This paper presents an overview of the SDC-I (Super Database Computer I) developed at the University of Tokyo, Japan. The purpose of the project was to build a high performance SQL server which emphasizes query processing over transaction processing. Recently relational database systems tend to be used for heavy decision support queries, which include many join, aggregation, and order-by operations. At present high-end mainframes are used for these applications requiring several hours in some cases. While the system architecture for high traffic transaction processing systems is well established, that for adhoc query processing has not yet adequately understood. SDC-I proved that a parallel machine could attain significant performance improvements over a coventional sequential machine through the exploitation of the high degree of parallelism present in relational query processing. A unique bucket spreading parallel hash join algorithm is employed in SDC, which makes the system very robust in the presense of data skew and allows SDC to attain almost linear performance scalability. SDC adopts a hybrid parallel architecture, where globally it is a shared nothing architecture, that is, modules are connected through the multistage network, but each module itself is a symmetric multiprocessor system. Although most of the hardware elements use commodity microprocessors for improved performance to cost, only the interconnection network incorporates the special function to support our parallel relational algorithm. Data movement over the memory and the network, rather than computation, is heavy for I/O intensive database processing. A dedicated software system was carefully designed for efficient data movement. The implemented prototype consists of two modules. Its hardware and software organization is described. The performance monitoring tool was developed to visualize the system activities, which showed that SDC-I works very efficiently.

This paper discusses a digital neuroprocessor named Quantizer Neuron Chip (QNC) employing the Quantizer Neuron model and two newly developed schemes; "concurrent processing of quantizer neuron" and "removal of ineffective calculations". QNC simulates neural networks named the Multi-Functional Layered Network (MFLN) with 64 output neurons, 4672 quantizer neurons and two million synaptic weights and can be used for character or image recognition and learning. The processing speed of the chip achieved 1.6 µseconds per output neuron for recognition and 20 million connections updated per second (MCUPS) for learning. In addition, QNC can execute multichip operation for increasing the size of networks. We applied QNC to handwritten numeral recognition and realized high speed recognition and learning. QNC is implemented in a 1.2 µm double metal CMOS with sea of gates' technology and contains 27,000 gates on a 10.9910.93 mm2 chip.

This paper examines the retransmission probability and throughput characteristics of slow-frequency-hopping spread spectrum transmission in Rayleigh fading indoor channels of multi-cell environments. Because signal strength in a Rayleigh fading indoor channel changes slowly, retransmission probability is little influenced by the retransmission unit length and error correction capability when retransmission unit length is shorter than the fading period. With the 83.5MHz bandwidth of the 2.4GHz ISM (Industrial, Scientific, and Medical) band in the USA, quaternary phase shift keying (QPSK) is expected to provide a throughput of nearly 1Mbps in each cell when we assume a shadowing margin of 18.1dB and we use 4 cells, 100 hopping frequencies, a transmission power of 600mW, and a transmission distance of 20m. And also with the 26MHz bandwidth of the 2.4GHz ISM band in Japan, QPSK is expected to provide a throughput of nearly 1Mbps in each cell under similar conditions but with 26 hopping frequencies and a transmission power of 260mW.

In this paper, the average error-rate characteristics are investigated as the number of users increases in the digital cellular DS/CDMA (Direct Sequence/Code Division Multiple Access) systems. Then, the performances of the various error control schemes applied to the data service of digital cellular DS/CDMA systems are compared and analyzed. That is, the performances of the conventional error control schemes such as Go-back-N ARQ (Automatic Repeat Request) and Selective-Repeat ARQ are analyzed in the circumstance of digital cellular DS/CDMA system. Also, the improved error control schemes which utilize the variable window size and/or variable data packet size are proposed and evaluated in order to improve the performances of the conventional error control scheme such as Quick-Repeat ARQ and WORM ARQ schemes in the digital cellular DS/CDMA system environments. According to the simulation results, the performances of the improved scheme with variable window and variable frame size are superior to those of the conventional scheme in the view of throughput and delay characteristics due to the robustness to the fading channel impairments.

The issue about document structure recognition and document understanding is today one of interesting subjects from a viewpoint of practical applications. The research objective is to extract the meaningful data from document images interpretatively and also classify them as the predefined item data automatically. In comparison with the traditional image-processing-based approaches, the knowledge-based approaches, which make use of various knowledge in order to interpret structural/constructive features of documents, have been currently investigated as more flexible and applicable methods. In this paper, we propose a totally integrated paradigm for understanding table-form documents from a viewpoint of the architectural framework.

A new design method for 2-D IIR digital filters, having a separable denominator,in the spatial domain is presented. The modified Gauss Method is applied in the iterating calculation of the filter coefficients. Also, the 1-D state space representation of the denominator is utilized in determining the impulse response of the designed IIR transfer function and its partial derivatives systematically while the numerator is expressed by a nonseparable polynomial. The error criterion function, which also includes the response outside the given region of support, is minimized in the least square sense. Convergence, together with the stability of the resulting filttr, are guaranteed.

An efficient algorithm is presented for solving nonlinear resistive networks. In this algorithm, the techniques of the piecewise-linear homotopy method are introduced to the Katzenelson algorithm, which is known to be globally convergent for a broad class of piecewise-linear resistive networks. The proposed algorithm has the following advantages over the original Katzenelson algorithm. First, it can be applied directly to nonlinear (not piecewise-linear) network equations. Secondly, it can find the accurate solutions of the nonlinear network equations with quadratic convergence. Therefore, accurate solutions can be computed efficiently without the piecewise-linear modeling process. The proposed algorithm is practically more advantageous than the piecewise-linear homotopy method because it is based on the Katzenelson algorithm that is very popular in circuit simulation and has been implemented on several circuit simulators.

This paper discusses the role of knowledge in document image understanding from the viewpoints of representation, utilization and acquisition. For the representation of knowledge, we propose two models, a layout model and a content model, which represent knowledge about the layout structure and content of a document, respectively. For the utilization of knowledge, we implement layout analysis and content analysis which utilize a layout model and a content model, respectively. The strategy of hypothesis generation and verification is introduced in order to integrate these two kinds of analysis. For the acquisition of knowledge, we propose a method of incremental acquisition of a layout model from a stream of example documents. From the experimental results of document image understanding and knowledge acquisition using 50 samples of visiting cards, we verified the effectiveness of the proposed method.

In this letter, we compare the three compound methods of the Multiple-scale technique to improve the quality of the scale values estimated by the method of fuzzy categories. The results show that the maximum compound method brings higher ability to estimate the scale values than the other methods despite categories used in the scale.

A new frequency hopped spread spectrum system is introduced. The frequency hopped signal is a combination of multi frequency and multi phase signals and is referred to as Frequency Hopped/Joint Frequency-Phase Modulation (FH/JFPM). A noncoherent receiver for the FH/JFPM signals is introduced and an exact expression for the bit error rate is obtained. A performance analysis of this system is given in the presence of broadband and partial-band noise jamming. The optimal jamming strategy is evaluated. The results show that under these jamming conditions the FH/JFPM perform better than the FH/M-ary DPSK and FH/M-ary FSK systems. It is also shown that for a given channel bandwidth and data rate, the FH/JFPM system has more processing gain than its FSK or DPSK counterparts.

This paper proposes a group demodulator that employs multi-symbol chirp Fourier transform to demodulate pulse shaped and time asynchronous signals without degradation; this is not possible with conventional group demodulators based on chirp Fourier transform. Computer simulation results show that the bit error rate degradation of the proposed group demodulator at BER=10-3 is less than 0.3dB even when a root Nyquist (α=0.5) filter is used as the transmission pulse shaping filter and the symbol timing offset between the desired channel and the chirp sweep is half the symbol period.

This paper proposes a new fully-digitalized π/4-shift QPSK modulator consisting of a digital pulse shaping filter and a baseband quadrature modulator. By employing a novel digital filter configuration, the required filter memory is reduced to just 6.25% of the conventional one. Moreover, since the proposed baseband modulation scheme does not employ analog mixers or an analog 90 divider, a very accurate, high-stable and compact modulator is realized. It is shown that the proposed scheme achieves excellent low power consumption characteristics and is more suitable for digital LSIC implementation of personal communication terminals than a direct RF modulation scheme and an analog IF modulation scheme.

This paper describes a nondestructive measurement method for complex permittivity of dielectric material at pseudo microwave frequencies. The resonator used in this study has a cylindrical cavity filled with a sapphire material of a well known complex permittivity. The resonator is divided into two parts at the center. A dielectric substrate specimen is clamped with these halves. Relative permittivity εand loss tangent tan δ of the specimen are obtained at 3 GHz using the TE011 resonance mode. The accuracy of the present method is evaluated through the comparison of the measured values by the new method with those at around 10 GHz by the conventional empty cavity resonator method. The errors of measurements are smaller than 1% and 1105 for εand tan δ, respectively.

The dielectric resonator method is now widely accepted as a precise measurement method for determining the dielectric properties at microwave frequencies. This paper describes the measurement results of εr, tan δ and TCf determined by a round robin test of this method. The resultant measurement errors were Δεr/εr0.10%, Δtan δ0.5105 and ΔTCf0.5 ppm/K, where Δdenotes a standard deviation. The causes of measurement errors and the conditions to improve the measurement accuracy are discussed.

An improvement of Q evaluation is discussed. The Resonance Curve Area method was confirmed to give a deviation in the order of 6104. The result was three times more accurate than the widely known Q evaluating method which utilizes the cursor function installed in a network analyzer. A discussion is also made on the physical validity of the RCA method. It is shown that the application of the RCA method improves the accuracy of the cavity perturbation method. Actual measurements have shown that the deviation of dielectric constant is less than 1% and that of the loss tangent is less than 3%, in the order of 104. The accuracy of the RCA method was estimated to be three times that of the conventional cavity perturbation technique. The consistency of the perturbation with other methods has also confirmed. The accuracy comparison to more accurate formulae derived from a rigorous solution have shown that the difference is sufficiently small.

A simple technique is proposed for improving the convergence of Newton's method in the spherical algorithms, which are metheods for tracing solution curves. A numerical example is given in order to show the effectiveness of the proposed technique.

An estimation method for efficiently calculating the field intensity in the Fresnel region of broadside colinear array antennas is developed, and its performance is experimentally verified. The calculation utilizes only the antenna design data, and is readily applicable to arbitrary array antennas. This method can provide a safety protection zone in the proximity of array antennas, in order to protect radio communication personnel and general public from the potentially hazardous radiofrequency exposure.

The plane wave diffraction by a two-dimensional parallel-plate waveguide cavity with partial material loading is rigorously analyzed for both the E and the H polarization using the Wiener-Hopf technique. Introducing the Fourier transform for the scattered field and applying boundary conditions in the transform domain, the problem is formulated in terms of the simultaneous Wiener-Hopf equations satisfied by the unknown spectral functions. The Wiener-Hopf equations are solved exactly via the factorization and decomposition procedure leading to the formal solution, which involves branch-cut integrals with unknown integrands as well as infinite series with unknown coefficients. Applying rigorous asymptotics with the aid of the edge condition, the approximate solution to the Wiener-Hopf equations is derived in the form suitable for numerical computations. The scattered field inside and outside the cavity is evaluated by taking the inverse Fourier transform together with the use of the saddle point method. Numerical examples of the radar cross section are presented for various physical parameters, and the far field backscattering characteristics of the cavity are discussed in detail. Some comparisons with a high-frequency technique are also given to validate the present method.

A new built-in drive circuit for superconducting Josephson LSI circuits has been designed and fabricated in a ceramic multichip module. The drive circuit consists of an impedance matching circuit and a DC bias current feeding circuit to supply a two-phase power current to Josephson chips at a microwave frequency. The impedance matching circuit was designed based on a quarter wavelength stripline. A balanced stripline configuration was introduced to reduce the fluctuation of ground potential. Tungsten layers were used to make the drive circuit in a multilayer ceramic substrate of the multichip module. Whole circuit was successfully packed in a volume of 76 mm38 mm1.7 mm. The gain of microwave current were 20 dB around 1.2 GHz and 23 dB around 3.6 GHz, which were in good agreement with the simulated current gain.

ADTs (Abstract Data Types) have been known as a promising feature for extending the database applications to CAD/CAM and other engineering areas. This extension has brought a new dimension to query optimization. Conventional query optimization methods, which considers only joins as the dominant cost factor, are based on the belief that the executions of selections and projections basically take no time. However, in databases that support ADTs, this may not be true since the execution of a selection involving ADT functions may be very time-cosuming. Thus selections with ADT functions should not be considered as inexpensive operations in queries, and the conventional optimization heuristics should be enhanced to correspond to the appearance of the queries of this kind. In this paper, we show the possibility that semijoins can be used as an effective means to reduce the number of evaluations of an ADT function and consequently optimize queries containing expensive ADT selections. We suggest the enhancement of an conventional optimization heuristics by adding a semijoins pre-stage which is an additional component corresponding to expensive ADT selections. By this way, the applicable range of the conventional heuristics are extended to hold the ability of handling queries with ADT functions. Several optimization algorithms are given and some simulation results show the effectiveness of our methods.