Nobuhiro DOI Takashi HORIYAMA Masaki NAKANISHI Shinji KIMURA Katsumasa WATANABE
In the hardware synthesis from a high-level language such as C, the bit length of variables is one of the key issues for the area and speed optimization. Usually, designers are required to optimize the bit-length of each variable manually using the time-consuming simulation on huge-data. In this paper, we propose an optimization method of the fractional bit length in the conversion from floating-point variables to fixed-point variables. The method is based on error propagation and the backward propagation of the accuracy limitation. The method is fully analytical and fast compared to simulation based methods.
Thanyapat SAKUNKONCHAK Satoshi KOMATSU Masahiro FUJITA
SpecC language is designated to handle the design of entire system from specification to implementation and of hardware/software co-design. Concurrency is one of the features of SpecC which expresses the parallel execution of processes. Describing the systems which contain concurrent behaviors would have some data exchanging or transferring among them. Therefore, the synchronization semantics (notify/wait) of events should be incorporated. The actual design, which is usually sophisticated by its characteristic and functionalities, may contain a bunch of event synchronization codes. This will make the design difficult and time-consuming to verify. In this paper, we introduce a technique which helps verifying the synchronization of events in SpecC. The original SpecC code containing synchronization semantics is parsed and translated into a Boolean SpecC code. The difference decision diagrams (DDDs) is used to verify for event synchronization on Boolean SpecC code. The counter examples for tracing back to the original source are given when the verification results turn out to be unsatisfied. Here we also introduce idea on automatically refinement when the results are unsatisfied and preset some preliminary results.
This paper proposes an efficient method for design space exploration of the global optimum configuration for parameterized ASIPs. The method not only guarantees the optimum configuration, but also provides robust speedup for a wide range of processor architectures such as SoC, ASIC as well as ASIP. The optimization procedure within this method takes a two-steps approach. Firstly, design parameters are partitioned into clusters of inter-dependent parameters using parameter dependency information. Secondly, parameters are optimized for each cluster, the results of which are merged for global optimum. In such optimization, inferior configurations are extensively pruned with a detailed optimality mapping between dependent parameters. Experimental results with mediabench applications show an optimization speedup of 4.1 times faster than the previous work on average, which is significant improvement for practical use.
Masashi SUGIYAMA Hidemitsu OGAWA
In supervised learning, the selection of sample points and models is crucial for acquiring a higher level of the generalization capability. So far, the problems of active learning and model selection have been independently studied. If sample points and models are simultaneously optimized, then a higher level of the generalization capability is expected. We call this problem active learning with model selection. However, active learning with model selection can not be generally solved by simply combining existing active learning and model selection techniques because of the active learning/model selection dilemma: the model should be fixed for selecting sample points and conversely the sample points should be fixed for selecting models. In this paper, we show that the dilemma can be dissolved if there is a set of sample points that is optimal for all models in consideration. Based on this idea, we give a practical procedure for active learning with model selection in trigonometric polynomial models. The effectiveness of the proposed procedure is demonstrated through computer simulations.
This paper addresses bitwidth optimization focusing on leakage power reduction for system-level low-power design. By means of tuning the design parameter, bitwidth tailored to a given application requirements, the datapath width of processors and size of memories are optimized resulting in significant leakage power reduction besides dynamic power reduction. Experimental results for several real embedded applications, show power reduction without performance penalty range from about 21.5% to 66.2% of leakage power, and 14.5% to 59.2% of dynamic power.
Minoru KURIBAYASHI Hatsukazu TANAKA
One of the important topics of watermarking technique is a robustness against geometrical transformations. In the previous schemes, a template matching is performed or an additional signal is embedded for the recovery of a synchronization loss. However, the former requires the original template, and the latter degrades the quality of image because both a watermark and a synchronization signal must be embedded. In the proposed scheme only a synchronization signal is embedded for the recovery of both a watermark and a synchronization loss. Then the embedded information depends on the distance between two embedded signal positions. The distance is not changed seriously by random geometrical transformations like StirMark attack unless the embedded signal is disturbed. Therefore, a watermark can be extracted correctly from such geometrically transformed image if the synchronization signal can be recovered.
A.A. Seyno SLUYTERMAN Tjerk G. SPANJER
The size of the microscopic electron spot is an important parameter for the white-uniformity of a CRT. It changes as a function of the focus voltage and beam repulsion. This paper explains the mechanism behind this phenomenon. The model is supported by means of measurements.
Digital Subtraction Angiography (DSA) is a technique used for enhancement of small details in angiogram imaging systems. In this approach, X-ray images of a subject, after injection, are subtracted from a reference X-ray image, taken from the same subject before injection. Due to the exponential absorption property of X-rays, effects of small details at different depth appear differently on X-ray images. Consequently, image subtraction cannot be employed on the original images without any adjustment or modification. Proper modification, in this case, is to use some form of logarithmic operation on images before subtraction. In medical imaging systems, the system designer has a choice to implement this logarithmic operation in the analog domain, before digitization of the video signal, or in the digital domain after analog-to-digital conversion (ADC) of the original video signal. In this paper, the difference between these two approaches is studied and upper bounds for quantization error in both cases are calculated. Based on this study, the best approach for utilization of the logarithmic function is proposed. The overall effects of these two approaches on the inherent signal noise are also addressed.
Parameter-optimized cubic convolution is used to accurately analyze the pitch center, rate and extent of vibrato tones. We interpolate the time-tracing fundamental frequencies of vibrato tones using parametric cubic convolution, and analytically estimate the positions and values of the extrema, which are used to analyze the characteristics of the vibrato. The optimal values of α, the parameter of the interpolation kernel, are also shown as a function of the normalized vibrato rates.
The purpose of this paper is to propose a novel cluster map based blind RBF equalizer for received signal constellation (RSC) independent channel, which belongs to RSC based blind equalization approach. Without channel estimator, firstly, the desired numbers of unlabeled RBF centers are obtained by an unsupervised clustering algorithm. Then a cluster map generated from the known RBF equalizer structure is used to partition the unlabeled centers into appropriate subsets merely by several simple sorting operations, which corresponds to the weight initialization. Finally, the weight is adjusted iteratively by an unsupervised least mean square (LMS) algorithm. Since the process of the weight initialization using the underlying structure of RBF equalizer is very effective, the proposed blind RBF equalizer can achieve almost identical performance with the optimal RBF equalizer. The validity of the proposed equalizer is also demonstrated by computer simulations.
Jiahai WANG Zheng TANG Qiping CAO Xinshun XU
Edge linking is a fundamental computer vision task, yet presents difficulties arising from the lack of information in the image. Viewed as a constrained optimization problem, it is NP hard-being isomorphic to the classical Traveling Salesman Problem. This paper proposes a gradient ascent learning algorithm of the elastic net approach for edge linking of images. The learning algorithm has two phases: an elastic net phase, and a gradient ascent phase. The elastic net phase minimizes the path through the edge points. The procedure is equivalent to gradient descent of an energy function, and leads to a local minimum of energy that represents a good solution to the problem. Once the elastic net gets stuck in local minima, the gradient ascent phase attempts to fill up the valley by modifying parameters in a gradient ascent direction of the energy function. Thus, these two phases are repeated until the elastic net gets out of local minima and produces the shortest or better contour through edge points. We test the algorithm on a set of artificial images devised with the aim of demonstrating the sort of features that may occur in real images. For all problems, the systems are shown to be capable of escaping from the elastic net local minima and producing more meaningful contours than the original elastic net.
Shigemasa TAKAI Toshimitsu USHIO
In this paper, we study reliable decentralized supervisory control of discrete event systems with a control architecture where certain controllable events are controlled under the conjunctive fusion rule, and certain others are controlled under the disjunctive fusion rule. We first introduce a notion of reliable co-observability with respect to such a partition of the controllable event set. We then prove that reliable co-observability together with Lm(G)-closure and controllability is a necessary and sufficient condition for the existence of a reliable decentralized supervisor under a given partition. Moreover, we present necessary and sufficient conditions for the existence of a partition of the controllable event set under which a given specification language is reliably co-observable.
A special group of voice application services (VASs) are promising contents for wireless as well as wireline networks. Without a designated call admission policy, VAS calls are expected to suffer from relatively high probability of blocking since they normally require better signal quality than ordinary voice calls. In this letter, we consider a prioritized call admission design in order to reduce the blocking probability of VAS calls, which makes the users feel the newly-provided VAS in belief. The VAS calls are given a priority by reserving a number of channel-processing hardwares. With the reservation, the blocking probability of prioritized VAS calls can be evidently reduced. That of ordinary calls, however, is increasing instead. This letter provides a system model that counts the blocking probabilities of VAS and ordinary calls simultaneously, and numerically examines an adequate level of the prioritization for VAS calls.
For the reception of MC-CDMA signals in a frequency-selective fading channel, frequency-domain equalization is necessary before despreading. In this paper, joint antenna diversity combining and one-tap frequency-domain equalization is considered (simply referred to as the joint antenna diversity & equalization, in this paper). A receiver structure for joint antenna diversity & equalization is presented and the unified weights based on minimum mean square error (MMSE) criterion are found in the presence of multi-users with different spreading factors and transmit powers. For comparison, antenna diversity combining after despreading using MMSE combining (MMSEC) is also considered. The achievable bit error rate (BER) performances with joint antenna diversity & equalization and with antenna diversity after MMSEC despreading in a frequency-selective Rayleigh fading channel are evaluated by computer simulations and compared.
Masayuki JYUMONJI Yoshinobu KIMURA Masato HIRAMATSU Yukio TANIGUCHI Masakiyo MATSUMURA
A two-dimensional laser beam profiler has been developed that can measure the intensity distribution on a sample surface of a single-shot of an excimer-laser light beam from not only the macroscopic viewpoint, but also the microscopic viewpoint, which is important to excimer-laser triggered lateral large-grain growth of Si. A resolution as fine as 0.4 µm was obtained with a field of view of as large as 30 µm 30 µm. The effects of homogenizers, phase-shifters, and their combination on beam profiles were quantitatively investigated by using this apparatus. The relationship between the microscopic beam profile and the surface morphology of laterally grown grains was also examined.
The adaptive cross-spectral (ACS) technique recently introduced by Okuno et al. provides an attractive solution to acoustic echo cancellation (AEC) as it does not require double-talk (DT) detection. In this paper, we first introduce a generalized ACS (GACS) technique where a step-size parameter is used to control the magnitude of the incremental correction applied to the coefficient vector of the adaptive filter. Based on the study of the effects of the step-size on the GACS convergence behaviour, a new variable step-size ACS (VSS-ACS) algorithm is proposed, where the value of the step-size is commanded dynamically by a special finite state machine. Furthermore, the proposed algorithm has a new adaptation scheme to improve the initial convergence rate when the network connection is created. Experimental results show that the new VSS-ACS algorithm outperforms the original ACS in terms of a higher acoustic echo attenuation during DT periods and faster convergence rate.
Zhibin PAN Koji KOTANI Tadahiro OHMI
A fast winner search method based on separating all codewords in the original codebook completely into a promising group and an impossible group is proposed. Group separation is realized by using sorted both L1 and L2 norms independently. As a result, the necessary search scope that guarantees full search equivalent PSNR can be limited to the common part of the 2 individual promising groups. The high search efficiency is confirmed by experimental results.
Naoki IMASAKI Ambalavanar THARUMARAJAH Shinsuke TAMURA Toshiaki TANAKA
This paper proposes a simulation framework suitable for holonic manufacturing systems, or HMS, based on the concept of distributed self-simulation. HMS is a distributed system that comprises autonomous and cooperative elements called holons, for the flexible and agile manufacturing. The simulation framework proposed here capitalizes on this distributed nature, where each holon functions similar to an independent simulator with self-simulation capabilities to maintain its own clock, handle events, and detect inter-holon state inconsistencies and perform rollback actions. This paper discusses the detailed architecture and design issues of such a simulator and reports on the results of a prototype.
Qihong GE Jianhua LU Shunliang MEI
The channel noise in OFDM systems affects the accuracy of channel estimation, deteriorating the performance of equalization. We present a novel algorithm with MMSE (Minimum Mean Square Error) channel estimation based on Hadamard Transform, to mitigate the effects of noise. The performance of the proposed algorithm is proved to be better than that with LS (Least Square) estimation, and very close to that with the MMSE estimation based on Fourier Transform, while the computation required is pretty small due to the use of Hadamard Transform.
Takeshi TATEYAMA Seiichi KAWATA Hideaki OHTA
In this paper, a new grouping method for Group Technology using Self-Organizing Map (SOM) is proposed. The purpose of our study is to divide machines in a factory into any number of cells so that the machines in each cell can process a similar set of parts to increase productivity. A main feature of our method is to specify not only the number of the cells but also the maximum and minimum numbers of machines in a cell. Some experimental results show effectiveness of our proposed algorithm.