電気学会論文誌C(電子・情報・システム部門誌)
Online ISSN : 1348-8155
Print ISSN : 0385-4221
ISSN-L : 0385-4221
122 巻, 1 号
選択された号の論文の24件中1~24を表示しています
  • 奥村 次徳
    2002 年 122 巻 1 号 p. 1
    発行日: 2002/01/01
    公開日: 2008/12/19
    ジャーナル フリー
  • 塩島 謙次
    2002 年 122 巻 1 号 p. 2-5
    発行日: 2002/01/01
    公開日: 2008/12/19
    ジャーナル フリー
  • 塩島 謙次
    2002 年 122 巻 1 号 p. 6-9
    発行日: 2002/01/01
    公開日: 2008/12/19
    ジャーナル フリー
  • 梅沢 仁, 谷内 寛直, 有馬 拓也, 石坂 博明, 藤原 直樹, 大庭 誉士和, 立木 実, 川原田 洋
    2002 年 122 巻 1 号 p. 10-16
    発行日: 2002/01/01
    公開日: 2008/12/19
    ジャーナル フリー
    Utilizing hydrogen-terminated surface-conductive layer as channel, high performance diamond FETs are demonstrated. High frequency operations of diamond transistors have been realized. 2μm gate diamond MESFET shows fT of 2.2GHz and fmax of 7GHz. Due to the effect of gate insulator insertion in series to source-gate capacitance of surface-channel, the source-gate capacitance of diamond MISFET is reduced by a factor of 2 as that of MESFET. The fT and fmax of 0.7μm gate diamond MISFET reach 11 and 18GHz respectively. Analysis of diamond short channel effect is also carried out. Down to 70nm channel diamond MISFETs are realized by utilizing new FET fabrication process for gate minimization. The short channel FETs with thick gate insulator more than 25nm shows short channel effect such as drain current increase by drain biasing, positive shift of VTH and degradation of subthreshold slope S. By the reduction of the thickness of gate insulator less than 20nm, VTH shift and degradation of Sare suppressed.
  • 木本 恒暢, 宮本 直, 松波 弘之
    2002 年 122 巻 1 号 p. 17-22
    発行日: 2002/01/01
    公開日: 2008/12/19
    ジャーナル フリー
    High-energy (MeV) implantation of Al+ and B+ into SiC epilayers has been investigated. A 3μm-deep pn junction is formed by room-temperature multiple Al+ or B+ implantation with implantation energies up to 6.2MeV or 3.4MeV, respectively. Rutherford backscattering channeling measurements have revealed that implantation-induced damages can be reduced close to a virgin level by high-temperature annealing at 1700_??_1800°. High electrical activation ratios over 90% have been achieved for both Al+- and B+-implanted layers by annealing at 1800°. Mesa pn diodes with a 15μm-thick n-layers formed by MeV implantation have exhibited high breakdown voltages of 2860_??_3080V. The reverse characteristics of the diodes have been substantially improved by increasing annealing temperature up to 1800°. The diode performance is discussed with the results of deep level analyses near the junctions.
  • 正戸 宏幸, 池田 義人, 松野 年伸, 西井 勝則, 井上 薫
    2002 年 122 巻 1 号 p. 23-28
    発行日: 2002/01/01
    公開日: 2008/12/19
    ジャーナル フリー
    Novel high breakdown voltage AlGaN/GaN HFETs have been developed by using a selective thermal oxidation process for the device isolation of AlGaN/GaN hetero-structure and by using silicon nitride as a dielectric film. The leakage current of thermal oxidation isolation process between two active islands separated by 6 micron drastically reduced 5 order of magnitude smaller than that of conventional mesa isolation. The breakdown voltage between islands of over 400V was obtained. The reverse leak current of gate to drain with SiN film deposited by plasma CVD method exhibited drastic reduction of 3 order of magnitude smaller as compare with SiO2 film. The fabricated 1.3μm-gatelength HFET showed maximum drain current of more than 500mA/mm, maximum transconductance (gmmax) of 130mS/mm and excellent pinch off characteristics at relatively high drain voltage of over 120V. These results indicate that AlGaN/GaN HFETs by using of thermal oxidation process and with SiN dielectric film are quite suitable for high frequency and high power applications.
  • 笠原 健資, 羽山 信幸, 国弘 和明, 宮本 広信, 高橋 裕之, 安藤 裕二, 中山 達峰, 岡本 康宏, 大野 泰夫, 葛原 正明
    2002 年 122 巻 1 号 p. 29-35
    発行日: 2002/01/01
    公開日: 2008/12/19
    ジャーナル フリー
    We investigated the resistivity of Ti/Al contact to n-AlGaN/GaN heterostructures changing the annealing temperature and the Al-fraction for high performance short channel FETs. A low resistivity of 1×10-7 Ω cm2 was obtained at an annealing temperature of 650°C with a 10nm Ti and 200nm Al structure for 1019cm-3 doped n-AlGaN/GaN. The mechanism of ohmic contact is studied by TEM observation. Surface oxide is removed when the thin Ti layer makes Al3Ti grains during 650°C annealing. After the oxide is removed, pure Al metal makes a low resistance contact to AlGaN/GaN heterostructures. K-Band Microwave performance of AlGaN/GaN HJFETs grown by MOVPE on sapphire substrate is also reported. A 0.25-μm T-shaped gate is fabricated by two-layer resist process using electron-beam lithography. The fabricated devices have an intrinsic cut-off frequency fT of 57GHz and an intrinsic maximum oscillation frequency fmax of 114GHz. The continuous wave output power density is 1.58W/mm with 5.9dB of gain and 11% power added efficiency at 25GHz. This output power density is about two times higher than GaAs-based pseudomorphic FETs.
  • 横山 清子, 大見 直史, 森本 陽子, 水野 康文, 高田 和之
    2002 年 122 巻 1 号 p. 36-41
    発行日: 2002/01/01
    公開日: 2008/12/19
    ジャーナル フリー
    In this paper, parameters of heart rate variability using wavelet transform, are employed to analyze mental and physical workload during long-distance driving. In experiment, subjects have driven 400km distance by automobile or motorcycle. The heart rate time series was recorded using Holter ECG. The power spectral density function of the heart rate time series was calculated at six-second interval. In this results, the power of the low-frequency component and the peak frequency of the high-frequency component were considered to evaluate mental and physical workload during long-distance driving of automobile and motorcycle.
  • 川田 昌武
    2002 年 122 巻 1 号 p. 42-51
    発行日: 2002/01/01
    公開日: 2008/12/19
    ジャーナル フリー
    Electroencephalogram (EEG) has been generally known as a non-destructive method to examine the function of human brain. The aim of this paper is to propose a new method using Wavelet Transform to analyze and to identify the components of Movement - Related Cortical Potential (MRCP) which is evoked accompanying with the human voluntary muscle behaviors. In this study three healthy-volunteers were participated. The EEG was recorded with the 5-channel referential derivations, Fz, Cz, C3, C4, Pz, referenced to the right earlobe based on the international 10-20 system. The voluntary activity was recorded with the electrode arranged on the palm side of the metacarpophalangeal joint when the subjects flexed the right-hand index finger and eyes closed.
    I have found the insights into the temporal dynamics of MRCP. About 10Hz signals were recorded at all electrodes when the voluntary activity started. The higher level signals of about 10Hz were recorded at Fz from about 0.4s after the activity started. About 5Hz signals appeared from about 0.9s before to 0.3s after the activity started. This method is also useful in point of the number of averaging, because it's not needed many subjects and trials to identify the components of the MRCP as shown in the past publications.
  • 鈴木 明美, 内山 明彦
    2002 年 122 巻 1 号 p. 52-59
    発行日: 2002/01/01
    公開日: 2008/12/19
    ジャーナル フリー
    Tissue impedance reflects the individual information related to the tissue structure and the function of the tissue. A fixed coefficient regularization algorithm is one of the impedance reconstruction algorithms to improve the ill-conditioning problem of the Newton-Raphson algorithm. However, the determination of a good coefficient needs a lot of experimental data and a long period of computation time. Because a good smoothing coefficient has to be manually chosen from a number of coefficients, and the fixed coefficient regularization algorithm sometimes distorts the information, or fails to obtain any effect. In this paper, a new automatic algorithm with successive determination of variables is introduced. This algorithm automatically calculates the smoothing coefficient from the average value of the ill-conditioned matrix. We have evaluated the performance of this average variable algorithm through our impedance imaging system and two phantoms. It shows that the average variable algorithm yields efficient superior images, within the short reconstruction time, and is more practical in application; as compared to the fixed coefficient regularization algorithm.
  • 小矢 美晴, 溝尻 勳
    2002 年 122 巻 1 号 p. 60-65
    発行日: 2002/01/01
    公開日: 2008/12/19
    ジャーナル フリー
    We obtain important information from movement of a left ventricle in an assessment of a cardiac function. Many of methods have been proposed for myocardial motion analysis before now. However, they are only able to make rough estimate such as center-line method.
    In this paper, we propose the method to estimate precise displacements and strain distribution on an inter ventricular septum using warping and characteristic points matching.
    We show to be able to specify myocardial infarction using this method.
  • 川勝 真喜, 小林 宏一郎, 内川 義則, 斎藤 正男, 小谷 誠, 藤平 誠一, 尾形 久直
    2002 年 122 巻 1 号 p. 66-70
    発行日: 2002/01/01
    公開日: 2008/12/19
    ジャーナル フリー
    According to increasing number of SQUID in Magnetoencephalography (MEG) system, cryostat and the amount of liquid helium consumption have been getting larger. At some large institution with liquid helium production center, it has carried out collection and re-liquefaction of helium. However, gasified helium is discharged in many institutions, because (1) large-scale systems are necessary for re-liquefied, (2) the measurement noise will increase. We developed a helium collection and re-liquefied system for MEG measurement system. The system does not increase any noise and it consists with minimum units. We have used the system applying to 195-channel SQUID system over 6 month. The helium purchase quantity became half the amount.
  • 林 華, 王 暁秋, 呂 建明, 谷萩 隆嗣
    2002 年 122 巻 1 号 p. 71-77
    発行日: 2002/01/01
    公開日: 2008/12/19
    ジャーナル フリー
    Interleaving is a practical technique to enhance the error correcting capability of coding. It has been widely used in conjunction with error control coding for channels that exhibit bursty error characteristics. In turbo codes, the interleaver can spread out burst errors. In personal communication systems using turbo codes, the interleaver design is very important when the short frame transmission is taken into consideration. The main purpose of the interleaver is to decrease the correlation characteristic between the input information sequence and the output sequence of the interleaver by changing the order of the input information sequence.
    In this paper, we will present a block interleaver based on low correlation characteristic. By statistical analysis, the correlation coefficient of an interleaver is obtained from the fixed points in the interleaving process. Simulation results have shown that the proposed interleaver has lower correlation coefficient and better BER characteristic than the traditional row-column interleaver except for square-type interleaver.
  • Tao LU, Shunichiro OE, Hiroshi MURAI
    2002 年 122 巻 1 号 p. 78-83
    発行日: 2002/01/01
    公開日: 2008/12/19
    ジャーナル フリー
    Two-dimensional phase unwrapping is a key step in extraction of digital elevation models (DEMs) from interferometric synthetic aperture radar (IFSAR) data. Least-squares (LS) algorithm is one of major approaches to phase unwrapping, and weighted LS algorithm offers great potential if weights are appropriately obtained. However, the determination of weights is still a critical problem. As noise increases, binary weights becomes more difficult to be determined, because of high possibility of coexisting of noise pixels and non-noise pixels in same histogram bin. Hence instead of determining binary weights, this paper presents an effective approach to determining continuous weights by incorporating weighted LS algorithm into a minimization method. The continuous weights are estimated by minimizing a norm of the differences between real values and calculated ones. The proposed approach is illustrated through a simulative example in which the input data are corrupted by both uniformly distributed noise and scattered noise. By comparing errors of phase unwrapping by binary weights and those by continuous weights, the latter turns out to be superior to the former.
  • 澤 和弘, 大山 航, 若林 哲史, 木村 文隆
    2002 年 122 巻 1 号 p. 84-89
    発行日: 2002/01/01
    公開日: 2008/12/19
    ジャーナル フリー
    In this paper, we propose a new recognition procedure for a degraded character string printed on a steel plate for factory automation. The character string consists of alphanumerals and '-'. The procedure includes the image emphasis with Laplacian Gaussian filter, the extraction of the character string image, the segmentation-recognition with the dynamic programming, and the high accuracy character recognition. We evaluated the accuracy for 1036 images (8, 806 characters) scanned by the monochrome video camera at an actual steel producting factory line. The average recognition rate was 99.2% for the character recognition, and 91.6% for the character string recognition.
  • 無限次元LMIアプローチ
    池田 光治郎, 東 剛人, 内田 健康
    2002 年 122 巻 1 号 p. 90-98
    発行日: 2002/01/01
    公開日: 2008/12/19
    ジャーナル フリー
    In this paper, we propose a new method of L2-gain analysis and state feedback control synthesis for time-delay systems via infinite-dimensional LMI, which reflects explicitly infinite-dimensionality features of time-delay systems. First, we show an infinite-dimensional LMI condition for L2-gain analysis. The infinite-dimensional LMI condition depends two parameters, and difficult to solve. Next, we propose a method of solving such infinite-dimensional LMI conditions using a convex polyhedron. The method consists of two steps: the first step is to make a convex polyhedron, and the second step is to reduce an infinite-dimensional LMI condition to a finite-dimensional LMI condition using the convex polyhedron. Finally, we discuss a state-feedback control synthesis based on the infinite-dimensional LMI conditions for L2-gain analysis and the method of solving infinite-dimensional LMIs, when the feedback gain is constrained.
  • 藤原 伸行, 恩田 寿和
    2002 年 122 巻 1 号 p. 99-104
    発行日: 2002/01/01
    公開日: 2008/12/19
    ジャーナル フリー
    We have developed an observation system to register virtual property lines on a TV monitor. The system tracks markers set at regular intervals on the baseline of the observation area in the input image, calculates the camera parameters from the positions of the detected markers, and registers the property lines in association with the camera motion. In this paper, we describe the conditions of our application to use an actual observation seen, the camera calibration and the registration algorithm for setting up property lines, marker tracking, and practical experiments at an actual construction site.
  • 平澤 宏太郎, 橋本 雅之, 胡 敬炉, 村田 純一, 金 春植
    2002 年 122 巻 1 号 p. 105-115
    発行日: 2002/01/01
    公開日: 2008/12/19
    ジャーナル フリー
    Recently, a number of methods on neural network control have been studied. Universal Learning Networks (ULNs) have been proposed, as the name indicates, to provide a universal framework for the class of neural networks and moreover to model and control complex systems. In addition to calculation of the first order derivatives of the signals flowing in the networks that are necessary in gradient-based learning, the generalized ULN learning algorithm is equipped with a systematic mechanism that calculates their second or higher order derivatives. Using the higher order derivatives, robust controlled systems have been studied, that are not much affected by changes in the plant parameters. On the other hand, in the control fields it is requested to design the systems to meet the plural specifications concurrently such as quick response, quick damping and small steady state errors. Until now, there have been thoroughly discussed on how to make a controller to deal with such plural specifications. Model Matching and Model Reference Control system are the typical conventional methods for aiming at such a goal. But, these method have been mainly applied to linear control systems. In this paper, a new control method formulated by ULNs has been proposed, which can overcome the conventional nonlinear problems by using the impulse response as the extended criterion function. In addition the impulse response can be easily calculated by the higher order derivatives of ULNs. In simulations of a nonlinear crane control system, it is shown that the proposed method is superior to the commonly used neural networks control.
  • 高橋 信補, 谷 繁幸, 石井 俊夫, 加藤木 工三
    2002 年 122 巻 1 号 p. 116-123
    発行日: 2002/01/01
    公開日: 2008/12/19
    ジャーナル フリー
    A distinction method is proposed for distinguishing the fuel property (volatility) of automotive engines based on a dynamic reference model. This method can determine whether the fuel property is heavy or light.
    The reference fuel response when an engine is accelerating is calculated based on the fuel injection amount and the dynamic reference model, which provides a medium response for heavy and light properties. The actual fuel response is detected based on the air intake amount and the measured air-fuel ratio. If the actual response is faster than the reference response, the fuel property is judged as light. If the actual response is slower, the fuel property is judged as heavy. This judgment is well-founded because the faster the fuel response, the lighter the fuel property. Simulations showed that the proposed method determines the fuel property accurately.
  • 村上 慎治, 矢野 京二, 高岡 佳市, 橋口 幸男, 入江 俊充, 後藤 純, 脇迫 仁, 長谷川 勉
    2002 年 122 巻 1 号 p. 124-132
    発行日: 2002/01/01
    公開日: 2008/12/19
    ジャーナル フリー
    This paper reports on the automation technique of live line re-connection included in the electricity distribution work using robot manipulator. There are two major technical issues ;visual measurement of the position of the object in an unstructured outdoor environment and insertion of the wire terminals to a metal sleeve for connection. As a distinctive structure of the work, the distance to the object from CCD camera is quite shorter than the distance to the background. Therefore, we can obtain a clear image of the object while the background is out of focus. We have developed a method of object detection by template matching combined with differential filtering technique of image.The direction of the electric wire and the direction of the sleeve were aligned and the tip of the wire is positioned to the tip of a sleeve using this method. Once the position is reached, the insertion process starts using the compliance control with estimation of the contact force In addition, we have developed procedures for error recovery based on the state transition analysis.
  • 川崎 紀宏, 長名 優子, 萩原 将文
    2002 年 122 巻 1 号 p. 133-140
    発行日: 2002/01/01
    公開日: 2008/12/19
    ジャーナル フリー
    In this paper, we propose a chaotic associative memory for successive learning (CAMSL) using internal patterns. In the CAMSL, the learning process and the recall process are not divided. When an unstored pattern is given to the network, the CAMSL can learn the pattern successively. The CAMSL distinguishes an unstored pattern from the stored patterns. When a stored pattern is given, the CAMSL recalls the pattern. When an unstored pattern is given, the CAMSL changes the internal pattern for the input pattern by chaos and presents the other pattern candidates. When the CAMSL cannot recall the desired pattern, it learns the input pattern as an unstored pattern. We carried out a series of computer simulations and confirmed the effectiveness of the CAMSL.
  • 寳来 俊介, 山田 泰司, 合原 一幸
    2002 年 122 巻 1 号 p. 141-147
    発行日: 2002/01/01
    公開日: 2008/12/19
    ジャーナル フリー
    In this paper, we propose the Iso-directional Recurrence Plots(IDRP) and its techniques. In IDRP, characterization of vectors on orbits reconstructed from the time series data is represented on a twodimensional plane, similarly to the Recurrence Plots(RP). In addition, we also propose Iso-directional Neighbors Plots(IDNP) which is the product set of RP and IDRP. IDNP shows the set of points that are neighbors and possess similar directions. We apply the proposed methods to various data such as chaotic and stochastic time series, and compare IDNP with the method proposed by Wayland et al. for recognizing determinism. The property of the results is similar to that of Wayland method, and thus IDNP is applicable to the analysis of determinism. Since the proposed methods in this paper are based on RP, not only quantitative analysis of determinism but also qualitative observation of the behavior of time series are possible with time series data.
  • Mohamed Anis Ben Amor, Watanabe Shigeyoshi
    2002 年 122 巻 1 号 p. 148-155
    発行日: 2002/01/01
    公開日: 2008/12/19
    ジャーナル フリー
    This paper is concerned with the development of a car-following model for traffic simulation. A new approach to emulate the driver's decision making mechanism is proposed. This approach uses the powerful techniques of fuzzy reasoning to deal with uncertainty and to smoothly blend behaviors induced by simultaneous goals, while being aware of the long history of equation based modeling. Moreover, an original algorithm is implemented in this system to imitate the driver's self-learning from his on-road driving experience and to mimic his maneuvers over the brake and the accelerator pedals which reflect his approximation of the appropriate acceleration for the actual situation.
  • 佐野 雄二, 依田 正樹, 向井 成彦, 島村 光明, 小野 芳明, 金澤 寧
    2002 年 122 巻 1 号 p. 156-162
    発行日: 2002/01/01
    公開日: 2008/12/19
    ジャーナル フリー
    The authors showed that the residual stress on the inside wall of a pipe with inner diameter of 15 mm could be improved by an underwater laser ablation process, in which laser pulses with energy of 130 mJ from a Q-switched and frequency-doubled Nd: YAG laser were irradiated with diameter of 0.7 mm and with 18 or 27 impacts for every 1 mm2 on material surface. A compact total-reflection mirror was invented, which reflected laser pulses in water with broad angle of incidence, and a device of 14 mm in outer diameter was assembled using the mirror in order to irradiate laser pulses to the inner surface of the pipe. The laser irradiation condition was optimized through experiments with Alloy 600 plates, and was applied to the cylindrical test piece of 15 mm in inner diameter. The results showed that a compressive residual stress was built on the inner surface of the test piece and the effect penetrated up to 1 mm from the surface. The compressive residual stress on surface is expected to prevent stress corrosion cracking (SCC) of structural materials in corrosive environments and to enhance fatigue strength of components under cyclic load.
feedback
Top