|
1. |
Fault‐tolerant associate memories |
|
Systems and Computers in Japan,
Volume 26,
Issue 7,
1995,
Page 1-12
Eiji Fujiwara,
Tsuyoshi Tanaka,
Preview
|
PDF (1033KB)
|
|
摘要:
AbstractBecause of its capability of high‐speed search, the associative memory (CAM) is expected to be used in a variety of information processing systems. In this paper, novel fault‐tolerant techniques which are effective for on‐line use are proposed for TLB which is an example of the application of CAM.First, fault and error models of the TLB consisting of the CAM part and the SRAM part are clarified. Then, the fault‐tolerant techniques for these faults and errors, such as distance separable technique, cod‐ing technique, simplified 1‐out‐of‐ncheck and graceful degradation, are proposed. The distance separable technique which encodes the data stored in the CAM part is the one which masks the faulty CAM part and prevents errors from propagating to the subsequent circuits. The coding technique checks the one‐to‐one correspondence between the data in the CAM and those in SRAM by using the SEC‐DED code with byte error detection capability, i.e., SEC‐DED‐SbED code, and at the same time it detects and corrects errors in the data stored in SRAM. The simplified 1‐out‐of‐ncheck processes association errors. The graceful degradation gives a flag in the faulty memory section and prevents it from being used.The methods proposed in this paper are evaluated from area augmentation and error detection capability perspectives. The results show that the fault‐tolerant TLB with 32 virtual address bits, 32 physical address bits and 128 entries gives single fault detection probability of nearly 99 perc
ISSN:0882-1666
DOI:10.1002/scj.4690260701
出版商:Wiley Subscription Services, Inc., A Wiley Company
年代:1995
数据来源: WILEY
|
2. |
WAKASHI/C: A storage system for multimedia database |
|
Systems and Computers in Japan,
Volume 26,
Issue 7,
1995,
Page 13-23
Guangyi Bai,
Hirofumi Amano,
Akifumi Makinouchi,
Preview
|
PDF (937KB)
|
|
摘要:
AbstractWAKASHI/C (centralized server version) is a sub‐system at the base level of the “Shusse Uo” project. It provided the C programmer with the basic functions to manipulate the distributed shared persistent data and the distributed shared volatile data. In WAKASHI/C, the virtual memory scheme and the distributed shared memory scheme are combined to realize the distributed shared persistent heap and the distributed and shared volatile heap, for the efficient handling of multimedia data.This paper discusses the realization of WAKASHI/C corresponding to the distributed computing environment. To evaluate the performance for the manipulation of the multimedia database, the extended object operation benchmark is proposed, and the execution performance of WAKASHI/C is eval
ISSN:0882-1666
DOI:10.1002/scj.4690260702
出版商:Wiley Subscription Services, Inc., A Wiley Company
年代:1995
数据来源: WILEY
|
3. |
Test pattern generation for crosstalk faults considering the gate delay |
|
Systems and Computers in Japan,
Volume 26,
Issue 7,
1995,
Page 24-33
Noriyoshi Itazaki,
Kozo Kinoshita,
Hisao Naitoh,
Preview
|
PDF (726KB)
|
|
摘要:
AbstractThis paper presents a new method of test pattern generation for detecting crosstalk faults. In general, to detect crosstalk faults, two successive test vectors are needed for exciting and propagating the crosstalk fault. A special signal corresponding to the two successive vectors is designed to simplify the test pattern generation algorithm and its implementation. By taking into consideration gate delay, more accurate test pattern generation can be expected. The algorithm was implemented in a C++ program, and applied to a few benchmark circuits. The experiments show that coverage for large‐scale integrated circuits is at least 70 percent, and for small‐scale integrated circuits up to 90 perc
ISSN:0882-1666
DOI:10.1002/scj.4690260703
出版商:Wiley Subscription Services, Inc., A Wiley Company
年代:1995
数据来源: WILEY
|
4. |
Efficiency of the conflict‐preserving serializable scheduler for object‐oriented databases |
|
Systems and Computers in Japan,
Volume 26,
Issue 7,
1995,
Page 34-46
Shinichi Taniguchi,
Shojiro Nishio,
Nobuya Kubo,
Preview
|
PDF (1113KB)
|
|
摘要:
AbstractThis paper discusses the concurrency control of the Object‐Oriented Database System (OODBS), and presents a simulation model for the following basic elements which should be considered from the viewpoint of the data structure: (1) the entity should be modeled, based on the concept of object; (2) the definition of the attribute and the method should be inherited based on the class hierarchy; and (3) the composite object should be realized. Based on this model, performance evaluations are conducted on Two‐Phase Locking protocol (2PL) and Conflict‐Preserving Serializable scheduler (CPS). From investigations, so far, on the flat structure, CPS has shown better performance than 2PL. As a result of this study, however, it is shown that in the prominent data structure of OODBS, there exist access conflicts where the optimistic concurrency control such as CPS works effectively, as well as access conflicts where the pessimistic concurrency control such as 2PL works effectively.Accordingly, this paper proposed a new scheduler called Delayed CPS (DCPS), which realizes an adequate control according to the type of access conflict. It is seen that CPS has good concurrency perfor
ISSN:0882-1666
DOI:10.1002/scj.4690260704
出版商:Wiley Subscription Services, Inc., A Wiley Company
年代:1995
数据来源: WILEY
|
5. |
Backpropagation learning algorithm with different learning coefficients for each layer |
|
Systems and Computers in Japan,
Volume 26,
Issue 7,
1995,
Page 47-56
Hirochika Takechi,
Kenji Murakami,
Masanori Izumida,
Preview
|
PDF (703KB)
|
|
摘要:
AbstractWhen considering the algorithm of error backpropagation learning, it is seen that the learning of each layer is not completed independently. This paper proposes a DLBP learning algorithm which uses different learning coefficients for each layer. By changing the ratio of the learning coefficients, the following can be controlled:•the learning progress of the hidden units;•the influence of the hidden units on the network. With DLBP learning, we can construct the corresponding networks to the different requirement such as;•to obtain a network with small number of hidden units; and•to obtain a fault tolerant
ISSN:0882-1666
DOI:10.1002/scj.4690260705
出版商:Wiley Subscription Services, Inc., A Wiley Company
年代:1995
数据来源: WILEY
|
6. |
Characteristics measurement for three‐dimensional surfaces based on multiple filtering |
|
Systems and Computers in Japan,
Volume 26,
Issue 7,
1995,
Page 57-68
Keisuke Iwasaki,
Fumio Kishino,
Preview
|
PDF (1304KB)
|
|
摘要:
AbstractMatching a 3‐D shape to a model in a database is one of the main issues in 3‐D computer vision, along with indexing 3‐D objects for the teleconferencing with realistic sensations. Concerning this issue, one of the keys for reducing the error and computational cost in the matching step is to carry out robust characteristics measurement of the 3‐D shape. This paper proposes a new multiple spatial filtering based characteristics measurement technique for 3‐D surfaces, and presents experiments based on qualitative and quantitative analyses comparing characteristics obtained by the technique and conventional differential geometric characteristics.This method uses the trajectories of a point on a surface caused by viewpoint invariant filtering. By using a set of low‐pass filters with multiple radii, the shape of the trajectory reflects some shape information around the point on the surface, from coarse to fine, while eliminating noise.Basic concepts, the characteristics measurement technique, and experimental results wi
ISSN:0882-1666
DOI:10.1002/scj.4690260706
出版商:Wiley Subscription Services, Inc., A Wiley Company
年代:1995
数据来源: WILEY
|
7. |
An algorithm for affine transformation of binary images stored in pxy tables by run format |
|
Systems and Computers in Japan,
Volume 26,
Issue 7,
1995,
Page 69-78
Kenji Shoji,
Preview
|
PDF (655KB)
|
|
摘要:
AbstractThe algorithm for an affine transformation (such as magnification, reduction, or rotation of a binary image) using a run format and a combination of a skew transformation and a transposition has been known. A transposition using this conventional algorithm requires a relatively large amount of computation time, although its skewing operation is simple and fast.This paper proposes an efficient method of transposition of an image represented by a run format, and its applications to an affine transformation (e.g., rotation, magnification, and reduction of an image). The proposed method uses a simple run data format “pxytables,” in which the start coordinate of black and white runs are stored alternatively in a one‐dimensional array. An experimental comparison of the proposed method with a conventional method shows that the computation time per run of the former is constant, while that of the latter increases with the length of the black runs. In other words, the computation time of an affine transformation of a binary image by using the proposed method is approximately proportional to the number of runs
ISSN:0882-1666
DOI:10.1002/scj.4690260707
出版商:Wiley Subscription Services, Inc., A Wiley Company
年代:1995
数据来源: WILEY
|
8. |
Limits of pictorial pattern discrimination caused by noise originating in image pickup process and the deduced optimum resolution |
|
Systems and Computers in Japan,
Volume 26,
Issue 7,
1995,
Page 79-87
Tsuneo Yoneyama,
Preview
|
PDF (659KB)
|
|
摘要:
AbstractIn an image processing system, the resolution of the input unit serves as an important design parameter in prescribing the processing speed and the scale of the entire system. However, no method to date has been established to derive the specification logically. This paper describes a method for calculating the best resolution in an objective manner by observing the relationship between the resolution and the discrimination limit of figures resulting from the fluctuation of images that are generated in light detection. It is found that the sensitivity in observation at the best resolution coincides with the features of the visual system.
ISSN:0882-1666
DOI:10.1002/scj.4690260708
出版商:Wiley Subscription Services, Inc., A Wiley Company
年代:1995
数据来源: WILEY
|
9. |
A method of motion analysis using spatiotemporal image‐directional temporal plane transform |
|
Systems and Computers in Japan,
Volume 26,
Issue 7,
1995,
Page 88-97
Hiroyasu Taniguchi,
Akinobu Seki,
Shin‐Ichi Kuroda,
Shigeki Ikebata,
Haruki Furusawa,
Preview
|
PDF (1109KB)
|
|
摘要:
AbstractSome of the recent image processing methods proposed for detecting object moving in 3‐D space form a spatiotemporal image by placing consecutive camera images side by side in temporal order. This paper presents one such method, the directional temporal plane transform method (DTT), an image processing method that is not influenced by environmental changes. Assuming that the objects are moving on a nearly constant path, DTT first extracts significant data about these objects from each image, and then makes a 1‐D data stream by projecting the data long a directional axis parallel to the moving loci. After that, by placing the 1‐D data from individual frames side by side in temporal order, the spatiotemporal image is transformed into a 2‐D image on a directional temporal plane. Since the object motion is represented by regions of this 2‐D image, moving objects can be detected by simple 2‐D image processing. This method was used on an experimental basis to detect vehicles running along a road. Experimental results show the effectiveness of the method. An advantage of this method is its ability to detect vehicles using the same processing during both daytime and nighttime—a feature that has not been achieved with convent
ISSN:0882-1666
DOI:10.1002/scj.4690260709
出版商:Wiley Subscription Services, Inc., A Wiley Company
年代:1995
数据来源: WILEY
|
10. |
Staging of awake and sleep based on feature map |
|
Systems and Computers in Japan,
Volume 26,
Issue 7,
1995,
Page 98-107
Takafumi Katayama,
Eiji Suzuki,
Masao Saito,
Preview
|
PDF (883KB)
|
|
摘要:
AbstractThe staging of sleep EEG is executed by the decision by human observations, and is utilized in the diagnosis of insomnia or psychiatric diseases. On the other hand, the sleep EEG includes a large amount of data, and the automation of the sleep staging is considered to be desirable. However, there has not been established a method for staging the biological state based on the awake EEG.This paper proposes an automatic method of staging the biological state based on the hypothesis that the biological state is the same if the same kind of waveforms occur with the same ratio. The proposed method is composed of two procedures: the first is to estimate the standard pattern for the waveform from the individual EEG and the cluster the waveforms; and the second is to examine the kind and ratio of the waveforms appearing in a specified interval of time so that the type of EEG can be identified. The results of staging the sleep EEG is compared to the result of the decision by a doctor, and an agreement of 79 percent is obtained. Generally, it is considered that there can be a difference of some 20 percent in the sleep staging by different doctors. Consequently, it is estimated that the proposed method has a performance comparable to that of the doctor. The staging of the awake EEG also is attempted, and the result is consistent with the skin potential response or the behavior of the subject. Thus it is concluded that the proposed method will be useful in the staging of the biological state.
ISSN:0882-1666
DOI:10.1002/scj.4690260710
出版商:Wiley Subscription Services, Inc., A Wiley Company
年代:1995
数据来源: WILEY
|
|