the desired decrease in interruption noise is confirmed.
ISSN:8756-6621
DOI:10.1002/ecja.4410640607
出版商:Wiley Subscription Services, Inc., A Wiley Company
年代:1981
数据来源: WILEY
|
7. |
A novel radar reflector with variable RCS |
|
Electronics and Communications in Japan (Part I: Communications),
Volume 64,
Issue 6,
1981,
Page 58-64
Hiroshi Kotajima,
Hirotoshi Fujita,
Preview
|
PDF (589KB)
|
|
摘要:
AbstractIt is convenient to have a standard radar reflector with a variable radar cross section when the reflection characteristics of radar reflectors and various objects are measured or when performance tests of radar systems are carried out. There have been several proposals for such radar reflectors in the past. However, since all of them make use of reflection of the conductor surface, their reflection efficiency is too low to be practical. The principle of the radar reflector reported in this paper is completely different from that of conventional ones and makes use of focusing and spherical symmetry of the Luneburg lens. The reflection efficiency is quite high and the radar cross section can be varied in response to the rotating angle of the lens. First a method is proposed to obtain a desired radar reflector and the design and fabrication method are described. Further, by means of a prototype development and subsequent measurement is is shown that the proposed method is realizable. As an example, the prototype reflector for a horizontally polarized wave of 9375 MHz is made using a Luneburg lens with diameter of 30.5 cm. The radar cross section can be varied from 43.28 m2to 0.42 m2as designed.
ISSN:8756-6621
DOI:10.1002/ecja.4410640608
出版商:Wiley Subscription Services, Inc., A Wiley Company
年代:1981
数据来源: WILEY
|
8. |
Hierarchical structure of telecommunication network with independent transversal trunks |
|
Electronics and Communications in Japan (Part I: Communications),
Volume 64,
Issue 6,
1981,
Page 65-73
Minoru Akiyama,
Yoshiaki Tanaka,
Mikio Yamashita,
Makoto Iio,
Preview
|
PDF (583KB)
|
|
摘要:
AbstractThe main objective of setting up the hierarchical structure for telecommunication network is to concentrate the traffic in order to increase the efficiency of transmission lines, thereby reducing the total cost of the network. The optimum structure of the network is determined by such factors as traffic, its flow pattern, the costs of transmission line and switching equipment. The costs of components of the switching network change rather rapidly due to recent technological development. The change of cost and the role of transversal trunks in handling traffic calls have made the reconsideration of installing transversal trunks worthwhile. This paper presents a geometrical model for the Japanese public telephone network. The optimization of the hierarchical structure of the network is performed taken into consideration such parameters as costs of transmission lines, switching facility, and traffic threshold for transversal trunk. It is concluded that the three‐stage network with end offices, toll centers, and district centers seems appropriate in the near future. The district center of this network should be about the same scale as the existing regional center. The traffic threshold for installing the transversal trunk should be almost several tens of erlang
ISSN:8756-6621
DOI:10.1002/ecja.4410640609
出版商:Wiley Subscription Services, Inc., A Wiley Company
年代:1981
数据来源: WILEY
|
9. |
Theoretical method for the analysis of queueing system with overflow traffic |
|
Electronics and Communications in Japan (Part I: Communications),
Volume 64,
Issue 6,
1981,
Page 74-83
Jun Matsumoto,
Yu Watanabe,
Preview
|
PDF (544KB)
|
|
摘要:
AbstractThe theoretical analysis and design of queueing systems with overflow traffic are presented. The non‐random traffic is approximated by an interrupted Poisson process with which the 3‐moment matching method is used for obtaining the average, variance and the 3rd‐order moment. The analysis of a queueing system with finite capacity with overflow traffic can be considered to be a G/M/1 type queueing matrix to which the interrupted Poisson process and known equations apply. For analysis, the overflow traffic moments and formulas for the overflow probability and average waiting time of the finite‐capacity queueing system with overflow traffic are so derived that they can be used in their practical forms very easily. In the case of finite‐capacity queueing systems, the theoretical overflow probability and average waiting time are derived for single‐ and multiple‐overflow traffic models, overflow and random traffic models as well as the two‐stage overflow traffic model. From comparison with the simulated experimental values, the accuracy of the theoretical approximations is shown
ISSN:8756-6621
DOI:10.1002/ecja.4410640610
出版商:Wiley Subscription Services, Inc., A Wiley Company
年代:1981
数据来源: WILEY
|
10. |
Computer simulated reproduction process with nonlinear magnetic heads |
|
Electronics and Communications in Japan (Part I: Communications),
Volume 64,
Issue 6,
1981,
Page 84-91
Yasuhiro Koshimoto,
Tetsuo Mikazuki,
Preview
|
PDF (577KB)
|
|
摘要:
AbstractTo upgrade magnetic recording equipment consideration anticipated is given to the construction of an electromagentic conversion using high‐permeability magnetic media in order to obtain high output with high resolution power. In the case of obtaining high reproduction output, the nonlinear magnetism of the reproduction head can affect the reproduction properties, but presently there is no report on the head operation with nonlinearity taken into account. In this paper, the reproduction process with nonlinear magnetic head is simultaed numerically and the magnetic field, permeability and magnetic flux distributions in the head as well as the effect of the nonlinearity of the head core material are analyzed. Results show that the core permeability decreases locally because of local concentration of magnetic flux due to the presence of recorded media and the permeability decreases significantly when the medium position giving rise to maximum output lies in the vicinity of the gap. The reason for degradation of the reproduction characteristics in heads using material with low saturation flux density can be explained by the fact that the low‐permeability region in the vicinity of the gap gives rise to an effective gap incre
ISSN:8756-6621
DOI:10.1002/ecja.4410640611
出版商:Wiley Subscription Services, Inc., A Wiley Company
年代:1981
数据来源: WILEY
|
|