The U. S. Federal Communication Commission (FCC) has required the routine specific absorption rate (SAR) evaluation of portable telephones since 1997. The applicable SAR limit is 1.6W/kg as averaged over any one gram of tissue, defined as a tissue volume in the shape of a cube. Because of the irregular surface of the human head and tissue heterogeneity, computing the one-gram averaged spatial peak SAR in the shape of a cube is not so easy. A lack of standardized procedure for computing the one-gram averaged spatial peak SAR is resulting in confusion in the SAR assessment. In this paper, the SAR in an anatomically based human head model is computed using the finite-difference time-domain (FDTD) method, and the uncertainties of different four kinds of one-gram average procedures are evaluated numerically. The results show that an uncertainty up to 2.4 times may exists at 900 MHz. A nearly 1×1×1 cm cube with a mass of just one gram, but not containing the air cells, seems to be reasonable for the one-gram average of SARs.

View full abstract

With the increase of electromagnetic interference (EMI) problems in digital information and communication devices, ferrite cores are often used to reduce the EMI level and improve the system immunity. For this reason, a need for measurement of complex permeability μ_γ of ferrite material, which determines the noise reduction effect, has been growing. In this paper, a practical transmission-line method is proposed for measurement of the complex permeability of ferrite cores in a broad-band frequency by scattering-parameter techniques. This measurement method tolerates the existence of air spaces between the measurement material and the coaxial-typed holder. Its validity is verified by fitting the measured results for commercially available ferrite cores having various dimensions to the Naito's formula and examining whether or not the measured results satisfy the Snoek's limit.

View full abstract

Using a photoacoustic effect, a non-destructive and non-contact defect detection has been studied. A low power semiconductor laser which has a power of 10mW is used as a light source. Because a photoacoustic signal is very low in this case, a proper condition of an acoustic sensor which detects the photoacoustic signal needs to be find. As a result of measuring a propagation characteristic of the photoacoustic signal, the photoacoustic signal propagates forward intensely, so the acoustic sensor is set in front of an object. An increase of the photoacoustic signal called an edge effect is observed on an edge of the object. A small defect which has a diameter of about 1.5mm can be detected. It has a high resolution in the defect detection and some adjacent defects which have a gap of 0.1mm can be detected because the laser light can be focused to the extremely small spot.

View full abstract

We present an active vision system which chooses and fixates the scene with the largest information. We use Shannon information to measure image features, and present an information map to quantify a saliency map. Particularly, we use intensity, hue and saturation as the image features to be quantified. The system also has a task dependent attention selection ability. To achieve a given task, the system selects some features, uses them to calculate the image information, constructs an information map, focuses on the area with the largest information, and zooms in it. Experimental results demonstrate that the system can acquire images from a real scene with the highest information and generate compact images with different resolution.

View full abstract

Some real-world entities such as human beings have different aspects in different situations (multiple aspects). They can decide one of their aspects dynamically and autonomously (autonomy). They can also initiate some behaviors according to their surroundings (pro-activeness). We call those entities multiple objects. The conventional object-oriented models can not represent the multiple objects appropriately. In this paper, we propose a novel agent model, called EAMMO, to represent the multiple objects, and implement an interpreter for the description language of EAMMO in Java. Our agent model consists of three types of agents as follows: (1) an upper-agent describes an autonomous and dynamic object; (2) a lower-agent describes a reflective and static object; and (3) an environmental-agent describes the autonomous surroundings of the other agents. As an example, we model soccer players (a kind of multiple objects in real world) in EAMMO. Three kinds of soccer agents are constructed with different powers. We simulate the soccer games to confirm the effectiveness of our model.

View full abstract

We propose a method for extracting human regions from a sequence of three images using motion, head detecting and simple human countor model. First the method calculates the subtraction between successive frames to find a moving region. Next it classifies the edge segments into curves and lines, and then represents the curves as circles of candidates for a human head. Simple contour model of human is placed at the head position which the moving region and the circles decide, then converges to make a rough human region. To obtain a more accurate region, the image is segmented into regions with different color, and the regions are merged to make a human region. We conducted experiment using real scene images to evaluate the proposed method, and showed that human regions were almost extracted successfully.

View full abstract

It is known that about three thousand Korean (Hangul) characters are needed to fully display or print for POS terminal in Korea. If POS terminal houses all the characters in character generator memory size increases and increases production cost, therefore conventional POS terminal houses only limited number of characters in its character generator. Although missing characters are supplemented by numeric or English character, when sold products are registered by POS it was difficult for customer to read printed or displayed products name and when POS prints out report it was difficult for operator to read printed report. These facts caused serious MMI (Man Machine Interface) related problems. To resolve this problems, by analyzing the Korean character structure and using about 150 letters this paper presents an algorithm to generate 11, 172 Korean characters and its code. This presentation surely resolves the MMI related problems.
It is also referred to the fact that because this method requires small size of memory and uses fairly simple algorithm this can be used by various microprocessor embedded equipment to improve its MMI in Korea.

View full abstract

This paper proposes a method of constructing a log-periodic microstrip antenna. Log-periodic technique is considered to be useful to improve the characteristics of a microstrip antenna (MSA) which is narrow band in nature, without preventing the advantage of low-profile and light weight. The radiaiting elements are co-ax fed and arranged on one side of the common feed line behind the ground plane. This way of arrangement necessarily decreases the element spacings, and increases the mutual coupling between elements. The coupling effect has acted to fill up the inactive frequency regions. Larger coupling between elements due to the necessarily dense arrangement gives advantage in obtaining wide-band frequency characteristics. An example of 5 elements case proves that the present method is useful giving gain of 6 to 10 dBi over about 20%, bandwidth.

View full abstract

Microactuators with multi degrees of freedom, which can provide a lot of benefits to precise engineering, are required for industrial applications. We propose two new actuators (V type and Δ type) which can move in x, y and θ directions. They consist of piezoelectric and electromagnetic actuators. The proposed actuators realize high resolution by using the piezoelectric actuators. We describe the structure and principle of the proposed actuators, and the feasibility of the proposed actuators is studied. The resolution in x direction was obtained to be 70nm and that in θ direction was 0.5 μrad by the Y type actuator. The resolution in x direction was obtained to be 90nm and that in θ direction was 0.6 μrad by the Δ type actuator.

View full abstract

In this paper, the method for the heart rate variability evaluation which uses the frequency information on electrocardiogram (ECG) is proposed. The proposal method does not extract the time series of the R-R intervals from the ECG, so the processing greatly becomes handy, and the method has the advantage of suitable for the real-time processing. The try by which the parameters of the heart rate variability are extracted directly from the frequency information of the ECG is not done so far. The heart-rate (HR) and the coefficient of variance of the RR intervals (CVrr) was targeted as parameters of heart rate variability. In the proposal method, the amplitude spectra corresponding to the R wave of the electrocardiogram is extracted from the all spectra of the ECG. The heart rate is obtained from the peak frequency, and the CVrr is obtained by using the standard deviation and the peak frequency of the spectra. The correlation coefficient of the parameter of the method and the parameter of the proposal method was obtained so far, so the correlation coefficient of the HR was 0.99 and the correlation coefficient of the CVrr was 0.92. It was able to be confirmed to be able to calculate the HR and CVrr by high accuracy by the proposal method. The possibility to the arrythmia diagnosis and the possibility of the evaluation by a new aspect which used frequency information of ECG was illustrated.

View full abstract

The micro-gravity environment causes some physiological changes on human body. Spatial disorientation (SD) is one of the most important aspects and it has strongly been desired to develop a monitoring method in current space medicine. Referring to the previous physiological and psychological studies, the pupil and eye movement have potential to be an objective indicator for the SD phenomenon.
In this paper, at first, a system for measuring pupil size and eye movement, by digital processing of eye images, was developed. Then, the system was applied to rotating-stripe and rotating-chair experiments, each of which causes SD of the subject, to evaluate the changes in pupil activity and eye movement during the SD phenomenon.

View full abstract

The propagation mechanism of very low latitude (geomagnetic latitudes of less than 20°) whistlers is poorly understood. This paper examines the propagation characteristics of non-ducted propagation, but we use three-dimensional ray tracing for realistic ionosphere/magnetosphere models, the electron density profile with latitudinal and longitudinal gradients and the International Geomagnetic Reference Field (IGRF) magnetic field model instead of the conventional dipole model. By assuming small possible tilts (in the latitudinal and longitudinal direction) of the initial wave normal angle in the input southern hemisphere, we have found that it is possible for us to detect, at a very low latitude position in the northern ionosphere, some whistler rays penetrate through the ionosphere, but some others can not. On the basis of systematic analysis of important parameters, we come to the general conclusion that the occurrence local times of whistlers are depend on the ionospheric model, especially latitudinal and longitudinal gradients of the electron density.

View full abstract

The electric power steering (EPS) has been used for smaller size automobiles, and it is considered for larger size ones. It is believed that the EPS will replace the hydrauric power steering soon or later. But the EPS control systems has not been studied enough to afford control theories. A new design method of EPS control systems using the scheduled H_∞ control is proposed in this paper. Firstly, an analytical model of the EPS is derived. Secondly, the EPS control systems are constructed with a loop shaping technique of closed loop transfer functions based on H_∞ criterion. Thirdly, as the system changes according as driving conditions, the scheduled H_∞ control is used to design the EPS control systems. Finally, the system is shown useful by simulations.

View full abstract

In radar technology, subsurface radar has found many important applications to identify various target objects, such as, cables for optical fiber communications, gas and water pipes. But due to dispersion and randomness of the underground media, it is difficult to presume an exact location and shape of the buried objects. And there is only a marginal improvement of detection characteristics reported in the literature and should be improved further.
In this study, we carried out experiments on subsurface radar with scale model which facilitate comparing real scale model with computer simulations. We used the impulse radar as subsurface radar. In real scale, we considered to use 1.5ns impulse and decided the size of scale for the present experiment. A scale size of one-third and 0.5ns impulse as transmitted wave have been used.
Transmitting and receiving antennas have been designed and the antennas are of Bow-Tie type which are commonly known as wide-band antennas. We measured the radiation pattern and frequency characteristics of these antennas and confirmed the applicability of use these antennas as subsurface radar antenna. We also measured the characteristics of soil. We used dry sand as soil and propose an easy way to measure the characteristics of the soil using a monopole-antenna. Various scattered waves using several targets have been investigated.

View full abstract

Detailed characterization of radio propagation channel is a major requirement for successful design of mobile communication systems. In this paper, mobile radio channel characterization process based on the FDTD method is presented. The merits and demerits of the currently used methods, namely impulse-response method and ray-tracing methods are briefly considered, and the total field formulation of the FDTD method is discussed. The simulation model consists of streets with concrete buildings. The wave propagation patterns in the whole channel and the received signals at some line of sight and out of sight locations are presented.

View full abstract

Recently X-rays are the focus of the next generation technologies. The development of X-ray optics, with the advent of the UNDULATOR such as SOR (Synchrotron Orbital Radiation), has been remarkable. X-ray optical devices are forced to step up the efficiency, and lenses, reflection mirrors and guides are being studied in the X-ray active region. In this paper scattering analysis is carried out by using wavelength of X-ray, and parameters such as radius of atoms, distance between atoms, angle of incidence X-ray. The model of atoms are considered to be cylindrical perfect conductor and dielectric, and we considered up to 15×15 cylinders using calculations based on the BEM technique.

View full abstract

This paper presents a stable motor speed control system in low speed region with a low resolution shaft encoder. The proposed control system has the models of the controller and the motor to estimate the rotor speed. The estimation of speed is performed by correcting the estimated position, speed and the integration term of controller in accordance with the position estimation error between the detected and estimated positions. The stability of the proposed control system and the design of the estimation gains are examined.
The proposed algorithm was verified by using a 4-pole, 2.2kW test motor drive system with a 256 pulse/rev encoder. In the experimental results, the stable speed control for the reference speed of 20 rpm was realized. The proposed algorithm was also stable against the step change of the disturbance torque.

View full abstract

The real time checking of the enormous netnews articles, which are distributed through the TCP/IP based computer network (the Internet) systems from not only domestic but overseas areas, is quite significant because the valuable information will be readily acquired. However the conventional netnews systems do not support the urgent notification of the new arrival of the news articles. In the present paper, we shall first illustrate the mechanism of the network news distribution system and propose the epoch-making algorithm of the real time checking program of the new arrival news articles and implements that based on the nntpd daemon process, which knows the arrival of the new news at any time. The developed convenient utilities called as “cnn” (Check New News articles) and “xcnn” (X client for Check New News articles) are freely distributed through the Internet, acquire many users and are archived in various Internet utility CD-ROMs.

View full abstract

This paper proposes FRL (Fundamental Recognition Learning) algorithm. FRL is knowledge acquisition, modification and unification. Knowledge acquisition and modification method are ABLE (Activation Bidirectional propagation LEarning) method using FAMOUS (Fuzzy Associative Memory Organizing Units Systems). Knowledge unification method is the revision of ABLE method. This unification method is to unify some membership functions and weights as knowledge. Experimental show the effectiveness of this method.

View full abstract

In this paper, we propose a novel parallel multiplication scheme using 2×2 submultipliers. It adopts a new encoding method which halves the number of partical products through 2×2 submultipliers and rearrangement of primitive partical products. We present the parallel multiplication algorithm and design a 16bit×16bit multiplier architecture of the proposed method, which is suitable for VLSI implementation. The proposed scheme can be easily extended to large-size multipliers with 32, 54 or even longer bits. This parallel mltiplication scheme can be applied to many areas such as DSP, MPEG and so on.

View full abstract

This paper proposes a new applied method to compress the VCV (Vowel-Consonant-Vowel) speech spectrum pattern for synthesis by rule using a Genetic Algorithm (GA). In this method synthesized speech is generated by using parameters from a subset of all of the analyzed VCV frame data. The subset which can generate high-quality synthesized speech is selected automatically by using an optimum combination technique based on GA. We regarded “individuals” in GA as the random selected subsets and defined “fitness” as the inverse of the perceptually based weighting error which is extracted from the difference between natural speech and synthesized speech. Furthermore by using a “generation-dependent mutation width”, the convergence speed can be greatly increased as compared to a ordinary mutation technique. The results from a listening experiment indicate that the synthetic speech generated from the compressed spectrum pattern is of sufficient quality to be of practical use.

View full abstract

This letter presents a technique to segment the character number region from the self-lighting in the night. From the histograms of R, G and B images, the image of the parts of lights are eliminated and another image derived from the Gauss-Laplacian operation and binarization operation of the G image are processed by Logic operation of AND, then characters are extracted. Our examinations conclude that it is a effective way to segment the character region on number and satisfactory results are obtained.

View full abstract