| Project/Area Number |
13224059
|
|---|
| Research Category |
Grant-in-Aid for Scientific Research on Priority Areas
|
| Allocation Type | Single-year Grants |
|---|
| Review Section |
Science and Engineering
|
|---|
| Research Institution | Cybermedia Center, Osaka University |
|---|
Principal Investigator |
SHIMOJO Shinji Osaka University, Cybermedia Center, Professor, サイバーメディアセンター, 教授 (00187478)
|
|---|
| Co-Investigator(Kenkyū-buntansha) |
MURATA Masayuki Osaka University, Graduate School of Information Science and Technology, Professor, 大学院・情報科学, 教授 (80200301)
NISHIHARA Kousyu Osaka University, Institute of Laser Engineering, Professor, レーザーエネルギー学研究センター, 教授 (40107131)
MORI Hirotaro Osaka University, Ultra High Voltage Electron Microscopy Center, Professor, 超高電圧電子顕微鏡センター, 教授 (10024366)
TUKAMOTO Masahiko Kobe University, Graduate School of Scince and Technology, Professor, 工学部電気電子工学科, 教授 (60273588)
AKIYAMA Toyokazu Osaka University, Cybermedia Center, Lecturer, サイバーメディアセンター, 講師 (80324862)
野崎 一徳 大阪大学, サイバーメディアセンター, 教務職員 (40379110)
|
|---|
| Project Period (FY) |
2001 – 2005
|
| Project Status |
Completed (Fiscal Year 2005)
|
| Budget Amount *help |
¥108,100,000 (Direct Cost: ¥108,100,000)
Fiscal Year 2005: ¥22,000,000 (Direct Cost: ¥22,000,000)
Fiscal Year 2004: ¥25,000,000 (Direct Cost: ¥25,000,000)
Fiscal Year 2003: ¥26,000,000 (Direct Cost: ¥26,000,000)
Fiscal Year 2002: ¥35,100,000 (Direct Cost: ¥35,100,000)
|
|---|
| Keywords | Grid / Large Scale simulation / Remote Control / Electron Microscopy / Datagrid / Data Management / Mobile Agent / metadata / 電子顕微鏡遠隔操作システム / トラヒック制御技術 / 可視化システム / スーパーコンピューター / ライフサイエンス / トラフィック制御技術 / 画像解析 / スーパーコンピュータ / 医用データ解析 / 大規模シュミレーション |
|---|
| Research Abstract |
We applied the Grid technology to large-scale simulations, remote operation of Ultra-High Voltage Electron Microscopy(UHVEM), and medical data analyses. In conducting research on these applications, we evaluated every Grid function and clarified newly required functions. We developed the technology to collaboratively use multiple supercomputers and to connect the visualization system and the observational equipment onto the Grid. We further managed data by utilizing computers distributed around the world and were building a system that allows us to use these computers as if we use a single huge database. Through this system, we conducted R＆D to discover and obtain new knowledge.
The detail of research is as follows:
1. UHVEM Remote Control
Datagrid has been established in UHVEM Center, Osaka University in order to share the image data using the remote operation applications. As UHVEM output a large datasets (32Mbyte/image x 181), a huge storage was needed in order to get volume images of
… More
specimens. We developed the original image enrollment application for 4Kx4K CCD camera control software by utilizing Storage Resource Broker(SRB) which is one of the datagrid middleware. Finally, we could have the tools to integrate facilities related with UHVEM, and also have remote control tools.
2. Ultra High-Speed Network Traffic Control
Nowadays, Large datasets such as UHVEM image datasets are distributed in world wide and are transferred between remote sites,. As TCP/IP is not suitable for long fat pipe, GridFTP was developed to realize the large datasets transfer. However, it requires to adjust number of parallel TCP connection. Hence, we proposed the efficient method of optimizing the number of connections, which realized effective large datasets transfer.
3. Large Scale Numerical Simulation on Grid
On the ultra high-speed networks, we have tired to perform large scale numerical simulations by using grid middleware. In this trial, we've focused on optimization of a molecular dynamics simulation for grid environment. Most important point of this trial, several related simulation codes could be integrated by this middleware and portal.
4. Autonomous Optimization for Resource Distribution
We propose a new model applying a mobile agent concept to Grid resource administration. Successfully, we could propose a platform technology for Grid applications, which works coordinate resource distribution on grid networks by mobile agents that works as an application. Less
|
