Search

This work describes the authors' approach at their university in the last three years for computer science second grade undergraduate students to experience parallel and distributed computing. The goal is to give a solid understanding of parallel and distributed processing technologies and to build up basic skills in the field, such as parallel algorithms, multi-thread/network programming, IP/socket communication, MVC paradigm, RPC/remote method invocation (RMI), Database/SQL, and Java/JDBC. The course features a combination of active experimental learning and N to N networking approach. Unlike typical laboratories where central parallel servers or parallel machines are used (N users to one system networking), our laboratories do without them and instead organize groups of student PCs to form virtual parallel/distributed systems (N users to N systems networking). All PCs work as servers as well as clients. Parallel bucket sorting and virtual shopping mall implementations are employed for the course projects. The course consists of 14 ninety-minutes sessions within a semester, including introductory Java network programming and two projects. As the time is limited, homework and pre-laboratory experiments are encouraged. Web based course material distribution and the virtual laboratory environment contributed to student success.

A 4-Gb/s track and hold (T/H) circuit with a parasitic capacitance canceller is proposed. The parasitic capacitance canceller is connected in parallel with the load capacitance of the T/H circuit and acts as a negative capacitance. The proposed T/H circuit can reduce by 26 % its chip area and by 37 % its power dissipation compared with those of a conventional one, since the cancellation circuit equivalently reduces the load capacitance of the T/H circuit. The proposed T/H circuit is applied to a 4-Gb/s 5-bit flash ADC, fabricated in a 90 nm CMOS process. Thanks to the cancellation circuit, there is not only a reduction of its power consumption but also an extension of its bandwidth. In particular, the bandwidth is extended up to 2 GHz. The measurement results show that the signal to noise and distortion ratio (SINAD) of the ADC, at 2 GHz, is improved to about 27 dB.

Novel ultra-low-k porous silica films were developed by use of a self-assembly technology. The mechanical properties of the porous silica films could be reinforced independently of the dielectric constant by introducing a tetramethyl-cyclo-tetra-siloxane (TMCTS) treatment. High modulus porous silica films, with an elastic modulus of 8 GPa and dielectric constant of 2, can be achieved simultaneously. Ultra-low-k/Cu damascene with sufficient mechanical strength was demonstrated for 45 nm BEOL (back-end-of-line) technology.

In this paper, to analyze the mechanism of the communication between a player and a listener, the relation between the time course of the acoustic level and the time course of the physiological level of a live performance were analyzed at the same time. The results revealed that, where scatter of notes was low, music rhythm of a player and handclap rhythm of a listener were likely to synchronize, and where scatter of notes was high, both respiration rhythms were likely synchronize. Then where scatter of notes was high, the coupling between acoustic level and physiological level became strong. To interpret these results, we hypothesized that where scatter of notes is high, a player and a listener have to anticipate the timing of sounding, and proposed a new communication model between a player and a listener.

In this research, to analyze "a feeling of unity" at a concert hall, we investigated the relationship among listener's respiration and handclap and the music. As a result, it was shown that listener's respiration and handclap preceded to musical note. This result suggests that the predictive actions have influence to listener's subjective mental condition.

A simple and accurate method to check the performance of an 8.5-digit digital voltmeter with high accuracy (/spl sim/1 pm) is proposed. It is demonstrated that this method is useful to check if such a digital voltmeter is working within the specification given in the catalogue or manual.

This paper proposes a tail current source dividing technique for increasing precision of a track-and-hold circuit without reducing the sample speed. The proposed technique is based on division of a tail current source into two current sources during a hold period to reduce signal errors caused by parasitic capacitance coupling between the input and output terminals. With the proposed technique, a SFDR of 62 dB for 25 MHz, 1 V/sub p-p/ input signal at 200 MS/s is achieved with 0.35 /spl mu/m CMOS parameters by HSPICE simulation.

A circuit named switched LNA which plays both the role of a low noise amplifier and a sampling circuit is proposed. Simulation results demonstrate an improvement of 20 dB on signal-to-spurious ratio. The proposed circuit occupies a smaller area and consumes less power than the conventional circuit.

In a mountainous and highly populated country like Japan, corner reflectors frequently must be placed near confounding objects. A procedure to detect singular corner reflectors in such situations is described. It is found that the initial reduction of speckle by the Enhanced SFP filter greatly facilitated the detection of these corner reflectors from both JERS-1 SAR images and ERS-2 images.

We have carried out JERS-1 differential interferometry (DinSAR) research in volcano and earthquake monitoring as 'The Earthquake Remote Sensing Frontier Project' since 1997. The potential of L-band SAR interferometry to detect subtle crustal movements even in mountainous regions in Japan and its geodetic accuracy were evaluated by comparison and combination with other geodetic and seismological observations (e.g. GPS network, leveling survey, precise gravity measurement, strong motion seismology). So far we have succeeded in several cases; (1) lateral fault motion by the Kagoshima earthquake (M=6.3) in 1997, (2) volcano inflation and the following earthquake (M=6.1) around Mt. Iwate in 1998, (3) tensile deformation associated with earthquake swarms around the monogenetic volcano area in the Izu Peninsula in 1997 and 1998.

Use of the Internet is growing explosively, not only for text-based applications such as e-mail and bulletin boards, but also for multimedia WWW applications involving video and audio, and multicast services. Recently, satellites have been used in communication networks as a means of avoiding terrestrial Internet traffic jams. Satellites are a very efficient means of delivering large volumes of content to multiple clients simultaneously. We show that fast throughput using TCP/IP connection is feasible in networks comprising both satellite and terrestrial links. Then we propose a UDLR (Uni-Directional Link Routing) mechanism for networks that include uni-directional links (such as satellite links), and show that general dynamic routing algorithms such as RIP and OSPF can be adopted even when using mi-directional links.

The authors discuss a functionally distributed processing system architecture based on a local area network (LAN) communication environment. They also describe the network resource database configuration essential to the end customer control management system. An implementation of an end customer control management system consisting of small-sized processing systems and the communication environment which interconnects these systems in a highly reliable and easily expandable manner are presented. The implementation is expected to facilitate the various management functions necessary for customer control. >

The authors investigate potential asynchronous transfer mode virtual path operation systems (ATM VP OpS) with a special focus on the leased-line services that will encourage the smooth penetration of ATM networks. They describe ATM VP OpS requirements after clarifying leased-line operation scenarios. An object-oriented modeling method is presented for operation system implementation. A highly effective ATM VP OpS system architecture is outlined. >

Japanese quail inoculated subcutaneously with Borrelia garinii responded to infection. B. garinii was reisolated mainly from skin and randomly from several organs between 7 and 56 days postinoculation. Skin lesions were occasionally observed in association with spirochete recovery. All birds were positive for antibody at 53 to 56 days postinoculation. These results suggest that the Japanese quail is a model for experimental infection with B. garinii and that birds are reservoirs for Lyme disease spirochetes.

Elevated temperature, normally a characteristic of inflammation, is a potential indicator of periodontal disease. The present investigation examined the relationship of subgingival temperature and various indicators of gingivitis in rats with spontaneous gingivitis (SUS rat) and in control rats (RES rat). In 24 SUS rats with gingivitis, the subgingival temperature rose, but not in 6 SUS and 20 RES rats with healthy gingiva. Subjects with high mean subgingival temperatures had significantly higher mean percentages of Prevotella intermedia and Fusobacterium nucleatum than age-matched RES rats. The results show that dental plaque, including specific bacteria P. intermedia and F. nucleatum, can play some role in causing the elevated temperatures. J Periodontol 1994;65:710–712.

The authors found that the inactive Zn in GaInP/(Al/sub 0.7/Ga/sub 0.3/)InP quantum well structures peculiarly degraded the carrier life time which strongly affected the laser's performance and reliability. The novel cap layer of n-(Al/sub 0.7/Ga/sub 0.3/)InP realized almost 100% activation of Zn in the p-(Al/sub 0.7/Ga/sub 0.3/)InP layer. The same temperature dependence on the threshold current density of visible lasers was obtained by introducing the novel n-cap layer. The kink level of 160 mW and the maximum light output power of 204 mW at 25/spl deg/C were realized by introducing the novel cap layer of the n-(Al/sub 0.7/Ga/sub 0.3/)InP for the window structured visible laser. This visible laser could achieve the finest reliability under stable continuous wave operation over 2600 hrs under 50 mW at 50/spl deg/C. >

The paper proposes a distributed operation system (OpS) model for the management of transmission networks. The model is based on the telecommunication management network (TMN), and uses directory service to settle some issues of the distributed system. The transmission networks need distributed OpSs that handle the huge number of network elements (NEs). An OpS manages managed objects that represent network resources. Identification of managed objects, NEs and OpSs is indispensable to establish management associations between them in the distributed environment. The proposed model uses the directory service to obtain those identifications. In order to effectively realize the directory function, the distributed directory realization method is proposed and evaluated. The simulation result confirmed the benefit of the distributed directory in reducing management traffic to obtain information. >

This paper presents a method to analyze and minimize finite wordlength errors in output signals of linear phase FIR filters. In the frequency domain, impulse responses corresponding to all levels of input impulses are optimized. If the probability densities of input signals are taken into account, the errors in the output signals of the filters are smaller than those in the conventional filters. For image signals, which have only positive values, the non-symmetry of their probability densities makes the proposed algorithm more effective. Many results of simulations confirm that the output signals of the proposed filters have far less errors than those of conventional filters.

Sufficient conditions of optimality in shape optimization problems are studied in the present paper. Shape optimization problems are the problems in which an objective function depends on the domain through the solution of the boundary-value problem defined on the domain. There are many such problems in various branches of science and high-technology industries whenever the best shape for a physical system is required. The method of deriving sufficient conditions suggested here is based on taking into account the symmetry that shape optimization problems have. We mean the objective function depends an the shape of the domain and does not depend on the place of the domain in the plane. That is why extremum in such problems is not proper and it makes obtaining of the effective sufficient conditions nontrivial. The main idea of our method is special projecting of the perturbated domain on the set of the domains having optimal shape. We prove here that this method leads to effective sufficient conditions for the problems having symmetry. In order to show the ideas of the method more clear we consider one of the classical problems as an illustrative example and establish the optimality in this problem proving all up to final point. Throughout the paper we are discussing how to use the general ideas of this method for other shape optimization problems.

Sufficient conditions of optimality in shape optimization problems are studied in present paper. The method of deriving sufficient conditions suggested here is based on taking into account the symmetry that shape optimization problems have. The main idea of our method is special projecting of the perturbed domain on the set of the domains having optimal shape. We prove here that this method leads to effective sufficient conditions for the problems having symmetry.

Operation scenarios running on a transport network operation system (OpS) need to be rapidly modified and changed in order to effectively support the deployment of new telecommunication services. This paper proposes an operation scenario program development method based on late binding technology. The method includes operation object definitions and stored operation object binding. Operation objects model the common management information service (CMIS) used in TMN-based network operation systems. An experiment confirms the method's effectiveness. The late binding technology enables us to dynamically bind operation objects and verify the performance of the designed scenarios.

The ATM transport network operation system described follows the TMN concept and supports CMIS/P interfaces to connect its subsystems. It contains about a hundred GDMO object classes, which were specially designed to achieve the operation system function. New technologies introduced to ensure effective and rapid system implementation include a directory system, to solve the problem of object instance location, and the use of GDMO object classes to design all system functions. In addition, an original common software platform is used to support the GDMO object intercommunication function. This operation system achieves a distributed object processing environment that supports CMIS/P interfaces. The system has been deployed in Japan and is now being used in multimedia experiments, the initial results of which are encouraging.

A telecommunications management network (TMN) based operation systems (OpS) development platform, including TNMS kernel (transport network management systems kernel) and OpS development support tools, is introduced. The platform adapts the object oriented design approach and the GDMO template based program development. In order to confirm the platform's capability, an ATM transport network OpS is designed and experimentally implemented. The experimental OpS uses a layered architecture of OpS subsystems to manage ATM virtual paths. For the evaluation measures of system development efficiency and operation response time, the platform and the experimental OpS are shown to be suitable for developing future OpSs.

A TMN based operations system development platform that addresses the problems pertaining to recursive operations system improvement is proposed. The platform consists of the object class library named TNMSkernel (transport network management systems kernel) and associated support tools. TNMSkernel provides the application program interfaces needed for management information manipulation while the support tools provide application program generation capability from GDMO object class definitions and extensions of behavior description. The proposed platform is implemented and used to develop an ATM based transport network experimental operations system. The results confirm the platform's efficiency in terms of reusability and productivity of system production. >

A CMOS operational transconductance amplifier and capacitor (OTA=C) integrator which is used for high-frequency operation has been designed and simulated by the SPICE 2G program. The authors have realized the continuous-time OTA=C elliptic low-pass filters by a cascade method and a signal flow graph (SFG) method using only capacitors and OTAs. The frequency characteristics of the OTA=C filters are compared. >

This paper describes current mirrors which consists of n-channel depletion-mode GaAs MESFETs. The newly proposed current mirrors have both inverting and non-inverting output terminals that enable us to realize current-mode circuits very simply. Current mode filters can be realized by using current-mode integrators which are obtained by adding a capacitor at the input port of the current mirror. As an example of filters, we show a leapfrog structure of a third-order low-pass filter. HSPICE simulation was performed to verify the validity of the proposed current-mode filter.

High frequency (higher than 100 MHz) and low power supply voltage (lower than 2.0 V) bipolar analog integrated filters are presented. The filters are based on a current mode signal processing architecture. 6th order low-pass and 6th order band-pass filters are simulated by SPICE.

A new CMOS companding current-mode integrator is proposed. The companding integrator is based on the MOS translinear principle and utilizing the nature of the MOSFET square-law. SPICE simulation results demonstrate good performances.

One of the important tasks in designing analog circuits is to select the most appropriate circuit topology from available knowledge base. The purpose of this paper is to develop a new tool for electronic circuit topology selection based on a neural network. The main advantages of this approach are: the system requires less designer's experience and it is capable of automatic incorporation of a new circuit topology into the system with little modification. Some circuit topology selection examples are shown.

This paper has proposed on-chip integration of an automatic tuning system for continuous-time filters using a switched-capacitor circuit and down-sampling techniques. The proposed system requires no external components and it is free of DC amplifier input-offset voltage and harmonics produced by the phase comparator. Its performance has been confirmed by computer simulations. An important property of an automatic tuning system is to keep its control signal appropriate. After the control signal converges at a desired value, operation of the system can be suspended. This will avoid noise produced by clock signals and save on power consumption. Therefore it is very effective to use the proposed system together with non-volatile analog memories.

Shape optimization problems with the functional depending on the gradient and Dirichlet problem for the Poisson equation as a constraint are studied. The problem statement reflects the situation in practical shape optimization where the symmetry is often present and the optimum is not proper. Sufficient conditions for a disk to be a solution of the optimization problems with highest symmetry are given along with necessary conditions.

This paper presents a method to minimize the finite wordlength error in output signals of linear phase 2-D FIR filters. The finite wordlength errors can be easily analyzed in the frequency domain when the input signal statistics are known. In the case of white input signals, impulse responses corresponding to all levels of input impulses are optimized so as to minimize the errors. A new ROM based filter structure is proposed in which the optimized impulse responses are stored in the ROM. The output signals are generated by superposing the impulse responses corresponding to the input levels. Many results of simulations confirm that the output signals of the proposed filters have far less errors than those of conventional filters.

GaAs MESFET is expected to be applied for high frequency analog integrated circuits. However, as n-channel FET is only now practically in use, the development of GaAs MESFET analog integrated circuits has been left behind. In this paper, novel low power GaAs MESFET current-mode lossy and lossless integrators using common-gate structure are proposed. The proposed circuits are very simple. A bandpass filter is composed as an application example. HSPICE simulation results show good performances.

A current-driver write driver for a +5/-5 V preamplifier is described in this paper. This IC, which incorporates RC load for harmonic oscillation to enlarge the voltage swing across the head and Super Push-Pull Logic (SPL), drives the write driver with a good rise/fall time. This write driver, built with a 0.35 /spl mu/m SOI-BiCMOS process, has demonstrated a rise-time as short as 0.45 ns with a 160 mA peak-to-peak output write current.

Class AB differential amplifiers are proposed. Class AB operation for differential amplifiers is realized by using a minimum voltage selecting circuit and an amplifier. An almost minimum tail current value is provided to a differential amplifier to reduce extra power dissipation. The simulation results show that the static power consumption of Wang's OTA [1990] using the proposed method can be reduced from 244 /spl mu/W to 46 /spl mu/W.

An operational transconductance amplifier (OTA) is one of the most important analog building blocks. This paper proposes a novel voltage-regulating circuit for an OTA realization. The voltage-regulating circuit allows one to implement a simple structure OTA using MOSFET's in the non-saturation region. The resultant OTA using the proposed voltage-regulating circuit operates under a single 1.5 V power supply voltage. Simulation results of the OTA show harmonic distortion less than 1% even when 0.7 V/sub p-p/ input voltage is applied. Its power consumption is 85.8 /spl mu/W and cutoff frequency is 7 MHz.

To detect antibodies to recombinant vacuolating cytotoxin (r-VacA) of Helicobacter pylori in cerebrospinal fluid (CSF) from patients with Guillain-Barre syndrome (GBS).CSF samples from 13 patients with GBS (electrophysiologically classified as eight acute inflammatory demyelinating polyradiculoneuropathy (AIDP), four acute motor axonal neuropathy (AMAN), and one unexcitable nerve conduction) and eight disease control patients were studied. The r-VacA protein was separated by SDS/PAGE, and Western blot analysis was carried out.Six of the 13 patients with GBS had a specific IgG antibody to VacA of H pylori, which was confirmed by absorption experiments using r-VacA. Every patient with positive CSF anti-r-VacA IgG had AIDP.The sequence homology previously found between VacA and human (Na(+)+K(+))-ATPase A subunit suggests that antibodies to VacA involve ion channels in abaxonal Schwann cell plasmalemma resulting in demyelination in some patients with GBS.

In Japan, the incidence of tuberculosis (TB), which had improved continuously from 1951, has shown an upturn since 1997.An Ad-Hoc National Survey was conducted to review the issues of TB control in Japan and to discuss the future direction of change.TB registries kept at health centres in all prefectures were reviewed for the year 1998. Vital statistics and BCG vaccination records were reviewed separately.The survey revealed a 34.7% under-estimation of TB deaths in TB registers, and a 11.2% over-estimation of TB cases registered due to a change in diagnosis after registration. For treatment, the four-drug regimen advocated for smear-positive cases was used in only 54.8% of indicated cases, and treatment duration was very long; for example, 18.4% of first-time hospitalised cases over 15 years of age received in-patient care for more than 6 months and 29.8% continued treatment for more than 12 months. Apart from an increasingly elderly population, other factors contributing to an increase in TB included underlying health risk factors for increased susceptibility to TB and socio-economic risk factors for treatment failure. Regarding preventive measures, active case finding by health screening contributed to detection of only 8.7% of new cases aged over 60 years, but BCG vaccination appeared to be preventive among infants under 1 year of age.The National Survey revealed a changing profile of TB in Japan and suggested a need to review and revitalise the TB control programme.

We report a 26-year-old woman who showed recurrent migrainous attacks and convulsions since her childhood. Neurological examination revealed no focal abnormality except mental retardation (MR). T2-and fluid-attenuated inversion-recovery (FLAIR)-weighted brain MRI revealed apparent high intensities in the deep subcotical white matter. Ultrastructual studies revealed an abnormal deposition of granular osmiophilic materials (GOM) on the surface of vascular smooth muscle cells in dermis. Her mother developed recurrent strokes without risk factor since age 41. A heterozygosis Arg133Cys mutation of Notch 3 gene has already presented in patient and her mother. This case might be an early stage in CADASIL before stroke onset and suggested that systemic vasculopathy was presented in this stage. The correlation between MR and phenotype of CADASIL were unclear.