This paper deals with the improvement of performance in the transform and subband image coding systems with negatively-correlated input signal. Using a more general source model than the AR(1) model as an input, the coding performance for the transform and subband coding schemes is evaluated in terms of the coding gain over PCM. The source model used here has such resonant band characteristics that its power spectrum has a peak at some frequency between 0 and π/2 for positive autocorrelation and between π/2 and π for negative autocorrelation. It is shown that coding schemes are classified into two classes; one has the pairwise mirror-image property in their filter banks and performs symmetrically regardless of the sign of the autocorrelation, and the other has no that property and performs asymmetrically with inferior performance for negative autocorrelation. Among the well-known transform and subband coding schemes, the DHT and QMF coding systems belong to the former class and the DCT and SSKF coding systems to the latter. In order to remedy the inferior performance, we propose the method in which one modulates the negatively-correlated signal sequences by the alternating sign signal with unity magnitude (-1)n to convert them into positively-correlated sequences. The algorithms are presented for the DCT and SSKF image coding systems with the adaptive signal modulation. In the DCT coding systems, we are particularly concerned with the DCT-based hierarchical progressive coding mode of operation, since the signal modulation works well for that coding mode. The SSKF image coding system has the regular quad-tree structure with three stages. The simulation results for test images show that our method can successfully be applied to the images with a considerable amount of energy in the frequency range higher than π/2 in horizontal or vertical direction, such as fingerprints and textile patterns sampled at a rate close to the Nyquist rate. The paper closes with a brief introduction to the modification of our DCT-based method.

This paper proposes and investigates a tap selectable Viterbi equalizer for mobile radio communications. When the multipath channel is modeled by a tapped delay line only, the taps which may seriously affect the data sequence estimation are selected and used to calculate the trellis metric in the Viterbi algorithm. The proposed equalization algorithm can reduce the number of path metric calculations and the number of path selections in the Viterbi algorithm. Moreover, we propose an extended equalizer which has antenna diversity. This equalizer calculates the path metric using the antenna outputs and results of channel estimators. Computer simulation is used to evaluate the BER performance of the proposed equalizer in a multipath radio channel.

Genetic Algorithm (GA) is the method of approaching optimization problem by modeling and simulating the biological evolution. As the genetic algorithm is rather time consuming, the use of a parallel genetic algorithm can be advantage. This paper describes new methods for fine-grained parallel genetic algorithm using a multiprocessor system FIN. FIN has a VLSI-oriented interconnection network, and is constructed from a viewpoint of fractal geometry so that self-similarity is considered in its configuration. The performance of the proposed methods on the Traveling Salesman Problem (TSP), which is an NP-hard problem in the field of combinatorial optimization, is compared to that of the simple genetic algorithm and the traditional fine-grained parallel genetic algorithm. The results indicate that the proposed methods yield improvement to find better solutions of the TSP.

Simplification of the Viterbi algorithm and the error rate performance are presented for a partial response maximum-likelihood (PRML) system employing the PR(1, 1) system as a PR system for (1, 7) run-length limited (RLL) code. The minimum run-length of 1's or O's in the output sequence of the precoder for (1, 7) RLL code is limited to 2. Two kinds of simplified Viterbi algorithms using this run-length constraint are proposed. One algorithm requires the path memory length of only two in the Viterbi detector. The Viterbi detector based on the other algorithm is equivalent to the simple PR(1, 1) system followed by a threshold detector. The bit-error rates of PRML systems with Viterbi detectors based on these algorithms are obtained by computer simulation and their performance is compared with that of conventional PRML systems for (1, 7) RLL code. It is shown that the proposed PRML system exhibits better performance than conventional PRML systems at high recording density.

Recently, some speech recognition methods using fusion of visual and auditory information have been researched. In this paper, a study on the mouth shape image suitable for fusion of visual and auditory information has been described. Features of mouth shape which are extracted from gray level image and binary image are adopted, and speech recognition using linear combination method has been performed. From results of speech recognition, the studies on the mouth shape features which are effective in fusion of visual and auditory information have been performed. And the effectiveness of using two kinds of mouth shape features also has been confirmed.

Tilt margins for disk and lens for a magneto-optical (MO) head were studied for designing a disk system for use with objective lenses having numerical apertures (NA) of 0.55, 0.60, and 0.65. The tilt margins were examined to determine the aberration characteristics of objective lenses and bit error rate (BER) by recording and reproducing signal. In preparing the optical head for testing disk and lens tilt margins, the aberrations were measured by image processing from the CCD area sensor for the spot image of the focused beam, and BER dependencies on the tilting of lens and disk were obtained at the velocity of the outer diameter of the MO disk at the bit rate of 80-Mbps (1, 7 code modulation) recording. According to the aberration and BER characteristics, the limitation for effective wavefront aberration would be 0.05λ rms, the tilt margins corresponded to BER limitation at the level of 3*10-5. The disk margins for NA=0.55, 0.6, and 0.65 were 0.4, 0.2, and 0.1 degrees. The lens tilt margins for NA=0.55, 0.6, and 0.65 were 0.2, 0.1, and 0.05 degrees.

The technique for estimating the parameters of multiple waves provides a convenient tool for analysis of multiple wave-fields and eventually for actual applications to mobile communications. Several algorithms have been proposed for those purposes. However, the best tactics to resolve multiple wave-fields are still imperfectly understood at present. This paper proposes a new method for estimating the angles and power levels of arrival waves based on the extended Kalman filter. A space-variable model which we call a spatial state equation is derived using array element locations and incident angles. It has been shown that by means of the model, the estimation of incident waves can be transformed into the problem of parameter identification in linear system which can be carried out by the extended Kalman filter conveniently. The algorithm is initiated directly by the signal received at each array element. The detailed procedure of an extended Kalman filter approach is given in the paper. The performance of the proposed approach is examined by a simulation study with two signals model. The simulation results show a good estimate performance, even in the case that two waves arrive from close directions.

A two-dimensional recording method that achieves double recording density by reducing the track pitch is described. This method uses a flat disk and the data are recorded with circular marks on lattice points. Two-dimensional interference consisting of crosstalk and inter-symbol interference is reduced by two-dimensional equalization. To minimize the two-dimensional interference, the optimum equalization coefficients are calculated dynamically with the reproduced signal of the training marks. Reproduction was simulated and this showed that the signal-to-noise ratio of the processed signal was 24.3 dB under ideal conditions and 19.8 dB under worst-case conditions with the usual magneto-optical media using double recording density. These simulation results were checked by a recording/reproduction experiment. The experimental result for the signal-to-noise ratio of the processed signal was 23.6 dB with an areal density of 2.3 Gbit/in2.

Tremendous progress has been made in magnetic data storage by applying theoretical considerations to technologies accumulated empirically through a great deal of research and development. In Japan, the recording demagnetization phenomenon was eagerly analyzed by many researchers because it was a serious problem in analogue signal recording such as video tape recording using a relatively thick magnetic recording medium. Consequently, perpendicular magnetic recording was proposed as a method for extremely high-bit-density recording. This paper describes the theoretical background which has resulted in the idea of perpendicular magnetic recording. Furthermore, the possibility of magnetic recording is discussed on the basis of the results obtained theoretically by magnetic recording simulators. Magnetic storage has the potential for extremely high-bit-density recording exceeding 1 Tb/cm2. We propose the idea of 'spinic data storage' in which binary digital data could be stored into each ferromagnetic single-domain columnar particle when the perpendicular magnetizing method is used.

The interstitial hyperthermia is an invasive heating method applied by inserting the applicator into the human body. We have been studying on coaxial-slot antennas for interstitial microwave hyperthermia. The characteristics of the square antenna array were theoretically examined. Firstly, the basic structure of the antenna, and a simplified analysis model taking account of the effect of the boundary surface were explained. Then analysis was performed by using the moment method. Finally, the calculated results were discussed. The catheter thickness has much effect on the characteristics and must be considered both in designing and in using the antenna. When the array spacing was increased, the effective heating area became larger and more uniform. As the insertion depth was increased, the effective heating area was also enlarged.

Conventional subband ADF's (adaptive digital filters) using filter banks have shown a degradation in performance because of the non-ideal nature of filters. To solve this problem, we propose a new type of subband ADF incorporating two types of analysis filter bank. In this paper, we show that we can design the optimum filter bank which minimizes the LMSE (least mean squared error). In other words, we can design a subband ADF with less MSE than that of conventional subband ADF's.

This letter describes the power absorption of a cylindrical man model placed near a flat reflector exposed to TE microwave. The numerical results show that the absorption is in some cases an order of magnitude or more greater than that of the man model without a reflector.

Theoretical and experimental results are presented for angle-multiplexed and wavelength-multiplexed holographic recording. The recording medium is a cerium doped Sr1-XBaXNb2O6 (SBN) single crystal, and the light sources are a laser diode excited second harmonic generation (SHG) laser and a tunable laser diode. The SBN single crystal has high recording sensitivity, high diffraction efficiency and high temperature stability. The laser diodes miniaturize the holographic recording system. Crosstalk between hologram pages is theoretically calculated by using modified coupled-wave equations, and is also experimentally measured. The experimental results agree well with the theoretical results. Two-dimensional alphabetical character images are recorded using angle- and wavelength-multiplexed holographic methods, and are successfully reconstructed. The theoretical results indicate that several hundred multiplexed holograms can be recorded with little crosstalk using the proposed system. This multiplexed holographic recording technique will enable high-density recording and high data-transfer rates.

I have examined factors for implementing a high-speed, low-power-consumption thermal head. In conventional thermal heads, a heat insulation layer is provided between the heating resistor and the radiator. I found it desirable to implement fast operation and low power consumption to lower the thermal conductivity of the heat insulation layer and to thin the heat insulation layer. I also found there is an optimum heat characteristic to the thickness of one heat insulation layer. I assumed polyimide as a material for the heat insulation layer which could materialize the hypothesis, and studied necessary items based on the thermal calculation. I manufactured a trial thermal head on the basis of this result and confirmed that our assumptions were correct. In addition, to confirm that the assumption is also ultimately correct, I fabricated a trial thermal head only consisting of a heating resistor and without a protective coat and a heat insulation layer. I confirmed that the structure with only the heating resistor exhibited excellent heat response and consumed less power necessary for heating.

The concept of a new visual communication environment is proposed. The proposed system is a shared and interactive multi-user virtual space that consists of a CG-based virtual space structure and video-based objects. Human interface design issues are discussed from the view point of creating a new reality for enhanced communication.

This paper describes an MOS charge-mode version of a T-Model neural-based PCM encoder. The neural-based PCM encoding networks are designed, simulated and implemented using MOS charge-mode circuits. Simulation results are given for both the T-Model and the Hopfield model CMOS charge-mode PCM encoders, and demonstrate the T-Model neural-based one performs the PCM encoding perfectly, while the Hopfield one fails to.

In this paper, we present for the first time two three-dimensional analytical electrostatic Green's functions for shielded and open arbitrarily multilayered medium structures. The analytical formulas for the Green's functions are simply expressed in the form of Fourier series and integrals, and are applicable to the arbitrary number of dielectric layers. In combination with the complex image charge method, we demonstrate an efficient application to analyze microstrip discontinuities in a three-layered dielectric structure. Numerical results for the capacitance associated with on open-end discontinuity show good agreement with those from a previous paper and the effectiveness of using the analytical Green's functions to analyze three-dimensional electrostatic problems.

Cellular Neural Networks (CNNs) are nonlinear dynamic array processors with mainly local interconnections. In most of the applications, the local interconnection pattern, called cloning template, is translation invariant. In this paper, an optimal ring-coding method for rotation invariant description of given set of objects, is introduced. The design methodology of the templates based on the ring-codes and the synthesis of CNN analogic algorithms to detect standing and moving objects in a rotationally invariant way, discussed in detail. It is shown that the algorithms can be implemented using the CNN Universal Machine, the recently invented analogic visual microprocessor. The estimated time performance and the parallel detecting capability is emphasized, the limitations are also thoroughly investigated.

In this paper, a new switched-current auto-tuning filter is proposed. Switched-current (SI) is a current-mode analog sampled-data circuit technique. An SI circuit can be realized using only standard digital CMOS technologies, and is capable of realizing high frequency circuits. The proposed filter is composed of SI-OTA (operational transconductance amplifier) integrators. The gain of an SI-OTA integrator can be electronically controlled by the bias current. The proposed filter is a current controlled filter (CCF) and a PLL technique was used as its tuning method. A 2nd-order SI auto-tuning low-pass filter with 100kHz cutoff frequency was designed assuming a 2µm CMOS process. The characteristics of this SI filter and its tuning characteristics were confirmed by SPICE simulations.

This paper proposes a masked data transferring method for the write-back controlled disk cache system employing a fixed-length recording disk drive, enabling data transfer of discontinuous sectors on the same track between the cache and the disk. This paper also evaluates the method. In write-back controlled disk cache sytems, random write requests cause dirty data (write-pending data on a cache) on discontinuous areas on the cache. It is likely that several sectors on the same track become dirty. These dirty sectors must be written onto the disk according to the cache management scheme. In conventional data transferring methods between a disk cache and a disk drive, plural sectors can be transferred in one single operation when the sectors are adjacent, but discrete sectors must be transferred by individual operations. In the methods, an address of the head sector and number of sectors to be transferred are given to the transfer unit. For example, when two sectors on the same track are located closely but not adjacently, and data transfer is requested for those two sectors, the transfer operation for the second sector must be prepared after the first transfer had completed and before the second sector arrives under the disk head. Although the time for the head to pass by the uninterested sector is often too short for the software overhead for the first transfer to be completed and the second transfer to be prepared, which leads to an unwanted extra rotation of the disk. With the masked transferring method proposed in this paper, the micro program creates a bit-map specifying the target sectors to be transferred and passes it to the data transfer unit, enabling to transfer the discontinuous sectors without latency. The method was evaluated using OLTP warkloads. Results show an improvement in random I/O throughput of between 8% and 27%. The masked transferring method is adopted in Hitachi's A-6521 disk subsytems, shipped since December 1993.