Language：Japanese  

CiNii Article

researchmap

Language：Japanese  

When working on computer, users often re-refer to past files. Some applications have a function to list recently opened files for file reuse support. However, when working using multiple files, it is generally to manage those files in folders, and it may be more useful to list the Working Directories than single files. Therefore, in this paper, we proposed a system that supports re-referencing by grouping Working Directories from a specific viewpoint and grouping them in a higher-level virtual folder. In order to realize the proposed system, the Working Directory discovering method and the classification method were described and evaluated. Finally, we show the design and implementation of the proposed system.

CiNii Article

CiNii Books

researchmap

Publishing type：Research paper (international conference proceedings)  

DOI： 10.1007/978-3-030-70639-5_31

researchmap

Other Link： https://dblp.uni-trier.de/db/conf/eidwt/eidwt2021.html#GotohISTOUN21

Language：Japanese  

A specialized operating system (OS) can provide some order-made service efficiently. However, development of the specialized OS takes a large amount of cost (e.g., device driver, file management functions). Thus, it is important to reduce the cost. In this paper, we describe a method for introducing the specialized OS and Linux run independently in multicore processors, and the specialized OS uses Linux file I/O functions. This paper describes the design and the implementation of the proposed method for AnT operating system as the specialized OS, and reports the evaluation results of the proposed method.

CiNii Article

CiNii Books

researchmap

Language：Japanese  

Since the e-mail system is a long-standing technology, it is highly interoperable and widely used as a basic communication tool between organizations. When transmitting confidential information such as personal information by e-mail, it is common in Japan to send an encrypted attachment-file and send the common key password separately in plaintext. In this paper, we name this as eZIP mail and point out the problem of e-mail attached with password. We also propose a method to solve the problem on the recipients side. Our proposed method efficiently finds the password mail corresponding to the encrypted mail and the password string contained therein and automates password decryption. Furthermore, we implemented the proposed method and evaluate the result of incorporating it into the personal e-mail system. In the evaluation, it shows that it is possible to decrypt many of mails with several password trials.

CiNii Article

CiNii Books

researchmap

Language：Japanese   Publishing type：Research paper (scientific journal)  

In research and development on calendar applications, there is a demand for a large number of calendar data that reflects characteristics of individuals or organizations. In this paper, we propose a method for generating dummy data of calendar information. First, we proposed a model that expresses group calendars reflecting the characteristics of organizations and individuals. Next, by reproducing the time series of planning behavior, we generated dummy calendars with reality. As a result, we showed that we can simulate how organizational changes and personal work changes will change existing calendars.

CiNii Article

CiNii Books

researchmap

Language：Japanese   Publishing type：Research paper (scientific journal)  

Social coding is getting widespread among software development societies. Various users propose contributions improving source code for social coding projects. Not only the proposal itself but also the contributor's reputation affects the decision on acceptance/rejection of these contributions. In this paper, we described an indicator of the quality of contributions excluding the bias due to contributor's reputation, and showed a method to measure the indicator by learning user's behavior related to the contribution. Then, by evaluating the indicator calculated by this method, we showed that the indicator performs effectively. Finally, we confirmed that the bias due to contributor's reputation may exist in actual decision on acceptance/rejection of contributions by analyzing evaluation results.

CiNii Article

CiNii Books

researchmap

Language：Japanese   Publishing type：Research paper (scientific journal)  

J-GLOBAL

researchmap

Language：Japanese   Publishing type：Research paper (scientific journal)  

For better productivity, it is crucial to manage the association between calendar-tasks and e-mails. However, if the management itself requires a large amount of effort, further improvements cannot be expected. In this paper, we propose a new method for reducing management time. Our method leverages the fact that e-mails and calendar-tasks are used repeatedly at periodic intervals. Our system equips a new interface to reuse the articles. Through the interface, the system learns the relationships among the articles. Utilizing the stored relationships, our system suggests the potential relationships to users. We also evaluated our system to confirm the reduction in management time.

CiNii Article

CiNii Books

researchmap

Language：Japanese   Publishing type：Research paper (scientific journal)   Publisher：Information Processing Society of Japan (IPSJ)  

We proposed a model that is capable of handling the patterns of ambiguous recurring tasks. Although, this model is believed useful for prediction of the tasks which might happen in the future, it is not clear how extract the necessary information from the model. This paper proposes a method for forecasting the future tasks by extracting statistical information from the model. First, some additional terms to express the statistical information are introduced. Next, we propose a new method for forecasting. Finally, we show typical examples of recurring tasks, and perform some evaluation of the method on these examples.

CiNii Article

CiNii Books

researchmap

Language：English   Publishing type：Research paper (international conference proceedings)   Publisher：IEEE  

In a typical office environment, employees are expected to collect information about their tasks, in the event that they must hand them over to others. This task information is usually stored in multiple business applications. In this paper, we investigated how task-related information is stored across multiple applications through an actual workflow. In addition, we address two problems encountered when collecting work information from applications. First, it is difficult to collect information that directly relates to the work unit or task unit. Second, it is unclear what type of information should be handed over at the beginning of tasks. To solve these problems, we propose a new method for acquiring work history from multiple applications. The proposed method manages the application history associated with work unit or task unit. Additionally, the proposed method manages the priorities of histories. Finally, we prove that the proposed method can effectively collect work information.

DOI： 10.1109/NBiS.2015.70

Web of Science

researchmap

Other Link： https://dblp.uni-trier.de/db/conf/nbis/nbis2015.html#OkadaN15

Language：Japanese  

Buffer cache is implemented to improve I/O performance with data in disks. As buffer cache management, there are many mechanisms based on access pattern of block. On the other hand, we proposed I/O buffer cache mechanism based on the frequency of file usage (FFU). Our previous proposed mechanism calculates file importance from the information of system-call of the file operation. Then, it controls two level buffer cache based on the file importance. In this paper, we propose a setting method of opportunity of updating file importance on FFU. The proposed method focuses on whether the file state is access intensive or not. This paper also describes a setting method of parameters of the proposed method based on access information of a target system. Finally, this paper reports the evaluation results of the proposed method by using typical access pattern data and real access patterns.

CiNii Article

CiNii Books

researchmap

Language：Japanese   Publishing type：Research paper (scientific journal)  

CiNii Article

CiNii Books

researchmap

Language：Japanese   Publishing type：Research paper (scientific journal)  

We proposed a model that is capable of handling the patterns of ambiguous recurring tasks. We also implemented a new calendaring system conforming with the ambiguous recurring tasks. By using our calendaring system, it would be useful for prediction of the tasks which might happen in the future, and also be the hint on handing them over to co-workers. In this paper, we have performed some evaluation of the calendaring system. First of all, we recall our model and the concept of our calendaring system. And we have shown some advantages of our system by performing a trial by comparing with conventional calendaring systems. In addition, we have inspected the effectiveness of our concept in the system.

CiNii Article

CiNii Books

researchmap

Other Link： http://id.nii.ac.jp/1001/00090271/

Language：English   Publishing type：Research paper (international conference proceedings)   Publisher：IEEE  

Today, the sharing of schedule information across individuals or organizations has been made simpler by the Internet and personal devices such as smartphones. However, it is not as easy in bring-your-own-device environments due to problems related to privacy and cultural gaps among organizations and individuals. Current calendar systems do not have any easy, fine-grained calendar sharing scheme to support such environments. In this paper, we propose a new calendar system that employs a fine-grained calendar sharing scheme called Virtual Calendar, which is based on our CalDAV proxy engine. Our calendar system satisfies such fine-grained demands as "I want to share my calendar with my wife for only those articles that start after 5 pm or on weekends."

DOI： 10.1109/BWCCA.2013.95

Web of Science

researchmap

Other Link： https://dblp.uni-trier.de/db/conf/bwcca/bwcca2013.html#NomuraMTUM13

Language：English   Publishing type：Research paper (international conference proceedings)   Publisher：IEEE Computer Society  

In order to utilize computer resources efficiently, technologies have been developed to enable running multiple OSes on a single computer. In our study, we developed the Mint operating system, which enables running multiple Linux instances on a multi-core processor. Mint is different from traditional virtualization technologies in that Mint can achieve performance isolation between Linux instances by segmenting computer resources. Currently, Mint statically segments and assigns its physical memory to each Linux instance. By transferring physical memory ranges among the Linux instances depending on their physical memory utilization, Mint can adapt to the load variations. In this paper, we describe the requirements for segmentation and transfer of physical memory and provide a solution. © 2013 IEEE.

DOI： 10.1109/NBiS.2013.58

Scopus

researchmap

Other Link： https://dblp.uni-trier.de/db/conf/nbis/nbis2013.html#MiyazakiNT13

Language：Japanese  

CiNii Article

J-GLOBAL

researchmap

Language：Japanese  

CiNii Article

J-GLOBAL

researchmap

Language：Japanese   Publisher：The Institute of Electronics, Information and Communication Engineers  

CiNii Article

CiNii Books

researchmap

Language：English   Publishing type：Research paper (international conference proceedings)   Publisher：IEEE COMPUTER SOC  

We proposed a model for detecting patterns in tasks that recur irregularly. This model may be useful for predicting future tasks, but it is not clear how to extract the necessary information from the model. This paper proposes a method for forecasting future tasks by extracting statistical information from the model. First, we introduce some additional terms for expressing the statistical information. Next, we propose a new method for forecasting. Finally, we provide typical examples of recurring tasks and evaluate the method using these examples.

DOI： 10.1109/BWCCA.2012.111

Web of Science

researchmap

Other Link： https://dblp.uni-trier.de/db/conf/bwcca/bwcca2012.html#YoshiiNT12

Language：English   Publishing type：Research paper (international conference proceedings)   Publisher：IEEE Computer Society  

We proposed a model that is capable of handling the patterns of ambiguous recurring tasks. If we can grasp the pattern, it would be useful for prediction of the tasks which might happen in the future, and also be the hint on handing them over to co-workers. To utilize our model, we have developed a concrete method for manipulation or visualization of the modeled data. This method is useful to build a nifty user interface to the calendaring systems. This paper shows an implementation of calendaring system conforming with ambiguous recurring tasks. First of all, we recall our model. Next, we propose some basic operations for the modeled data, EDIT, READ, and INHERIT. Finally, we describe the implementation of the method as an interface to our calendaring system. © 2012 IEEE.

DOI： 10.1109/WAINA.2012.119

Scopus

researchmap

Other Link： https://dblp.uni-trier.de/db/conf/aina/ainaw2012.html#NomuraMT12

Language：English   Publishing type：Research paper (international conference proceedings)   Publisher：IEEE Computer Society  

In a typical office, there are many recurring tasks. If the recurrence pattern can be determined, then this pattern can be used to forecast which task might occur or provide guidance for a coworker who takes on the given responsibilities. A calendaring system is a popular tool for task enumeration. However, the existing major implementations are not capable of handling fluctuation cycle in the recurrences or the relationships between the tasks, and thus these systems do not give the user any analytical perspective on the tasks. Therefore, this paper proposes a new calendaring system that can handle changes in the recurrences and display the relationships between tasks. Three new concepts, Task, Mission, and Job are introduced to describe the relationships between the tasks. Interoperability with the existing systems is also described. © 2011 IEEE.

DOI： 10.1109/NBiS.2011.91

Scopus

researchmap

Other Link： https://dblp.uni-trier.de/db/conf/nbis/nbis2011.html#NomuraMT11

Language：English   Publishing type：Research paper (international conference proceedings)   Publisher：IEEE Computer Society  

In this paper, we describe the implementation and evaluation of the Mint operating system, which runs multiple Linux instances by partitioning cores, memory blocks, and I/O devices. The proposed method is not a type of virtualization, hence, it makes the most efficient use of available cores. By performing some evaluations, we confirmed that all OS instances in Mint have performances equal to those in the original Vanilla Linux. In addition, because all the instances are completely independent of one another, Mint will not be affected by performance issues, even under the forthcoming many-core environments. In addition, each instance is fully armored Linux. This means that users can still use existing software assets. © 2011 IEEE.

DOI： 10.1109/BWCCA.2011.90

Scopus

researchmap

Other Link： https://dblp.uni-trier.de/db/conf/bwcca/bwcca2011.html#NomuraSNUKT11

Language：Japanese  

CiNii Article

J-GLOBAL

researchmap

Language：Japanese   Publisher：The Institute of Electronics, Information and Communication Engineers  

CiNii Article

CiNii Books

researchmap

Publishing type：Research paper (international conference proceedings)   Publisher：Springer  

DOI： 10.1007/978-3-642-12165-4_26

researchmap

Other Link： https://dblp.uni-trier.de/db/conf/iccsa/iccsa2010-2.html#NomuraOGNTYM10

Language：English  

DOI： 10.1109/CSA.2009.5404183

researchmap

Language：Japanese  

CiNii Article

J-GLOBAL

researchmap

Language：English   Publishing type：Research paper (international conference proceedings)   Publisher：IEEE  

Recent operating systems are capable of using dynamic loadable module functions to update the device driver programs without shutting down the entire system. However, even if a system continues to operate, some services provided by the device driver will be temporarily unavailable during the run-time update. To reduce the period during which the services are unavailable, we propose a new method for run-time updating of device drivers; this method uses context migration from an old driver to an updated one. In this study, we describe the implementation of a network-device driver using our method. In addition, we have shown some advantages of our method by performing a trial on the FreeBSD operating system.

DOI： 10.1109/NBiS.2009.53

Web of Science

J-GLOBAL

researchmap

Other Link： https://dblp.uni-trier.de/db/conf/nbis/nbis2009.html#TanakaNT09

Language：Japanese  

CiNii Article

J-GLOBAL

researchmap

Language：Japanese   Publisher：The Institute of Electronics, Information and Communication Engineers  

CiNii Article

CiNii Books

researchmap

Language：Japanese   Publisher：The Institute of Electronics, Information and Communication Engineers  

CiNii Article

CiNii Books

J-GLOBAL

researchmap

Language：Japanese  

CiNii Article

J-GLOBAL

researchmap

Publishing type：Research paper (international conference proceedings)   Publisher：IEEE  

Fine-grained multithreading based on a natural model, such as dataflow model, is promising in achieving high efficiency and high programming productivity. In this paper, we discuss operating system issues for fine-grained multithread programs. We are developing an operating system called CEFOS based on a dataflow based computation model. A program on CEFOS consists of zero-wait threads which run to completion without suspension once started. Firing control among such threads is performed in a dataflow manner along with continuation relations in the program. Target platforms include Fuce processor, which is dedicated to fine-grained multithreading, and commodity processors such as Intel. x86. In this paper, after introducing our basic model and our operating system model, we discuss implementation issues on Fuce and commodity platforms. The evaluation results indicate that our approach on commodity platforms is effective in reducing overheads while our approach on a special architecture naturally exploit parallelism even in I/O handling. © 2007 IEEE.

DOI： 10.1109/IPDPS.2007.370678

Scopus

researchmap

Other Link： https://dblp.uni-trier.de/db/conf/ipps/ipdps2007.html#KusakabeYAIANTA07

Publishing type：Research paper (international conference proceedings)   Publisher：ACM  

DOI： 10.1145/1242531.1242564

researchmap

Other Link： https://dblp.uni-trier.de/db/conf/cf/cf2007.html#KusakabeAIANTA07

Language：English   Publishing type：Research paper (scientific journal)   Publisher：SCRIPTA TECHNICA-JOHN WILEY & SONS  

A method to independently run two coexisting Linux systems on one computer is proposed. Up to the present tirne a virtual computer method has been used as a method to run multiple operating systems on one computer. This method has a problem, however, in which the guest operating system depends on the running state of the host operating system making it impossible to efficiently utilize the processing performance of each operating system. In contrast to this, because each operating system runs independently in the proposed method by time sharing the processes and exclusively dividing the other hardware resources, one of the operating systems and applications running on the operating systems are not influenced by the other operating system and the applications running on the operating system. This paper presents the proposed method and reports the results of performance evaluations when actually implernented in Linux. (C) 2007 Wiley Periodicals, Inc. Electron Comm Jpn Pt 3,90(9):56-68,2007; Published online in Wiley InterScience (www.interscience. wiley.com).

DOI： 10.1002/ecjc.20343

Web of Science

researchmap

Publishing type：Research paper (scientific journal)  

With increases in the performance of computers, it has become possible to provide a large number of services on a single computer. However, since the required execution performance can vary from service to service, it is necessary to guarantee execution performance for each service individually. In addition, it is common for a single service to be composed of multiple processes. Consequently, the ability to regulate execution performance for units consisting of multiple processes is desirable. Therefore, in this paper we propose a mechanism for regulating the execution performance of process groups using execution resources that encapsulate the assignable processor units designed for the Tender operating system. Specifically, executions are managed in a tree-structure and we then regulate execution performance by associating processes with these executions. In addition, we show via an implementation and evaluations that the proposed method is able to regulate the execution performance of process groups well and present an evaluation that makes use of a Web server. © 2007 Wiley Periodicals, Inc.

DOI： 10.1002/scj.20403

Scopus

researchmap

Language：Japanese  

J-GLOBAL

researchmap

Language：English   Publishing type：Research paper (international conference proceedings)   Publisher：INT INST INFORMATICS & SYSTEMICS  

Recently, leakage of classified information such as personal information has become serious problem. To prevent the leakage, it is important to know how the information diffuses in a computer system. In this paper, we propose a method for tracing the diffusion of the classified information.
We assumed that the classified information exists in a set of files on a file system. These files will be diffused by processes (programs). Once a process touched any classified file, it becomes a suspicious carrier process. The carrier process will make other classified files. Therefore, to keep track of the diffusion, we have to watch all processes to see if they touch any classified files. That is, we have to insert some probes into an operating system to watch file manipulation system-calls of processes such as OPEN and CLOSE.
We describe the algorithm of the method and an implementation of the method on the Linux kernel. We also show that the performance of our implementation is efficient.

Web of Science

researchmap

Language：Japanese   Publisher：Information Processing Society of Japan (IPSJ)  

Operating systems had been supposed to provide a single environment to their applications concealing details of various hardware architectures. However, in recent years, we can observe a different relationship between operating systems and hardwares: users run various operating systems concurrently or alternatively on a single PC. And more, these operating systems are able to share hardware components of a PC using a kind of hardware sharing mechanism. Based on this mechanism, we propose a program design framework to monitor PCI devices. Such kind of softwares were conventionally embedded in device drivers, on the other hand, our framework is on off-the-OS level, and has a lower level inspection method which brings us an effective and precise monitoring.

CiNii Article

CiNii Books

researchmap

Other Link： http://id.nii.ac.jp/1001/00018323/

Language：Japanese   Publisher：The Institute of Electronics, Information and Communication Engineers  

CiNii Article

CiNii Books

researchmap

Language：Japanese   Publisher：The Institute of Electronics, Information and Communication Engineers  

CiNii Article

CiNii Books

researchmap

Language：Japanese   Publisher：Information Processing Society of Japan (IPSJ)  

This paper presents a new mechanism to reduce the overhead about processing in which a lot of system calls are issued. The proposed method changes the process of an application program running in the user mode to the process running in the supervisor mode at any time. In a word, the process can use two running modes of the user mode and the supervisor mode during executing the application program. The application program running in the supervisor mode uses the system call by the function call. As a result, the overhead of the system call processing can be reduced by executing the process in the supervisor mode for the application program to which efficient execution is required.

CiNii Article

CiNii Books

J-GLOBAL

researchmap

Language：Japanese   Publisher：Information Processing Society of Japan (IPSJ)  

According to the increasing amount of data transferred between computers and peripheral devices, the frequency of interrupts from peripheral devices increases. Thus, efficiency of interrupt handling is one of the key issues to realize high performance computing environments. In order to reduce interrupt overheads, we propose an interrupt packaging mechanism that aggregates main handlers of a series of interrupts. We have designed and implemented an interrupt packaging mechanism for interrupts from NIC (Network Interface Card) of Myrinet (1.28Gbps). In this paper, we report design and implementation of an interrupt packaging mechanism, and our performance evaluation.

CiNii Article

CiNii Books

researchmap

Other Link： http://id.nii.ac.jp/1001/00011100/

Language：Japanese   Publisher：Information Processing Society of Japan (IPSJ)  

Software is continuously changed and evolved at users' requests. Consequently, software tends to lose its original concrete structure and become complex. Maintaining such software is extremely costly. Programmer needs to restructure the software before losing control of it. We can use a technique called "Refactoring". Refactoring is to reorganize a program without changing its function. In a refactoring procedure, it is difficult for a novice programmer to find out where he applys refactoring to and which refactoring method he should use. In this paper, we propose a mechanism for supporting automated refactoring process.

CiNii Article

CiNii Books

researchmap

Other Link： http://id.nii.ac.jp/1001/00011210/

Language：Japanese   Publisher：Information Processing Society of Japan (IPSJ)  

By an appearance of various parallel processing types, effective execution of parallel processing with heavy interaction between user and kernel is demanded. So this paper proposes a wrapped system-call mechanism that executes multiple requests as one system-call in a mass. This mechanism has the characteristics that can reduce load of processing with a system-call, and that can reduce kernel call frequency. And, by realization example of this mechanism, a wrapped system-call mechanism shows that parallel processing is executed effectively in high performance processor with many pipeline stages.

CiNii Article

CiNii Books

researchmap

Publishing type：Research paper (international conference proceedings)  

With the spread of the Internet, services that communicate to other services are increasing. Multimedia applications such as video on demand also ask for network Quality of Service (QoS). Thus, operating systems have to guarantee the allocation of computer resources to services. The computer resources are CPU, disk, network devices and so on. We suppose that the communications have to be controlled well, because services using network are increasing. This paper proposes the communication bandwidth control mechanism by regulating program execution speed. Our proposed mechanism is based on the process schedule method for regulating program execution speed. In the process schedule method, the operating system reserves the amount of CPU time of target processes and guarantees the allocation of CPU time on a sending host. Our proposed mechanism can guarantee a required data rate of target processes by allocating enough CPU time for communications. Because operating systems manage computer resources, they guarantee the allocation of CPU time if the process schedule method is implemented. The allocation of CPU time is not almost affected by non-target processes. This paper introduces the process schedule method and the implementation of it. This paper also shows how to control the communication bandwidth of target processes. Then this paper describes about an evaluation of our proposed mechanism.

Scopus

researchmap

Publishing type：Research paper (international conference proceedings)   Publisher：IASTED/ACTA Press  

We are developing an operating system, CEFOS in order to realize an efficient parallel and distributed computing environment based on a fine-grain multi-threading model. In CEFOS, a user process consists of a set of threads which are executed concurrently according to thread dependence. A thread in CEFOS is very fine-grain compared to a conventional thread such as Pthreads. By consisting a process with fine-grain threads, we can employ flexible scheduling schemes as long as thread dependence allows. In this paper, we propose Wrapped System Call mechanism using a cooperation mechanism between user-mode and kernel-mode. This mechanism aggregates multiple system-calls and executes them as a single system-call. In order to reduce overhead of context switches and mode changes between user-mode and kernel-mode. We also show the performance evaluation results and discuss the effectiveness of our mechanism.

Scopus

researchmap

Other Link： https://dblp.uni-trier.de/db/conf/appinf/appinf2003.html#KusakabeNTA03

Language：Japanese   Publisher：Kyushu University  

Software flexibility and extendibility reflect program architecture. For example, it is difficult to add functions and to maintain programs because of their complexity. We need to improve a software carefully if we want to keep it in high quality. We can use "Refactoring" which is the technique to reorganize a program without changing its functions. Refactoring methods allow existing programs to change easily against future modification and maintenance. We have to find out where we apply refactoring and which refactoring method we should use. In this paper, we propose a tool to support refactoring for Object Pascal programs using software metrics.

DOI： 10.15017/1525462

CiNii Article

CiNii Books

J-GLOBAL

researchmap

Other Link： http://hdl.handle.net/2324/1525462

Language：Japanese   Publisher：Information Processing Society of Japan (IPSJ)  

Software flexibility and expandablity reflect program architecture. For example, it is difficult to add functions and to maintenance programs because of its complexity. We need to improve software carefully if we want to keep it high quality. We can use a technique called "Refactoring" . "Refactoring" is to reorganize a program without changing its functions. Refactorings are used to make other program changes easier for future modification and maintenance. We have to find out where we apply refactoring and which refactoring we should use. This presentation proposes a mechanism to support automated refactoring process. Our approach to indentifying the refactoring candidate is based on software metrics. This mechanism help us to choose an appropriate refactoring method when we improve programs without experience and knowledge.

CiNii Article

CiNii Books

researchmap

researchmap

Language：Japanese  

CiNii Article

J-GLOBAL

researchmap

Publishing type：Research paper (scientific journal)  

DOI： 10.1002/scj.10043

researchmap

Language：Japanese   Publisher：Information Processing Society of Japan (IPSJ)  

Recently, computer networks such as the Internet become more popular. We have more opportunities of exchanging schedule information or arranging schedules on a network. In the past, we were able to use "groupware" for this purpose. But nowadays, computer network is not as like the groupware demands for. This makes the situation more difficult. In this paper, we describe what kind of problems would happen when we use schedule information on a network. Then we describe their solutions. And finally, we describe how the schedule management system, MHC (Message Harmonized Calendaring system) which we are developing, works to solve their problems.

CiNii Article

CiNii Books

researchmap

Other Link： http://id.nii.ac.jp/1001/00011855/

Language：Japanese   Publisher：Kyushu University  

Internet e-mail facility is explosively getting more and more popular today. Moreover, because of growing wireless communication facilities and handy mobile computers, we can have easy e-mail access from anywhere. In case one is away from his office and using wireless line, the network bandwidth would be narrower than usual. Therefore, we need to reduce the size of e-mails. In this paper, we propose a mechanism to reduce the size of e-mails using WWW and it's applications. Also, we describe our experimental implementation of the mechanism and show the effect of it. We found that the overhead of our implementation is at most 0.01 seconds for each e-mail. We also show that our mechanism can prevent fetch time of each e-mail from rising more than 10 seconds at 9600bps.

DOI： 10.15017/1515804

CiNii Article

CiNii Books

J-GLOBAL

researchmap

Other Link： http://hdl.handle.net/2324/1515804

Language：Japanese   Publisher：The Institute of Electronics, Information and Communication Engineers  

CiNii Article

CiNii Books

researchmap

Language：Japanese   Publisher：The Institute of Electronics, Information and Communication Engineers  

CiNii Article

CiNii Books

researchmap

Other Link： http://hdl.handle.net/10228/4539

Publishing type：Research paper (international conference proceedings)   Publisher：IEEE Computer Society  

DOI： 10.1109/APSEC.1999.809635

researchmap

Other Link： https://dblp.uni-trier.de/db/conf/apsec/apsec1999.html#NonakaHNCU99

Publishing type：Research paper (international conference proceedings)   Publisher：IEEE Computer Society  

DOI： 10.1109/APSEC.1995.496974

researchmap

Other Link： https://dblp.uni-trier.de/db/conf/apsec/apsec1995.html#KasaharaNKCU95

Language：Japanese  

CiNii Article

J-GLOBAL

researchmap

Language：English   Publishing type：Research paper (international conference proceedings)   Publisher：INT ACADEMIC PUBL  

A dependence-based program representation has many applications in various software development activities, and therefore, it can be used as a unified internal representation of various target programs in an integrated software development environment. This paper presents an overview of a general-purpose system to compile target programs written in various programming languages including C, Pascal, Ada, and Occam 2 into their dependence-based representations.

Web of Science

Scopus

researchmap

J-GLOBAL

researchmap

J-GLOBAL

researchmap

Language：Japanese  

HPCで使われるジョブスケジューラSlurmは，GANG Scheduling機能を有している．その実装は，ジョブ内の全てのプロセスにSIGSTOP/SIGCONTを送ることで実現されており，頻繁なジョブの切替を想定していない．本稿では，MPIプログラムをターゲットとし，MPI profiling interfaceを利用してジョブの性質や途中状態をジョブスケジューラに通知する機構を提案する．この機構によりジョブ切替に伴うオーバヘッドを推定したり，ジョブ切替のタイミングを見極めたりすることが可能になるため，既存システムに比べてより細粒度なスケジューリングの実現を期待できる．

researchmap

Language：Japanese  

マイクロサービスといった分散システム内においてプロセス間の結合を疎にするためにメッセージブローカがしばしば用いられ，その性能はシステム全体の性能に大きく影響する．また，実装のしやすさと実行速度はしばしばトレードオフの関係にあるため，どちらを優先するべきかは条件によって異なる．そのため今回はC言語とRuby をもちいてスループットと使用するネットワーク帯域を測定し，各実装において必要となるネットワーク環境を評価する．

CiNii Books

researchmap

Language：Japanese  

スマートフォンのフラッシュライトを利用した可視光通信は，計算機と接続するためのケーブルの用意やWi-Fi等の接続が不要である点から手間の少ない通信が可能であるが，フラッシュライトの点滅制御が不正確で，通信精度が高くないという問題がある．そこで本稿では，スマートフォンのフラッシュライトの特性について調査し，通信プロトコルの改善を行う．具体的には，フラッシュライトの点滅を受光素子に当てた際に発生する電圧を測定し，送信したい点滅と実際に受け取った点滅のズレについて調査した．その結果，点滅動作中の消灯時間のズレが大きいことが分かったため，データの復号に消灯時間を利用しない通信プロトコルを提案する．

CiNii Books

researchmap

J-GLOBAL

researchmap

J-GLOBAL

researchmap

J-GLOBAL

researchmap

J-GLOBAL

researchmap

J-GLOBAL

researchmap

J-GLOBAL

researchmap

J-GLOBAL

researchmap

J-GLOBAL

researchmap

J-GLOBAL

researchmap

J-GLOBAL

researchmap

J-GLOBAL

researchmap

Language：Japanese  

researchmap

Language：Japanese  

researchmap

Language：Japanese  

researchmap

Language：Japanese  

researchmap

Language：Japanese  

researchmap

Language：Japanese  

researchmap

Language：Japanese  

researchmap

Language：Japanese  

researchmap

Language：Japanese  

researchmap

Language：Japanese  

researchmap

researchmap

J-GLOBAL

researchmap

researchmap

J-GLOBAL

researchmap

J-GLOBAL

researchmap

J-GLOBAL

researchmap

J-GLOBAL

researchmap

J-GLOBAL

researchmap

J-GLOBAL

researchmap

J-GLOBAL

researchmap

J-GLOBAL

researchmap

J-GLOBAL

researchmap

J-GLOBAL

researchmap

J-GLOBAL

researchmap

J-GLOBAL

researchmap

J-GLOBAL

researchmap

J-GLOBAL

researchmap

J-GLOBAL

researchmap

J-GLOBAL

researchmap

J-GLOBAL

researchmap

J-GLOBAL

researchmap

J-GLOBAL

researchmap

J-GLOBAL

researchmap

Language：Japanese  

CiNii Article

researchmap

J-GLOBAL

researchmap

J-GLOBAL

researchmap

J-GLOBAL

researchmap

J-GLOBAL

researchmap

Language：Japanese   Publisher：電子情報通信学会  

CiNii Article

J-GLOBAL

researchmap

Language：Japanese  

利用者が優先したい処理の実行処理時間を短縮する方式として，ディレクトリ優先方式を提案した．ディレクトリ優先方式は，入出力バッファを優先領域とその他の領域に分割し，指定したディレクトリ下のファイルを優先的にキャッシュする．しかし，優先的にキャッシュするファイルが入出力バッファを圧迫し，それらのファイル以外のファイルのキャッシュヒット率が低下することにより，計算機全体の性能低下を招いてしまう問題がある．そこで，一定期間内に参照されていないバッファ数に着目し，入出力バッファを分割する方式を提案した．提案方式では，優先的にキャッシュされたファイルをキャッシュする領域の下限と減少量を設定することで，優先ファイルのキャッシュヒット率を低下しすぎないようにする．本稿では，提案方式について，Webサーバを使用した評価結果を報告する．

CiNii Article

CiNii Books

J-GLOBAL

researchmap

Language：Japanese  

オフィスにおいて，多くの作業は繰返し発生する．このような作業では，毎回同様のファイルが必要となり，多くの場合，過去の作業で作成したファイルの再利用が行われる．しかし，過去の作業で作成したファイルについてユーザの記憶が曖昧な場合，再利用するファイルの探索に手間がかかる．そこで，作業が繰返し発生することに着目し，ファイルの再利用をとらえて履歴として蓄積し，再利用を支援する手法を提案する．具体的には，再利用の履歴を時系列順にユーザに提示し，履歴を選択して再利用を実行できるようにする．これにより，1度行った再利用であれば，ファイルの探索に手間をかけることなく再実行できる．

CiNii Article

CiNii Books

J-GLOBAL

researchmap

J-GLOBAL

researchmap

J-GLOBAL

researchmap

J-GLOBAL

researchmap

J-GLOBAL

researchmap

J-GLOBAL

researchmap

J-GLOBAL

researchmap

J-GLOBAL

researchmap

J-GLOBAL

researchmap

Language：Japanese  

Since the e-mail system is a long-standing technology, it is highly interoperable and widely used as a basic communication tool between organizations. When transmitting confidential information such as personal information by e-mail, it is common in Japan to send an encrypted attachment-file and send the common key password separately in plaintext. In this paper, we name this as eZIP mail and point out the problem of e-mail attached with password. We also propose a method to solve the problem on the recipients side. Our proposed method efficiently finds the password mail corresponding to the encrypted mail and the password string contained therein and automates password decryption. Furthermore, we implemented the proposed method and evaluate the result of incorporating it into the personal e-mail system. In the evaluation, it shows that it is possible to decrypt many of mails with several password trials.

CiNii Article

CiNii Books

J-GLOBAL

researchmap

Language：Japanese   Publisher：電子情報通信学会  

CiNii Article

J-GLOBAL

researchmap

Language：Japanese   Publisher：電子情報通信学会  

CiNii Article

J-GLOBAL

researchmap

Language：Japanese  

近年，ゲームのプレイ動画を配信するサービスが活発になっており，鑑賞の対象や情報共有の場として利用されている．しかし，動画はゲームの詳細な状態を記録していないので，共有できる情報が限定的である．そこで，ゲームの途中状態を動画とは別に記録することを考える．これにより，以下のユーザ体験が期待できる．（1）プレイ動画視聴中に，ゲームを途中状態から操作することで，有名なプレイヤのプレイを追体験できる．（2）ゲーム状況の配信が容易になる．そこで，我々は，あるユーザがプレイしているゲームの途中状態を複製・共有することで別のユーザが対象のゲームをその続きから操作できるシステムを提案する．本稿では，提案システムの設計について述べる．

CiNii Article

CiNii Books

J-GLOBAL

researchmap

J-GLOBAL

researchmap

J-GLOBAL

researchmap

Language：Japanese   Publisher：電子情報通信学会  

CiNii Article

J-GLOBAL

researchmap

Language：Japanese   Publisher：一般社団法人情報処理学会  

我々は，これまでの研究で過去のカレンダ情報に基づいて周期的に発生する作業から将来の作業の発生を予測する手法を提案した．しかし，この手法による予測結果には予測の妥当性を示す指標が含まれておらず，予測の妥当性を利用者が吟味し，判断する必要がある．そこで，予測結果に対する信頼性の指標を導入することで，利用者の予測結果の妥当性に対する判断を支援する．本稿では，作業予測の結果に対する信頼性評価手法を提案し，評価した．

CiNii Article

CiNii Books

J-GLOBAL

researchmap

Language：Japanese   Publisher：一般社団法人情報処理学会  

オフィス環境において,業務の担当者が代わる際,業務の引継ぎに先立って前任者は業務に関する情報を収集する.このとき,業務で利用した複数のアプリケーションプログラムには業務に関する情報が収集されている.本稿では,実際の業務を例に挙げ,業務に関する情報がどのように複数のアプリケーションプログラムに収集されるのかについて説明した.また,複数のアプリケーションプログラムから業務に関する情報を収集する際の問題を明らかにした.1つは,業務や作業単位での情報収集が困難な問題である.もう1つは,業務に関する情報の内どれを引き継ぐべきかわからない問題である.これらの問題に対処するために,業務や作業単位で情報を管理し,情報の重要度を管理する作業履歴の保存方式を提案した.さらに,提案方式を実現したDesktop Bookmarkから業務に関する情報が収集できることを示した.

CiNii Article

CiNii Books

J-GLOBAL

researchmap

Language：Japanese   Publisher：一般社団法人情報処理学会  

特定のサービスに適した独自 OS を利用し,さらにマルチコアプロセッサを活かして独自 OS からの既存 OS プログラム利用による独自 OS の利便性向上を目指した Linux と AnT オペレーティングシステムの混載システムを提案した.Linux は,多種の入出力機器の利用を可能とし,かつ既存 OS として普及しており,多くのサービスを実現している.一方,AnT は,マイクロカーネル構造を有し,適応性と堅牢性を特徴とする独自 OS である.この両者の特徴を生かしたサービス提供のために,両 OS の連携機構を設計した.この連携機構は,AnT から Linux にシステムコール代行実行を処理依頼する.本稿では,AnT からの Linux システムコール代行実行の評価について報告する.具体的には,連携機構実現における工数と Linux システムコール代行実行の性能について述べる.

CiNii Article

CiNii Books

J-GLOBAL

researchmap

Language：Japanese   Publisher：一般社団法人情報処理学会  

計算機資源を効率的に利用するため，1 台の計算機上で複数の OS を走行させる方式が研究されている．仮想計算機方式では，OS 間で 1つの I/O デバイスを共有可能だが，ハイパーバイザによる処理負荷により I/O 性能が低下する．一方，Mint オペレーティングシステムは，仮想化によらず I/O デバイスを分割占有することで，性能低下を抑制できる．しかし，OS と I/O デバイスの関係は，起動時に静的に固定されるため，I/O デバイスの接続形態の変更のしやすさは仮想計算機方式に及ばない．そこで本稿では，移譲方式，つまり Mint において I/O デバイスを占有する OS を動作中に切替える方式を提案する．具体的には，I/O デバイスの 1 つである NIC を対象に，Mint において NIC を占有する OS を動作中に切り替える方式について，割込ルーティングの変更による移譲方式を提案し，Loadable Kernel Module を利用した移譲方式と比較する．

CiNii Article

CiNii Books

J-GLOBAL

researchmap

J-GLOBAL

researchmap

J-GLOBAL

researchmap

J-GLOBAL

researchmap

Language：Japanese   Publisher：一般社団法人情報処理学会  

ソフトウェア開発において，ソーシャルコーディングと呼ばれる手法が広がりつつある．ソーシャルコーディングでは，プロジェクトに対する理解度の低い会員が，過去に議論された内容と重複する無益な提案をすることがある．このような無益な提案はプロジェクトオーナーの手間を増やし，プロジェクトの進行の妨げになる．こうしたプロジェクトオーナーに対する負担を軽減するため，プロジェクトにとって有益な提案を抽出し，提示することは有用であると考えられる．本稿では，ソーシャルコーディングサービスである GitHub を対象として，著名なプロジェクト内で有益な提案が占める割合を調査した．調査の結果，プロジェクト内の有益な提案が占める割合は 50％ を上回っていないことがわかった．また，有益な提案の特徴について考察し，GitHub のシステムから機械的に有益な提案を取得する方式を示した．提案方式に基づいて GitHub 上のプロジェクトの提案数と有益な提案数の推移について考察した．考察の結果，有益な提案が占める割合は，開発初期をのぞけばプロジェクト全体を通してほぼ一定していることがわかった．

CiNii Article

CiNii Books

J-GLOBAL

researchmap

Language：Japanese   Publisher：一般社団法人情報処理学会  

我々は，これまでに，整理されたカレンダ情報とは何かについて論じ，これを扱うモデルを提案した．整理されたカレンダ情報は，将来の計画立案や仕事の引継ぎに有用である．従来のシステムでは，インタフェースとして，紙のカレンダを模した形態をとっているため，これによるカレンダ情報の整理は難しい．本稿では，まず，情報整理の際のカレンダインタフェースの問題について述べる．次に，このカレンダインタフェースの問題に対処する方法として，Inbox による整理手法を提案する．Inbox による整理手法では，発生日時に依存しない方法で予定を一覧表示したり，整理されていない予定のみを表示したりすることで，カレンダ情報の整理を支援できることを示す．

CiNii Article

CiNii Books

researchmap

Language：Japanese   Publisher：一般社団法人情報処理学会  

我々は，これまでに，整理されたカレンダ情報とは何かについて論じ，これを扱うモデルを提案した．整理されたカレンダ情報は，将来の計画立案や仕事の引継ぎに有用である．従来のシステムでは，インタフェースとして，紙のカレンダを模した形態をとっているため，これによるカレンダ情報の整理は難しい．本稿では，まず，情報整理の際のカレンダインタフェースの問題について述べる．次に，このカレンダインタフェースの問題に対処する方法として，Inbox による整理手法を提案する．Inbox による整理手法では，発生日時に依存しない方法で予定を一覧表示したり，整理されていない予定のみを表示したりすることで，カレンダ情報の整理を支援できることを示す．

CiNii Article

CiNii Books

researchmap

Language：Japanese   Publisher：一般社団法人情報処理学会  

我々は，これまでに，整理されたカレンダ情報とは何かについて論じ，これを扱うモデルを提案した．整理されたカレンダ情報は，将来の計画立案や仕事の引継ぎに有用である．従来のシステムでは，インタフェースとして，紙のカレンダを模した形態をとっているため，これによるカレンダ情報の整理は難しい．本稿では，まず，情報整理の際のカレンダインタフェースの問題について述べる．次に，このカレンダインタフェースの問題に対処する方法として，Inbox による整理手法を提案する．Inbox による整理手法では，発生日時に依存しない方法で予定を一覧表示したり，整理されていない予定のみを表示したりすることで，カレンダ情報の整理を支援できることを示す．

CiNii Article

CiNii Books

J-GLOBAL

researchmap

Language：Japanese   Publisher：一般社団法人情報処理学会  

計算機の性能の向上にともない，１ 台の計算機上で複数の OS を動作させる方式が研究されている．これらの研究において，複数の種類の OS を動作させることで，それぞれの OS が持つ特性を同時に利用できる．しかし，OS 混載のために仮想計算機方式を用いた場合，OS 間で依存関係が発生する．そこで，1 台の計算機上で Linux ベースの OS を複数混載する方式を提案する．本提案方式では，Linux の機能に制限をかけることなく動作可能である．本稿では，Linux ベースの OS を混載する場合に，改変を加える必要がある部分を明らかにする．また，本提案方式を実現するために必要な改変量は小さいことを示す．最後に本提案方式の実例として，Linux と Android の混載と 32/64bit Linux の混載について述べる．

CiNii Article

CiNii Books

J-GLOBAL

researchmap

Language：Japanese   Publisher：一般社団法人電子情報通信学会  

個人のカレンダに含まれる情報は,他者と共有することで,より有益に活用できる.しかし,個人のカレンダにはプライベートな予定情報が含まれるため,単にすべての情報を共有することは好ましくない.既存カレンダシステムのカレンダ共有の仕組みでは,カレンダに含まれる情報と共有したいグループの組み合わせに対応させ,目的に応じて予定情報を共有することは難しい.例えば,個人のカレンダの17時以前の予定情報のみを職場で共有し,17時以降の予定情報のみを家族で共有するといったことができない.そこで,目的に応じた予定情報の共有を表現するモデルとして仮想カレンダを提案する.また,仮想カレンダを既存カレンダシステムと親和性の高い方式で実現するための設計と処理の流れを述べる.

CiNii Article

CiNii Books

J-GLOBAL

researchmap

Language：Japanese  

個人のカレンダに含まれる情報は，他者と共有することで，より有益に活用できる．しかし，個人のカレンダにはプライベートな予定情報が含まれるため，単にすべての情報を共有することは好ましくない．既存カレンダシステムのカレンダ共有の仕組みでは，カレンダに含まれる情報と共有したいグループの組み合わせに対応させ，目的に応じて予定情報を共有することは難しい．例えば，個人のカレンダの 17 時以前の予定情報のみを職場で共有し，17 時以降の予定情報のみを家族で共有するといったことができない．そこで，目的に応じた予定情報の共有を表現するモデルとして仮想カレンダを提案する．また，仮想カレンダを既存カレンダシステムと親和性の高い方式で実現するための設計と処理の流れを述べる．

CiNii Article

CiNii Books

researchmap

Language：Japanese   Publisher：一般社団法人情報処理学会  

我々は，1 台のマルチコア計算機で複数 Linux を独立に走行させる方式の OS として，Mint オペレーティングシステムを開発している．本稿では，Mint で走行する各 Linux に仮想計算機方式の KVM を導入し，1 台の計算機上で複数のハイパーバイザを走行させる方式について述べる．これにより，ハイパーバイザ間の処理負荷の影響を軽減しつつ，各ハイパーバイザ上の OS 間では分割された資源を共有して利用できる．また，ハイパーバイザの障害時の影響を限定でき，Mint 上で Linux 以外のカーネルを走行可能にする．基本評価では，Mint 上の KVM と通常の Linux の KVM の性能が同等であることを示した．また，KVM を用いて複数の OS を同時走行させる場合，提案方式を利用して KVM 上で動作させる OS を分散させることで処理負荷の影響が軽減できることを示した．具体的には，I/O 処理の処理負荷の影響を限定でき，レジスタ操作処理どデータ複写処理の性能は通常の Linux で動作する KVM と同等であることを示した．

CiNii Article

CiNii Books

J-GLOBAL

researchmap

Language：Japanese  

CiNii Article

CiNii Books

J-GLOBAL

researchmap

J-GLOBAL

researchmap

J-GLOBAL

researchmap

Language：Japanese  

CiNii Article

CiNii Books

J-GLOBAL

researchmap

Language：Japanese   Publisher：The Institute of Electronics, Information and Communication Engineers  

CiNii Article

CiNii Books

J-GLOBAL

researchmap

Language：Japanese   Publisher：The Institute of Electronics, Information and Communication Engineers  

CiNii Article

CiNii Books

J-GLOBAL

researchmap

Language：Japanese   Publisher：The Institute of Electronics, Information and Communication Engineers  

CiNii Article

CiNii Books

J-GLOBAL

researchmap

Language：Japanese   Publisher：The Institute of Electronics, Information and Communication Engineers  

CiNii Article

CiNii Books

J-GLOBAL

researchmap

Language：Japanese   Publisher：The Institute of Electronics, Information and Communication Engineers  

CiNii Article

CiNii Books

J-GLOBAL

researchmap

J-GLOBAL

researchmap

Language：Japanese  

CiNii Article

CiNii Books

J-GLOBAL

researchmap

Language：Japanese   Publisher：The Institute of Electronics, Information and Communication Engineers  

As wide spreading of 64-bit CPUs, Operating System (OS) is changing its architecture-base from 32-bit to 64-bit. Software running under 64-bit OSes can use the vast address space, and can process a large amount of data at one time, which can improve the performance of system significantly. However, 64-bit OSes are not compatible with some software written for 32-bit OSes. Especially, system software including kernels and device drivers written for 32-bit OSes can't operate on 64-bit OSes. We propose a method for running simultaneously 32-bit kernels and 64-bit kernels on a multi-core processor by using Mint operating system. Using this method, users can use software written for 32-bit OS and the address space of 64-bit OS as the whole system. In this paper, we report the design, implementation, and some typical use of this method.

CiNii Article

CiNii Books

J-GLOBAL

researchmap

Language：Japanese   Publisher：一般社団法人情報処理学会  

次世代ネットワークNGN/IMSが提供するXDMSは,携帯電話のコンタクトリストやIM(Instant Messaging)の通信記録など,比較的小さなデータチャンクを扱うためのデータベースである.XDMSにおけるデータは,その種別に応じた単独のAS(Application Server)のみに操作されることが通例のため,XDMS自身は排他制御などの複雑な制御機能を提供していない.また,アクセスプロトコルとしてHTMLベースとしたXCAPを採用しているため操作が単純で分かりやすい反面,AS がリッチなサービスを提供する場合には,AS開発の工数が大きい.そこで,本稿では,ASがXDMSをより高水準なインタフェースで扱えるようにするためのフレームワークを提案する.

CiNii Article

CiNii Books

J-GLOBAL

researchmap

Language：Japanese   Publisher：The Institute of Electronics, Information and Communication Engineers  

We develop the Mint operating system that have the function of running multiple OSes simultaneously on a multi-core processor. In the Mint, first OS boots other OSes, and Mint needs a kernel image each OSes. In other words, multiple kernel images are necessary. In this paper, we propose a method for booting later running OS by using Kexec, and we describe that multiple kernel images of Mint can be unified. In addition, we evaluate the code change quantity and the time for booting OS about this method for booting.

CiNii Article

CiNii Books

J-GLOBAL

researchmap

Language：Japanese   Publisher：The Institute of Electronics, Information and Communication Engineers  

The computer is providing various services as the computer becomes high performance. The method that run in two or more OS on one computer is researched activity to execute various services flexibly. In these methods, a virtual machine is kwon widely. In a virtual machine method, one or more OSes can run on a hardware by running on a virtual computer. However, each OSes influence other OS and it is low independence between each OSes in a virtual machine method. In this paper, we proposes the "Mint" operating system that run independently two or more Linux by using the multi-core processor on one computer. Moreover, we report the performance when two or more OS run.

CiNii Article

CiNii Books

J-GLOBAL

researchmap

Language：Japanese   Publisher：The Institute of Electronics, Information and Communication Engineers  

In chip multi-processor based on the continuation concept, the hardware scheduler controls threads and achieves the high performance on thread-scheduling. However, the execution of each thread can not consider priority to threads because the hardware scheduler schedules threads in FIFO manner. Therefore, when multiple services execute simultaneously, the execution of each service can not consider priority to service. It needs the software support to control the execution of each service. We proposed the software scheduler to support the hardware scheduler. In this paper, we evaluate the software scheduler, which targets various parallel level services.

CiNii Article

CiNii Books

J-GLOBAL

researchmap

Language：Japanese  

CiNii Article

J-GLOBAL

researchmap

Language：Japanese   Publisher：Forum on Information Technology  

CiNii Article

CiNii Books

J-GLOBAL

researchmap

Language：Japanese   Publisher：Forum on Information Technology  

CiNii Article

CiNii Books

J-GLOBAL

researchmap

Language：Japanese   Publisher：The Institute of Electronics, Information and Communication Engineers  

Recent OS has a large number of device drivers. It is important to reduce cost for developing new device drivers in OS development. When we port an existing device driver into the OS which realizing device driver process to reduce the development cost, we need to not only port but also form it into process. In this paper, we show a method for converting an existing Linux LKM device driver into process in AnT operating system. Furthermore, we evaluated the proposal method using drivers such as the FD, Intel 1GbE, and Realtek 1GbE driver.

CiNii Article

CiNii Books

researchmap

Language：Japanese   Publisher：The Institute of Electronics, Information and Communication Engineers  

Recent OS has a large number of device drivers. It is important to reduce cost for developing new device drivers in OS development. When we port an existing device driver into the OS which realizing device driver process to reduce the development cost, we need to not only port but also form it into process. In this paper, we show a method for converting an existing Linux LKM device driver into process in AnT operating system. Furthermore, we evaluated the proposal method using drivers such as the FD, Intel 1GbE, and Realtek 1GbE driver.

CiNii Article

CiNii Books

J-GLOBAL

researchmap

Language：Japanese  

CiNii Article

CiNii Books

J-GLOBAL

researchmap

Language：Japanese   Publisher：IEEE Computer Society  

DOI： 10.1109/BWCCA.2010.157

CiNii Article

CiNii Books

researchmap

Other Link： https://dblp.uni-trier.de/db/conf/bwcca/bwcca2010.html#FujiwaraNT10

Language：Japanese   Publisher：The Institute of Electronics, Information and Communication Engineers  

Fuce processor is a variant of CMP processor, which has multiple thread execution cores. The computation model of the Fuce is built on the basis of continuation concept, and its thread-execution unit is realized by hardware logic. Hereby, Fuce achieves the high performance on thread-scheduling. By utilizing this characteristic, Fuce is capable of dealing with quite many simultaneous threads in nature. However, the thread execution unit of the Fuce implemented by hardware logic is too simple to control the precise priority of individual thread. Therefore, Fuse needs some additional software components on the top of its hardware scheduler. In this paper, we have implemented a CMP-oriented software thread-scheduler on Fuce, and evaluate the performance of it.

CiNii Article

CiNii Books

J-GLOBAL

researchmap

Language：Japanese   Publisher：The Institute of Electronics, Information and Communication Engineers  

Due to the increasing of illegal attacks from outside, network monitoring component is one of the largest concern to the operating system construction. Since the monitoring works as a system-lifelong component, the overheads of the monitoring should be minimum. However, one of the most popular monitoring technique using standard Virtual Machine (VM) has handsome overheads. This paper proposes a new method for packet monitoring using TwinOS, which alleviates the overheads of the VM based systems.

CiNii Article

CiNii Books

J-GLOBAL

researchmap

Language：Japanese   Publisher：The Institute of Electronics, Information and Communication Engineers  

CiNii Article

CiNii Books

J-GLOBAL

researchmap

Language：Japanese   Publisher：The Institute of Electronics, Information and Communication Engineers  

CiNii Article

CiNii Books

J-GLOBAL

researchmap

Language：Japanese   Publisher：The Institute of Electronics, Information and Communication Engineers  

CiNii Article

CiNii Books

J-GLOBAL

researchmap

J-GLOBAL

researchmap

Language：Japanese   Publisher：Forum on Information Technology  

CiNii Article

CiNii Books

J-GLOBAL

researchmap

Language：Japanese   Publisher：Information Processing Society of Japan (IPSJ)  

In a fine grain multi-thread environment based on continuation model, H/W scheduler takes dependency among threads as continuation order. In such environment, execution of service requires scheduling of many thread-groups by S/W. In addition, reduction of scheduling overhead is important so that a large number of fine-grain threads repeat generation and extinction regularly. In this paper, we focus on continuation that is dependency among threads, and show effective S/W scheduling method for reduction of scheduling overhead based on continuation flow.

CiNii Article

CiNii Books

J-GLOBAL

researchmap

Language：Japanese   Publisher：Information Processing Society of Japan (IPSJ)  

Recently, illegal attacks to OSs are increasing. A lot of intruder detection method have been proposed. In this paper, we propose a method for monitoring communication using TwinOS. TwinOS runs two Linux on a single PC, and has a virtual NIC that provides one Linux with the other Linux's NIC. Our proposed method monitors one Linux's transmission from the other Linux via virtual NIC, and keep intruders away from log files. In addition, as basic evaluation, evaluated influence on communication performance by the implementation of the monitor.

CiNii Article

CiNii Books

J-GLOBAL

researchmap

Language：Japanese   Publisher：The Institute of Electronics, Information and Communication Engineers  

Recently, operating systems have become increasingly multi-functional by the providing of various devices. Its functions are provided as driver programs. Even though average users do not use all functions, they do not customize and shrink their operating systems because of its job difficulty. Accordingly, they usually use wasteful and fat operating systems: much memory, slow down bootup time. In this paper, we propose an effective method for controlling invocation timing of device driver programs. As a strategy for the control, we use operation history on device drivers. And we present this design on AnT operating system.

CiNii Article

CiNii Books

researchmap

Language：Japanese   Publisher：The Institute of Electronics, Information and Communication Engineers  

Recently, operating systems have become increasingly multi-functional by the providing of various devices. Its functions are provided as driver programs. Even though average users do not use all functions, they do not customize and shrink their operating systems because of its job difficulty. Accordingly, they usually use wasteful and fat operating systems: much memory, slow down bootup time. In this paper, we propose an effective method for controlling invocation timing of device driver programs. As a strategy for the control, we use operation history on device drivers. And we present this design on AnT operating system.

CiNii Article

CiNii Books

J-GLOBAL

researchmap

Language：Japanese   Publisher：The Institute of Electronics, Information and Communication Engineers  

CiNii Article

CiNii Books

J-GLOBAL

researchmap

Language：Japanese  

CiNii Article

J-GLOBAL

researchmap

Language：Japanese   Publisher：Information Processing Society of Japan (IPSJ)  

Recently, operating systems have become increasingly multi-functional by the providing of various devices. Its functions are provided as driver programs. Even though average users do not use all functions, they do not customize and shrink their operating systems because of its job difficulty. Accordingly, they usually use wasteful and fat operating systems: much memory, slow down bootup time. In this paper, we propose an effective method for controlling invocation timing of device driver programs. As a strategy for the control, we use operation history on device drivers. And we present this design on AnT operating system.

CiNii Article

CiNii Books

J-GLOBAL

researchmap

Language：Japanese   Publisher：Information Processing Society of Japan (IPSJ)  

Most operating systems manage a buffer cache for I/O processing by a block unit. On the other hand, application programs handle a data from a viewpoint of files. I/O buffer cache mechanism considered this difference is effective to execute I/O processing efficiently because I/O processing is slower than CPU processing. In this paper, we describe I/O buffer cache mechanism that decides blocks protected based on use frequency of file of application programs.

CiNii Article

CiNii Books

J-GLOBAL

researchmap

J-GLOBAL

researchmap

J-GLOBAL

researchmap

J-GLOBAL

researchmap

Language：Japanese   Publisher：The Institute of Electronics, Information and Communication Engineers  

As an application of TwinOS system, we are developing a network packet filtering scheme. one sibling OS of the TwinOS provides a pseudo NIC hardware for the other OS and performs packet filtering in the middle of it. To improve the performance of the packet filtering scheme, we propose a method for CPU evacuation of idle OS and several possible efficient methods for packet disposal.

CiNii Article

CiNii Books

J-GLOBAL

researchmap

Language：Japanese   Publisher：Information Processing Society of Japan (IPSJ)  

Microprocessors based on the fine grain multithread architecture are characterized by its thread level parallelism. It essentially generates a large number of threads in a second. Therefore, it is very difficult to schedule all the threads effectively by a conventional software process scheduler. In this paper, we propose a cost-effective process scheduling method suitable for the fine grain multithread architecture based on actual thread behavior.

CiNii Article

CiNii Books

J-GLOBAL

researchmap

Language：Japanese   Publisher：Information Processing Society of Japan (IPSJ)  

Run-time updating of software component is a key technology for non-stop systems. We have already proposed a method for updating running user processes, however that method could not be applied to operating systems. In this paper, we propose a new method for run-time updating of network device driver, which provides context migration from an old driver to updated one. This paper also describes requirements for our method and a basic design built on the FreeBSD operating system.

CiNii Article

CiNii Books

J-GLOBAL

researchmap

Language：Japanese   Publisher：The Institute of Electronics, Information and Communication Engineers  

Run-time updating of software component is a key technology for non-stop systems. We have already proposed a method for updating running user processes, however that method could not be applied to operating systems. In this paper, we propose a new method for run-time updating of network device driver, which provides context migration from an old driver to updated one. This paper also describes requirements for our method and a basic design built on the FreeBSD operating system.

CiNii Article

CiNii Books

researchmap

Language：Japanese   Publisher：The Institute of Electronics, Information and Communication Engineers  

CiNii Article

CiNii Books

J-GLOBAL

researchmap

Language：Japanese  

CiNii Article

CiNii Books

J-GLOBAL

researchmap

Language：Japanese  

CiNii Article

CiNii Books

J-GLOBAL

researchmap

Language：Japanese  

CiNii Article

CiNii Books

J-GLOBAL

researchmap

Language：Japanese   Publisher：Information Processing Society of Japan (IPSJ)  

We proposed the composite filter which mixes up existing spam filters, and evaluated its basic performance. As a method for reflecting user's characteristic to the composite filter's judgment, we describe the method for setting a weight to each spam filter's judgment. The weight is based on classification log of spam filter and judgment of user. In this paper, we describe the evaluation of the composite filter applied the weighting method in actual case. The composite filter had no misclassification of legitimate email, and classified 43.07% of spam. Then the composite filter obtained the best spam detection rate using classification log of the past 24th.

CiNii Article

CiNii Books

J-GLOBAL

researchmap

Language：Japanese   Publisher：Information Processing Society of Japan (IPSJ)  

AnT is an operating system based on micro-kernel architecture. Filesystem, device driver and other operating system modules run as a process named external-core. AnT can adapt various kind of computer environment by changing set of external-core. In this paper, we present process creation mechanism for AnT to make AnT adapt computer environment and create process need by AnT. And we present efficient mechanism of process control.

CiNii Article

CiNii Books

J-GLOBAL

researchmap

Language：Japanese   Publisher：Information Processing Society of Japan (IPSJ)  

AnT is an operating system based on micro-kernel architecture. Therefore, AnT can provide various kinds of service. However, there is a problem about a data transferring overhead caused by a lot of inter-process communications. Therefore, we need to examine memory management so that AnT can transfer data with low overhead. In this paper, we present design, implementation and evaluation of memory management. Resulting from the design and the implementation, AnT can realize memory protection from each program, high speed inter-process communication and high speed rebooting. We evaluate data transferring and rebooting.

CiNii Article

CiNii Books

J-GLOBAL

researchmap

Language：Japanese   Publisher：Forum on Information Technology  

CiNii Article

CiNii Books

J-GLOBAL

researchmap

Language：Japanese   Publisher：Information Processing Society of Japan (IPSJ)  

Zero-Wait Thread is a continuation-based thread and supposed to run to completion without suspention. We have implemented handler routines for multiple I/O requests based on Zero-Wait Threads for the Fuce processor. The Fuce processor supports a continuation model which is a varient of dataflow computing model. Handling I/O requests causes high latencies in operating systems, implementation based on Zero-Wait Thread allows performance improvement and effective in threads parallel execution. This paper explains our impelementation and shows evaluation for effectiveness of Zero-Wait Thread on the Fuce processor.

CiNii Article

CiNii Books

J-GLOBAL

researchmap

Language：Japanese   Publisher：Information Processing Society of Japan (IPSJ)  

Recently, we are supposed to manage information about all PCs in every office, such as administrator, location, mac-address, list of installed softwares... However, it is difficult to do that strictly, some sites give up the management and a DHCP server supplies a free-to-connect network inside the LAN. Besides security risks, the situation is ugly. That is, we have to be accountable for compliance with laws: information leakage protection, software license management. In this paper, we describe a system to help changing the situation gradually. We introduce IP address based classify of users using DHCP server. It is naturally acceptable to free-to-connect network. This IP address based restriction gradually lead users to manage their PCs by themselves.

CiNii Article

CiNii Books

J-GLOBAL

researchmap

Language：Japanese   Publisher：Information Processing Society of Japan (IPSJ)  

Recentl, illegal attacks to OSs are increasing. A lot of intruder detection method have been proposed. In this paper, we propose a method for monitoring communication using TwinOS, TwinOS runs two Linux on a single PC, and has a virtual NIC that provides one Linux with the other Linux's NIC. Our proposed method monitors one Linux's communication from the other Linux via virtual NIC, and keep log from intruders.

CiNii Article

CiNii Books

J-GLOBAL

researchmap

Language：Japanese  

CiNii Article

CiNii Books

J-GLOBAL

researchmap

Language：Japanese  

CiNii Article

CiNii Books

J-GLOBAL

researchmap

Language：Japanese   Publisher：Information Processing Society of Japan (IPSJ)  

Average operating systems manage disk I/O cache by a block unit with LRU method. On the other hand, Application programs (APs) handle data from a viewpoint of file unit. To fill up the gap, we propose a directory oriented buffer cache mechanism that reflects on the demands of APs. Our mechanism is based on a thought that a disk access pattern occurred from a series of job will be settled in a corresponding particular directory. We give a high priority to the important directory, which is associated with an important job. This paper describes an implementation and an evaluation of our method.

CiNii Article

CiNii Books

J-GLOBAL

researchmap

Language：Japanese   Publisher：Information Processing Society of Japan (IPSJ)  

Conventional operating systems used to implement device drivers as kernel modules or embedded objects. Therefore, maturity of a device driver influences the reliability of a whole system. In this paper, we propose a method for constructing isolated device drivers in the form of user process. Our proposed method uses the running mode switch mechanism of application programs. User process device drivers enhance the reliability of the operating system. We evaluated the basic overhead of our method.

CiNii Article

CiNii Books

J-GLOBAL

researchmap

Language：Japanese   Publisher：Information Processing Society of Japan (IPSJ)  

In a fine grain multi-thread environment, operating systems are required to control innumerable threads efficiently. As a solution for the thread scheduling overhead problem, it is a good idea to control threads by injecting extra delays to their inter-thread synchronization. In the manner of this idea, we have to consider the optimal frequency and pattern of the delay injection. In this paper, we consider and evaluate some patterns of the delay injection.

CiNii Article

CiNii Books

J-GLOBAL

researchmap

Language：Japanese  

CiNii Article

J-GLOBAL

researchmap

Language：Japanese   Publisher：Forum on Information Technology  

CiNii Article

CiNii Books

J-GLOBAL

researchmap

J-GLOBAL

researchmap

J-GLOBAL

researchmap

J-GLOBAL

researchmap

Language：Japanese   Publisher：Information Processing Society of Japan (IPSJ)  

There is a remarkable thing for progress of a microprocessor and input-output hardware. Furthermore, improvement of transmission speed of a channel is remarkable, too. In addition, a computer is necessary in various scenes, and service classification to offer increases drastically. It is demanded that base software can use a function and performance of these hardware effectively. Therefore it was started development the AnT operating system (An operating system with adaptability and toughness) to have both adaptability and solidity. This article explains a design policy of the AnT operating system and a characteristic function.

CiNii Article

CiNii Books

J-GLOBAL

researchmap

Language：Japanese   Publisher：Forum on Information Technology  

CiNii Article

CiNii Books

J-GLOBAL

researchmap

J-GLOBAL

researchmap

J-GLOBAL

researchmap

J-GLOBAL

researchmap

J-GLOBAL

researchmap

J-GLOBAL

researchmap

J-GLOBAL

researchmap

Language：Japanese   Publisher：Information Processing Society of Japan (IPSJ)  

System-call is a high cost operation because it causes a running mode switch of a process. This paper presents an implementation and evaluation of a dynamic running mode switch mechanism of application programs. Our proposed mechanism has a characteristic of on-demand and on-the-fly mode switching of a process from user mode to supervisor mode and vice versa. While a process is in the supervisor mode, it is able to emit a system-call as a corresponding simple function call. Thus, it eliminates the overhead of a conventional trap-based system-call procedure. The evaluation shows that our implementation runs 800-900 clocks faster than a conventional system-call based implementation does.

CiNii Article

CiNii Books

J-GLOBAL

researchmap

Language：Japanese   Publisher：Information Processing Society of Japan (IPSJ)  

Recently, leakage of classified information such as personal information has become serious problem. To prevent the leakage, it is important to know where the information diffuses in the computer environment. In this paper, we propose the tracing method of the classified information diffusion. The tracing starts when a process invokes the open system call for the file, and checks if whole or a part of the classified information may be copied to the opening file. We describe the algorithm of the method, Implementation of the method on the Linux kernel, and the overhead to the system calls.

CiNii Article

CiNii Books

J-GLOBAL

researchmap

Language：Japanese   Publisher：Information Processing Society of Japan (IPSJ)  

In this paper, we propose an operating system architecture using continuation-based zero-wait threads. A thread is supposed to run to completion without suspension, and execution control between threads are performed based on continuations attached to threads. This approach enables us to reduce thread management overhead and to realize flexible scheduling schemes. We discuss the operating system architecture and examines its effectiveness in handling I/O requests.

CiNii Article

CiNii Books

J-GLOBAL

researchmap

Language：Japanese  

CiNii Article

CiNii Books

J-GLOBAL

researchmap

Language：Japanese  

CiNii Article

CiNii Books

J-GLOBAL

researchmap

Language：Japanese  

CiNii Article

CiNii Books

J-GLOBAL

researchmap

Language：Japanese  

CiNii Article

CiNii Books

J-GLOBAL

researchmap

Language：Japanese   Publisher：Information Processing Society of Japan (IPSJ)  

Groupware tools have enabled us to share information about location of each member. It would be useful for visitors to display this information on the door of our office. However, it is difficult to place a PC-style groupware terminal in front of the door for cost and appearance reason. Also, we will have a security issue if we open the bare groupware information to the visitors. In this paper, I propose a smart doorplate which solves these problems.

CiNii Article

CiNii Books

J-GLOBAL

researchmap

Language：Japanese  

CiNii Article

J-GLOBAL

researchmap

Language：Japanese   Publisher：Information Processing Society of Japan (IPSJ)  

A large number of threads run concurrently in the multithreading environments such as the processors with thread-level parallel execution mechanism. In those environments, the execution unit which a processor supports may differ from the schedule unit which operating system manages. Especially, for the machine which has the mechanism to control parallel non-preemptive event-driven thread execution, a new operating system mechanism has to be developed to control program execution with less overhead. 0ne of the key issues of this operating system mechanism is how to schedule the conventional threads or processes which are constructed with non-preemptive event-driven threads. In this paper, we propose a program execution controlling method, which covers the waiting for an event, based on continuity and character of program processing in multithread execution mechanism.

CiNii Article

CiNii Books

J-GLOBAL

researchmap

Language：Japanese   Publisher：Information Processing Society of Japan (IPSJ)  

Microprocessors based on the fine grain multithread architecture are characterized by its thread level parallelism. lt essentially generates a large number of threads in a second. Therefore, it is very difficult to schedule all the threads effectively by a conventional software process scheduler. For this reason, these microprocessors have a sort of embedded hardware thread control mechanism. However, if operating system does not pay any cost for the scheduling, system would be uncontrollable. In this paper, we propose a cost-effective process scheduling method suitable for the fine grain multithread architecture.

CiNii Article

CiNii Books

J-GLOBAL

researchmap

Language：Japanese   Publisher：Information Processing Society of Japan (IPSJ)  

In recent PC hardware's develop leads supplying of various services. These various services sometimes need to be performed in a PC, then OS has be customized in the existing way. Existing method like virtual machine, enables multiple OS's runs on a single PC and solves the problem. But, these method diminish. I/O performance of each OS. We propose a DualOS method that two different types of OSes can run on a single PC for providing two different types of service or function. And we apply our method on two linuxes and name it 「NINJA」.We evaluate NINJA's performance and mention its features.

CiNii Article

CiNii Books

J-GLOBAL

researchmap

Language：Japanese  

CiNii Article

CiNii Books

J-GLOBAL

researchmap

Language：Japanese   Publisher：Information Processing Society of Japan (IPSJ)  

Multiple operating system has become popular Major existing method is that secondary operating systems run on a primary operating system This relation between primary operating system and others causes some problem such as an accidental stop of primary operating system leads to the others We proposed a new framework that operating systems can run independently and simultaneously on a single processor In basic concept of this framework, coexistent operating systems don't share hardwares for their independency But for better availability, it is desirable that they can share hardwares This paper describes a method for sharing hardware resource among the coexistent operating system and its implementation focused on PCI hardware In this method, each operating system occupies some hardwares but can provide arbitrary hardware resource to the others safely

CiNii Article

CiNii Books

J-GLOBAL

researchmap

Language：Japanese   Publisher：Information Processing Society of Japan (IPSJ)  

We have been developing and implementing a prototype for running multiple operating systems on a single computer. This method improves the virtual machine method's problem that has dependence of child operating systems on their mother VM. This paper describes how to boot up multiple operating systems and how to switch operating systems from one to another. Under this method, physical memory is exclusively divided and reserved for each operating system from beginning and the trigger of switching is hardware interrupts which have no dependence on any operating system.

CiNii Article

CiNii Books

J-GLOBAL

researchmap

Language：Japanese  

CiNii Article

CiNii Books

J-GLOBAL

researchmap

Language：Japanese  

CiNii Article

CiNii Books

J-GLOBAL

researchmap

Language：Japanese   Publisher：九州大学情報基盤センター  

DOI： 10.15017/1470657

CiNii Article

CiNii Books

researchmap

Language：Japanese   Publisher：Information Processing Society of Japan (IPSJ)  

Due to the domestic growth of average PCs. virtual machine software for running multiple operating systems on a single system has become popular, today. Existing visual machine softwares are built on the similar concept of the conventional VM, which has dependence of child operating systems on their mother VM. This fact causes some problems such as performance limitation and collective security. In other works, accidental stop of the mother operating system leads to the corruption of the whole family. We have been research on the new formwork for solving these problems. As a key technology of the formwork, This paper describes a method for switching operating system from one to another safely. Under the method, the trigger of switching is a hardware interrput which has no dependence on any operating system. Also we propose 3 method for sharing hardware resource among the coexistent operating systems.

CiNii Article

CiNii Books

J-GLOBAL

researchmap

Language：Japanese   Publisher：Information Processing Society of Japan (IPSJ)  

We haw been developing a method for running multiple operating systems on a single computer for utilizing their individual application programs effectively. This paper describes how to boot up multiple operating systems safely in succession by sharing physical memory among them. Under the method physical memory is exclusively divided and reserved for each operating system from beginning. Each operating system is controlled to boot up using their own memory region without having any bad effect on other operating systems.

CiNii Article

CiNii Books

J-GLOBAL

researchmap

J-GLOBAL

researchmap

J-GLOBAL

researchmap

J-GLOBAL

researchmap

J-GLOBAL

researchmap

J-GLOBAL

researchmap

J-GLOBAL

researchmap

Language：Japanese   Publisher：Information Processing Society of Japan (IPSJ)  

We usually produce many documents during software development, which are supposed to be referred in the maintenance phase. Reusability of the documents is important. Therefore, an electronic document management system is developed. In this paper, we describe the new electronic document management system which has an e-mail interface for document registration. Thereby, we can manage many types of documents effectively such as design, specifications, minutes or discussions in meetings. In this system, user can easily register his documents to the system. All he have to do is simply sending an e-mail which has the documents as attachments. Therefore, the time and effort of registration decreases very much, and can share information effectively.

CiNii Article

CiNii Books

J-GLOBAL

researchmap

Language：Japanese   Publisher：Information Processing Society of Japan (IPSJ)  

Personal computer (PC) is one of the fastest growing specialty in industry. Various kinds of I/O devices, application software packages (AP) and OSs are available under PC. Therefore, sometimes, we have a problem to find a good combination of devices, APs and an OS. To solve the problem, we need to run multiple OSs at the same time. In this paper, we describe a new technique to run multiple OSs in one box. We also illustrate some applications of this new technique. Our technique has three strong points - (1) no need to modify original OSs, (2) good I/O performance as bare OS, (3) no parent-child relationship among all OSs.

CiNii Article

CiNii Books

J-GLOBAL

researchmap

Language：Japanese  

CiNii Article

CiNii Books

J-GLOBAL

researchmap

J-GLOBAL

researchmap

J-GLOBAL

researchmap

J-GLOBAL

researchmap

Language：Japanese  

CiNii Article

CiNii Books

J-GLOBAL

researchmap

Language：Japanese  

CiNii Article

CiNii Books

J-GLOBAL

researchmap

Language：Japanese  

CiNii Article

CiNii Books

J-GLOBAL

researchmap

Language：Japanese   Publisher：Information Processing Society of Japan (IPSJ)  

Internet e-mail facility is explosively getting more and more popular today. The size of e-mails has also been growing, and attachment of e-mail influences the size. Meanwhile, we often use a handy mobile computer to fetch e-mails from a mail server through a narrow wireless communication line. In case of using wireless telephone line, it takes a long time to fetch e-mails because the network bandwidth would be narrower than usual. In this paper, we propose a mechanism to reduce the attachment size of e-mails using Web and it's applications. Also, we describe our experimental implementation of the mechanism and show the effect of it. We found that the overhead of our implementation is at most 1 seconds for each e-mail. We also show that our mechanism can make e-mail fetch time 1/16 shorter.

CiNii Article

CiNii Books

J-GLOBAL

researchmap

J-GLOBAL

researchmap

J-GLOBAL

researchmap

J-GLOBAL

researchmap

Language：Japanese  

CiNii Article

CiNii Books

researchmap

Language：Japanese  

CiNii Article

CiNii Books

J-GLOBAL

researchmap

Language：Japanese   Publisher：公益社団法人 日本工学教育協会  

DOI： 10.20549/jseejaarc.1998.0_111

CiNii Article

researchmap

Language：Japanese  

CiNii Article

CiNii Books

J-GLOBAL

researchmap

Language：Japanese  

CiNii Article

CiNii Books

J-GLOBAL

researchmap

Language：Japanese   Presentation type：Oral presentation (general)  

researchmap

Language：Japanese   Presentation type：Oral presentation (general)  

researchmap

Language：Japanese   Presentation type：Oral presentation (general)  

researchmap

Language：Japanese   Presentation type：Oral presentation (general)  

researchmap

Language：Japanese   Presentation type：Oral presentation (general)  

researchmap

Language：Japanese   Presentation type：Oral presentation (general)  

researchmap

Language：Japanese   Presentation type：Oral presentation (general)  

researchmap

Language：Japanese   Presentation type：Oral presentation (general)  

researchmap

Language：Japanese   Presentation type：Oral presentation (general)  

researchmap

Language：Japanese   Presentation type：Oral presentation (general)  

researchmap

Language：Japanese   Presentation type：Oral presentation (general)  

researchmap

Language：Japanese   Presentation type：Oral presentation (general)  

researchmap

Language：Japanese   Presentation type：Oral presentation (general)  

researchmap

Language：Japanese   Presentation type：Oral presentation (general)  

researchmap

Language：Japanese   Presentation type：Oral presentation (general)  

researchmap

Language：Japanese   Presentation type：Oral presentation (general)  

researchmap

Language：Japanese   Presentation type：Oral presentation (general)  

researchmap

Language：Japanese   Presentation type：Oral presentation (general)  

researchmap

Language：Japanese   Presentation type：Oral presentation (general)  

researchmap

Language：Japanese   Presentation type：Oral presentation (general)  

researchmap

Language：Japanese   Presentation type：Oral presentation (general)  

researchmap

Language：Japanese   Presentation type：Oral presentation (general)  

researchmap

Language：Japanese   Presentation type：Oral presentation (general)  

researchmap

Language：Japanese   Presentation type：Oral presentation (general)  

researchmap

Language：Japanese   Presentation type：Oral presentation (general)  

researchmap

Language：Japanese   Presentation type：Oral presentation (general)  

researchmap

Language：Japanese   Presentation type：Oral presentation (general)  

researchmap

Language：Japanese   Presentation type：Oral presentation (general)  

researchmap