Workshop/Sessions - Technology
2005.1.25 (tue) ~ 2005.1.27 (thur)                                                                                                 <updated on 2005.2.1>

Medical Network BoF  | Medical Teleconference  |
End to end performance monitoring and improving BoF  | Network Performance Measurement and Monitoring
HDTV Session              | Grid Workshop           | Global Collaboration           |  NGI Applications Workshop       | G-H Workshop    
Multimedia Session     | Lambda BoF               | BoF on SIP Networking       | BoF on H323 Networking            | Network Engineering and NOC Session

Medical Network BoF

Time: 2005.1.26 (wed)
Place: Room D
(original proposal)



Organizing chairs:
Shuji Shimizu (Kyushu University Hospital, Japan)
We have established a medical network system with high-quality moving image on a super broad-banded Internet between Korea and Japan since February 2003. Teleconference or demonstration of live surgery using DVTS was revolutionary in medical fields. As our activity is known at various places, there have been strong demands to join our project from many countries. In order to promote our system to whole Asia-Pacific regions, it is essential for both medical and engineering people get together on site and discuss our next strategy. Medical staffs need to talk on institutions to be connected and their interesting contents on coming teleconferenes. Engineers are expected to give their practical comments to medical team on possibility of infrastructural as well as technical issues. This system is unquestionably useful for exchanging views in medicine. The purpose of this session is to share the opinions to work systemically to expand the current project.
Session Plan:
1.       Introduction of APAN: Mizushima H /Tokyo NCC
2.       Current broadband network in AP: Kitamura Y/NICT
3.       Situation in Korea-Japan: Telemedical activity in KJ: (on network) Nakashima N /Kyushu U Hospital, Japan
4.       Situation in Thailand: Continuous Medical Education Exchanges between the Royal Thai- US Armies via Videoteleconference;
                  from ISDN to Internet2 to IronGrid, COL Suwicha Tim Chitpatima, Ph.D. (Genetic Engineering), Phramongkutklao Hospital, Bangkok
5.       Situation in Taiwan: Robot-assisted laparoscopic surgery with Zeus system: preliminary results and technical aspects Hurng-Sheng
                  Wu, Show-Chwan Memorial Hospital, Taiwan
6.        Discussion
1)     Discussion on Charter
2)      Regional network: Thai/Taiwan/
3)      Contact person; matching with local engineer
4)      Homepage/ including event record/mailing list
5)      Event plans
6)      Collect slides


International medical teleconference between

Time: 2005.1.26 (wed)
Place: Room D

(original proposal)


Organizing Chairs:
Syuji Shimizu (Kyushu University Hospital)
Naoki Nakashima (Kyushu University Hospital)

We have established a medical network in Asia-Pacific area with high-quality moving image on a super broad-banded internet line since February 2003. Broadcasting real-time surgery and teleconference with medical-quality videos with Digital Video Transport System (DVTS) were so useful to learn surgical techniques and other medical procedures beyond borders. We have conducted 22 events in categories of teleconference (8 times), live transmission of medical procedure (7 times), and remote attendance to scientific medical meetings (7 times) until October 2004. We had teleconferences among 3 countries in APAN-Hawaii in January 2004 (Japan, Korea and USA), and in APAN-Cairns in July 2004 (Japan, Korea and Australia). There have been no serious transmission/ ethical problems so far. We always have, at least, one trial as of network technique and content in every event. With these experiences and current achievements, we hope to extend this advanced network system to whole Asia-Pacific countries to share medical skills and knowledge.

Session Plan:
             Beijing (Dr Dong M), Seoul (Dr Han HS),
             Fukuoka (Dr Nakashima N), Bangkok (Dr Shimizu S)
1.       Opening remarks (Dr Mizushima H)
2.       Message and introduction from each station
3.       Video presentation of Endoscopic Surgery (Seoul/Beijing)
      Dr. Ho-Seong Han, Seoul National University Bundang Hospital, Seoul
      Dr Lin Chen, General Hospital of People’s Liberation Army, Beijing
4.       Hematology (Thai/Kyushu)
 Regulatory Role of Apoptosis in Erythropoiesis   ( Koichiro Muta, M.D., Kyushu University, Fukuoka
2) Thalassemia; from genetic background to phenotypic diversity ( Suthat Fucharoen, M.D., Mahidol University, Bangkok )
5. Closing remarks (Thailand)


 End to end performance monitoring and improving BoF

Time: 2005.1.25 (tue)
Place: Room D

This BoF is the pre-meeting of the session of Network Performance Measurement and Monitoring



Network Performance Measurement and Monitoring:  creating a global fabric across APAN, the US and Europe to support high-performance global network collaboration

Time: 2005.1.27 (thur)
Place: Room B

Organizing Chair:
Heather Boyles, Internet2 

Session description:
Performing regular analysis of the end-to-end path across domains creates a normal operational mode where network operators and end-users can easily determine network capabilities and restrictions.  In turn, this will significantly improve the likelihood that advanced Internet applications operate at peak performance and thereby advance the productivity of researchers.  The Internet2 End-to-End Performance Initiative (E2Epi) is establishing a performance measurement infrastructure across and beyond university campuses and labs.  International collaboration has already generated testing points in Europe, Asia and Latin America allowing for performance monitoring across multiple, international domains.  In Europe, the GEANT2 project is deploying similar performance measurement and monitoring infrastructures.  And in Asia, APAN and KREONET have deployed measurement points compatible with the Internet2 Performance Initiative Performance Environment Software (piPEs). 

The latest piPEs-related project developments will be presented, including alpha release of the piPEs visualization package and alpha release of the piPEs Performance Measurement Domain software package. In addition, case studies will be explored, suggesting ways in which the measurement infrastructure can be used.  Special emphasis will be on American/European collaboration to create measurement frameworks for transatlantic partial path analysis as well as multi-domain measurement activities and inter-connectivity with a number of countries in Asia and Latin America.  Speakers from each of these regions will talk about their deployments and ways in which they are working together to make these deployments inter-operable. 

1. Masaki Hirabaru:"Advanced TCP Performance Measurement with BWCTL"
It is important to measure TCP performance before transferring huge
data between long distant places over Internet. Combining BWCTL with
Web100 enables us to examine detailed TCP dynamic behavior and helps
us to identify the bottleneck of performance. A case study of BWCTL
deployment in e-VLBI as well as Tokyo NOC will be presented.

2. Minki Noh: E2E performance measurement In KREONET
Within the context of KREONET (National Research Network of Rep. Korea) we will take a more in depth look at
existing measurement programs (AMP, Netflow, NDT, Network performance tester, etc.) Special attention will be given to the players involved, how the services are used, and the operation policies in place.

3. Matt Zekauskas: Internet2 Activities Toward a Global Measurement Infrastructure
Monitoring performance across multiple, international domains is made possible through the ongoing collaborative efforts of Internet2, GEANT2 JRA1, GGF NMWG, and other groups. Recent achievements by these groups and the latest piPEs-related project developments will be presented with an eye towards implementation and local benefits.

4. Nicolas Simar: GN2-JRA1 Performance Monitoring [Remote Presentation]

Matt Zekauskas, Internet2 End to End Performance Initiative, USA
Masaki Hirabaru, NICT, Japan
Minki Noh, KISTI, Korea
Nicolas Simar, DANTE, Europe (invited, TBC)


HDTV & Access Grid Session
Time: 2005.1.25 (tue)
Place: Room F
(original proposal)

JongWon Kim (GIST)
#1. HDTV & Access Grid session 1
This AG and HDTV session is designed to provide attendees with an opportunity to discuss the latest development on the enhanced media support for AG (Access Grid). In this session, several on-going efforts on the high-quality media services for AG will be presented and discussed.
- AGPager -- Advanced Display Management (speaker: Tokyo Institute of Technology Hiroyuki Komatsu)
Abstract : The Access Grid uses many windows such as video windows, presentation slides, and control panels. To control those windows, an operator has to move, resize and iconify each windows separately at the present. AGPager provides an easy window manipulation GUI as well as
an automatic window layout system under a specified layout rule. This talk will present the usage of AGPager, the outline of the implementation, and the goal.

 - MPEG4 codec for Access Grid (speaker: NCHC Barz Hsu)
Abstract : Access Grid are solutions most directly targeted to large and complecollaborations. Video and audio communications are doubtless the most components in Access Grid. The mbone tools used in Access Gird includes "vic" for video and "rat" for audio. However, the video codecs inside vic are somewhat out-of-date and lack of exhaustive optimization. For these reasons, we have tried to incorporate MPEG4 codec into vic
and adopt several video filters such as deinterlacing and interpolation filters to provide better visual quality. For the source of NTSC resolution (720x480) with 20 frames per second, the typical bit rate of MPEG4 codec is only 1 Mb/sec and the PSNR value is about 40 dB. Furthmore, we also paid attentation on minimizing the encoding latency and the ability of error resilience. In conclusion, modified vic indeed provides a better "sense of presence" than before and runs very efficiently. Thus, Access Grid can perform a large scale and high quality video conferenceing.

- AGconnect: Toward better connectivity for the AG (GIST Namgon Kim and JongWon Kim)
Abstract: Access Grid (AG) uses multicast to deliver multimedia data. By using multicast, AG can achieve bandwidth efficiency. But multicast is an
obstacle for users in multicast-disabled network. To solve this problem, AG provides multicast bridge for users in unicast network. Multicast bridge only treats users in unicast network. However, there are many kinds of AG-disabled users. For example, users behind firewall, users with insufficient network capacity and users with weak system performance can't use AG efficiently. Our goal is to provide connectivity to AG for these users. As one effort for this, we modified QuickBridge to support AG with High-Quality Video. In this talk, we will show the structure of modified AG to support connectivity solution for AG with High-Quality Video.

#2. HDTV & Access Grid session 2
Objective :

This HDTV session is designed to provide attendees with an opportunity to discuss the latest development around the network-based delivery of HD (high-definition) video technology. Aiming to provide immersive experience to end users with HD quality video, lots of efforts are taking place to enable HD video contents over high-speed research networks. In this session, several on-going efforts on HD video delivery will be presented and discussed.
Full rate Uncompressed HDTV Transport: Experiences and Implementation
[Remote Presentation]
  (speaker:  USC ISI Ladan Gharai and University of Glasgow Colin Perkins) Abstract:In this talk we present the architecture and implementation of our  full-rate UltraGrid nodes capable of sending and receiving up to 1.485  Gbps of uncompressed HDTV content.  Previously, UltraGrid nodes were  limited to sending and receiving (close to) 1 Gbps of sub-sampled HDTV  content due to hardware limitations. However,  the proliferation of 
PCI-X technology and the advent of 10 Gbps NICs, has removed all  hardware barriers to building full-rate UltraGrid nodes.  We discuss 
the technical challenges of transporting full-rate HDTV from the  network and end-system's point of view.

- AARNet and ResearchChannel TransPacific HD video (speaker: AARNet Andrew Howard)
Abstract: Over the past six months AARNet (Australia's Academic and Research Network) and ResearchChannel have collaborated on a number of projects to develop compressed and uncompressed High Definition video conferencing. Following the demonstration at the Cairns APAN meeting the first transpacific uncompressed HD video conference was operated for four days during the SuperComputing 2004 event between Canberra (Australia), Pittsburgh and Seattle (USA) using the SXTransport network. Compressed HD video between Australia and Hawaii was demonstraited at the PTC05 conference. The demonstrations were built from "off the shelf" hardware with software developed by ResearchChannel.

- Implementation of Low-Cost Stereo High-Definition Video Delivery  System   (speaker: GIST Kiyoung Lee and JongWon Kim)
Abstract: In GIST, we has implemented a stereoscopic high-definition video delivery system using modified VideoLAN S/W. We are using two JVC HDV cameras(JVC GR-HD1). The format of this camera is MPEG-2 TS at MP@HL and the bandwidth is 19.2Mbps from each camera. The total bandwidth of this system is 38.4Mbps with two 19.2Mbps MPEG-2 TS streams. In the sending part two streams from each camera are muxed and sent to the client. The client receives these streams, then each streams are decoded after demuxing process. Using 3D display devices end users can feel the HD stereo video. (For more detail, please contact 
Special arrangement
1) network setup for demos (local and international)
2) additional projector or PDP display (if possible HD display support) for demo
3) Polycom-based video conferencing support  


Grid WG
Grid Workshop

Time: 2005.1.25 (tue)
Place: Room F
(original proposal)

Kento Aida (Titech)
Putchong Uthayapas (KU)
Seishi Ninomiya (NARO-NARC)
Grid computing is an important emerging paradigm for organizing and sharing computing, network, storage and instrument resources in large-scale applications. The Grid workshop brings together participants from academia, industry and government to learn and to discuss about grid activities in the APAN countries. The workshop is organized by two sessions, the tutorial session of Grid middleware and the technical session of updated grid research topics.

Session Plan:
Tutorial session (chair: TBA)
  speaker and title (tentative):
  - Yoshio Tanaka (AIST)
    "Programming on the Grid using GridRPC"      Abstract

Technical session (chair: TBA)
  speakers and titles (tentatvie):
  - J.G. Wang (NARC)    
    "New Implementations of Agricultural Models Using Mediate Architecture"     Abstract
   - Jedsada Phengsuwan (NECTEC)
    "Development of Distributed MetBroker toward Information Grid"   Abstract
  - Vara Varavithya (KMITNB)
    "Thailand National Grid Project"          Abstract
  - Kiyoshi Honda (AIT)
    "Cluster Computing for SWAP Crop Model Parameter Identification using Remote Sensing"    Abstract


Global Collaboration

Time: 2005.1.26 (wed)
Session 1: 11:00-12:30
Session 2: 14:00-15:30
Place: Room C

(original proposal)

Jacqueline Brown
Provide an update on existing and emerging networking and high-end applications collaborations (such as GLORIAD, PRAGMA, GLIF, etc.)

Session Introduction:
Advanced networks have become one of the cornerstones of collaboration for research and education in the 21st century as the commercial Internet has been unable to support the bandwidth intensive and other advanced needs of those constituencies. In many ways, the technologies developed and deployed in these advanced networks are leading the way to the next generation of the commercial Internet, such as with early deployment of 10Gbps circuits, multicast deployment across autonomous systems and advances in user control of lightpaths. Examples are multiplying of cross-sector and cross-boundary collaborations in the creation of those advanced networks. The challenge will be to insure that these national and multi-national efforts interoperate seamlessly.

Session Plan:
1: Malcolm Read
Executive Secretary JISC, UK
Title: Update on United Kingdom Network Provision and International Connectivity
Abstract: The presentation will provide an update on UK network provision and international connectivity.
The changing requirements of the academic user communities (eg research, universities and colleges) will be discussed as will the need to develop international collaboration in relevant generic middleware applications such as authentication and authorization.

2: George McLaughlin
Director, International Developments, AARNet Canberra, AU
Title: Driving bandwidth-hungry applications:
SX-TransPORT - the infrastructure to support e-science
Abstract: High capacity networks have changed the way that global scientific collaborative research programs are undertaken. Unique high-end instruments (astronomical telescopes, particle accelerators) costing hundreds of millions of dollars and capable of generating terabytes of data each day, are now able to be accessed from anywhere with a very high speed connection to the facility (gigabits per second). Analysing and visualising these massive datasets can now be carried out in a different part of the world from where the data is collected. Correlating data in real time from telescopes in different parts of the world is providing major new discoveries in astronomy. Multimedia immersive environments, streaming several hundred megabits per second, coupled with haptic technology (adding the concept of touch to 3-d video and stereo audio) enable people at different part of the globe to work with the same experience as if they were in the same room.
The national and regional Research and Education (R&E) Networks form a global mesh of high capacity circuits on which we are building the next generation of advanced communications services, middleware and applications. Several demonstrators will be held in the broadband playground showing the future capabilities that will flow from this kind of infrastructure.
Southern Cross Cable Networks (SX) and AARNet (Australia's Academic and Research Network) formed an alliance to facilitate participation by Australian scientists in these new global collaborative initiatives by announcing the Southern Cross Trans-Pacific Optical Research Testbed (SX TransPORT) comprising dual 10Gbps circuits linking AARNet Australian 10Gbps network to Hawaii and the 10Gbps R&E networks in North America, Europe and the Asia Pacific.

3: Dai Davies
Title: Summary of GEANT2
Abstract: The GEANT network brought 10 Gbps networking to Europe in 2000. We are now well advanced with the implementation of its successor network GEANT2. Instead of building a yet faster network GEANT2 is moving away from traditional IP networking to offering a hybrid design. The presentation will look at the background behind the planning decisions for GEANT2, will describe progress in terms of its implementation and look at some of the issues we face.

4: David West
Title: European Outreach projects
Abstract: There are European initiatives to develop research networking with regions where it is less well developed today: Mediterranean (EUMEDCONNECT), Latin America (ALICE) and Asia (TEIN2). The presentation will briefly review the model developed to undertake these projects, and describe the progress and plans

5: Akira Kato
Title: GLIF: Global collaboration efforts on lambda networking
Abstract: GLIF is an informal collaboration to connect lambda network links provided by the members in order to provide international lambda networking for demonstration and experiments. One of the notable efforts based on GLIF was performed in Oct 2004, in which an OC-192 lightpath was formed from Geneva in Switzerland to Japan. This talk will cover how such a global collaboration was implemented as well as issues involved to the experiments

6: Baoping Yan
Title: Recent developments at the Chinese Academy of Sciences
Abstract: The speaker will address such diverse projects as GLORIAD and HK-light and well as initiatives in e-science.

7: Yoshio Tanaka
AIST Grid Technology Research Center
Title: PRAGMA: Cyberinfrastructure, Applications, People
Abstract: The Pacific Rim Application and Grid Middleware Assembly (PRAGMA) is an open, institution-based organization, founded in 2002, to establish sustained collaborations and to advance the use of grid technologies in applications among a community of investigators around the Pacific Rim. PRAGMA accomplishes its mission by conducting joint projects, sharing resources, integrating middleware, and exchanging and training researchers.
We give an overview of PRAGMA's activities, accomplishments, and structure. Furthermore we suggest area of future growth via applications and collaborations of interest to the Asia Pacific Advanced Networking participants. In particular, we will emphasis the PRAGMA experiment of making the grid usable on a routine basis, the lessons learned from the application developers, the grid middleware developers, and the network systems. Furthermore, we emphasize the role of applications in focusing development of our activities. Finally, we stress the role of human interactions to make this possible.

8: Yasuhiro Koyama [remote presentation]
Kashima Space Research Center / NICT
Co-Authors :
Yasuhiro Koyama, Tetsuro Kondo, Hiroshi Takeuchi, Masaki Hirabaru (NICT), Kazuhiro Takashima (Geographical Survey Institute), Jason SooHoo, Kevin Dudevoir, Alan Whitney (Haystack Observatory), Steven Tingay, Craig West (Swinburne Univ. Tech.), An Tao (Shanghai Observatory), and Jouko Ritakari (Metsahovi Radio Observatory).
Title : e-VLBI with the High Speed International Research Networks
Abstract: Developments of the Very Long Baseline Interferometry (VLBI) technology had been practically realized along with the improvements of the data recording technology using magnetic tapes. In the conventional VLBI
observations, radio signals from celestial radio sources are received at two or more radio telescopes and are recorded to magnetic tapes.
The tapes are then shipped to a correlation site and processed to calculate cross correlation functions between data taken at different radio telescopes. In this observation scheme, the sensitivity and the timeliness to obtain results from observations are limited by the maximum speed of data recording on magnetic tapes and the slow delivery of the tapes from the observing sites to the correlation site. To overcome these limitations, efforts have been made to transfer the VLBI data from observing sites to the correlation site electrically. The observed data are transferred either in real-time if the connecting network is fast enough to transfer the observed data rate, or in near real-time if any kinds of data buffering is required at the observing site. In either case, the technique is called as e-VLBI to distinguish it from the traditional tape-based VLBI.
Many radio telescopes around the glove are regularly participating in the various VLBI sessions for research in geodesy, astrometry, and astronomy by using the tape-based systems. If we can connect these observatories with high-speed research networks and the e-VLBI observing systems, it will dramatically improve the timeliness to obtain results from observations. Especially, the value of UT1 can only be determined by geodetic VLBI observations and the routine e-VLBI observation is desired by various research fields which require accurate satellite orbits or rely on accurate terrestrial and celestial reference frames. In addition, if the data transfer rate becomes faster than the maximum recording speed on magnetic tapes, the sensitivity of the VLBI observation will be dramatically enhanced and it will expand the horizon of the radio astronomical observations. In recent years, we have been performing various e-VLBI demonstration experiments and routine sessions using the international high speed research networks organized under the APAN. Initial results of these activities and future plans will be reported in the presentation

9: Ted Hanss
University of Michigan Medical School
Title: Global Medical Education Research Network
What if collaboration around medical education could occur without regard for time or location? Complementing this morning's presentation focused on hospital-based clinicians cooperating through videoconferencing, a group of medical schools have been discussing plans to develop a Global Medical Education Research Network. This is not a "virtual medical school". Rather,
given the global issues surrounding human health, can we establish a collaboration environment for both medical school faculty and students cooperating with colleagues around the world? The infrastructure being discussed will support everything from students working on team projects to formal seminars with speakers and audience members participating from their
home sites.
The technology to produce this event exists and is in use on our campuses.However, it will take work to package the technology together in a way that it is not itself the focus of attention----such that the ease-of-use allows the participants to focus on the content of the presentations. We are currently distributing a scenario document, which will be reviewed in the
presentation. This document is available at <>. In addition, technical staff have been establishing plans for trial uses of technology to demonstrate the capabilities to faculty and students. The intention of this presentation is to recruit a larger group of participants in this emerging


NGI Applications Workshop
Time: 2005.1.26 (wed)
Place: Room E

(original proposal)

Joonbok Lee (KAIST, Korea) Koji OKAMURA (Kyushu Univ., Japan)
For several years ago, we can not have enough bandwidth networks but recently the bandwidth grows quickly then we can not use its full performance. During we suffered narrow band networks, we have developed compression technologies , cache technologies etc but sacrificed delay time. Now a day, if we make good use for high broadband networks such as 10G, we must change the strategies, we may not need compression or cache technologies longer. In this workshop, we have presentations such the applications for the NGI. And these broad-bandization will comes to every economics soon or eventually but quickly. Then this topics is not interesting in by only economics who already get NGI but also every economics.  

Session Plan:
+ SC2004 Reports
  Koji OKAMURA (Kyushu Univ.,Japan)

+ Korean NGI Applications over Asia
  BH Seo (KISDI, Japan)

+ Japanse NGI Applications over Asia
  IGARASHI Kiyoshi (NICT, Japan)

+ Chinese NGI Applications over Asia


G-H Workshop

Time: 2005.1.26 (wed)
Place: Room E

(original proposal)

Koji OKAMURA (Kyushu Univ., Japan)
We have used the Gigabit Link between Japan and Korea for two years and can have many useful and future oriented experimentations using this link. Now we are planning to extend link capacity and project. This workshop will have many presentation based on such advanced experimentations using G-H link and have discussion about our future. 

Session Plan:
+ Reports of the Applications over G-H
 + e-lerning
 + Medical
    + Shuji Shimizu (Kyushu Univ.,Japan)
    + NCC
+ Updates of the G-H link
 + IGARASHI Kiyoshi (NICT, Japan)
 + BH Seo (KISDI, Japan) 


Multimedia WG
Multimedia Networking

Time: 2005.1.26 (wed)
Place: Room B

(original proposal)

Dr. Omar Amer Abouabdalla 
- Invite speakers to give presentation on Multimedia Networking
- Exchange updates and related information regarding to the Multimedia WG

Speakers :
1. Mr. Salah AL-Naemi  -  from University Science Malaysia
Reliable File Transfer Method Using Data Size as a Sequence Indicator
Abstract: With block transfers, a file is transferred in fixed chunks size made up of a header portion (transmission sequence number and sometimes with a block length) followed by a data portion, these sequence numbers are used by the sender and the receiver sides to coordinate and to recover from loss or duplicate chunks and sometime for reordering purpose as well. This paper proposes a reliable file transfer method using data size as a sequence indicator, our proposed method also allows for more efficient use of the available bandwidth, as we are able to remove the sequence number portion that is currently used within most algorithms like FTP.
2. Dr. David Ngo -  from Multmedia University in Malaysia (not confirmed yet)


Lambda Networking BoF

Time: 2005.1.26 (wed)
Place: Room C

(original proposal)

Akira Kato 
Exchange updates and related information regarding to the Lambda networking. Possible agenda will include, lambda networking in KR/TW/JP/..., IEEAF, brief report of GLIF workshop, Tokyo--CERN OC-192 experiment, etc.  

Session Plan:

Activities in T-LEX, Akira Kato WIDE
TWAREN/TAIWANLight : Research and Development, Te-Lung Liu


BoF on SIP Networking in APAN

Time: 2005.1.26 (wed)
Place: Room D

(original proposal)

Stephen Kingham
Update by delegates on present SIP deployments, Open Discussion on peering SIP Networks.  

Session Plan:
Speakers -
Quincy Wu, Francis Lee, Stephen Kingham, others


BoF on H323 Networking in APAN

Time: 2005.1.26 (wed)
Place: Room D

(original proposal)

Stephen Kingham
Update on H.323 in Region, discuss peering, and issues raised by Kewin's Tutorial.

Session Plan:
Topic- H.323 NAT/firewall traversal A chance for APAN?
Speaker- Kewin Stoeckigt / Max Plank Institute


Network Engineering and NOC Sessions
Time: 2005.1.27 (thur)
Place: Room C

Kazunori Konishi (APAN-JP), Chris Myers (AARNet), Lee Jaehwa (ANF)  
This session will encourage the collaborations among the engineers of applications, middleware and network.
Session Introduction:
This session will focuses on the advanced network infrastructure, including measurement, security and end systems. The experiments at SC2004 Bandwidth Challenge and the lessons learned will be discussed for further development of advanced network infrastructure. The status and plan of APAN member networks will also be shared for further collaborations.

Session Plan:

1.  Doug Pearson <>         REN-ISAC: Network Information and the Security
The REN-ISAC (Research and Education Networking Information Sharing and Analysis Center) is an integral part of the U.S. higher education strategy to improve network security through information collection, analysis, dissemination, timely warning, and response, specifically designed to support the unique environment and needs of organizations connected to served higher education and research networks. Supported by Indiana University and through relationship with EDUCAUSE and Internet2, the REN-ISAC supports efforts to protect cyber infrastructure by participating in the formal framework of U.S. ISACs. This presentation will describe REN-ISAC activities, with an aim to explore opportunities for cooperation with Asia-Pacific research and education networks.

2.  Greg Wickham <>   Keeping Network Monitoring Current using Automated Nagios Configurations
A significant challenge managing a large and growing network is ensuring that all network devices are appropriately monitored. If attempts are not made to keep the monitoring current then it is possible that major events on the network will go unnoticed.
To solve this problem a proof of concept was undertaken where the important attributes of the network devices were stored in an XML file. This file was then used as the key configuration repository for both validating the status / configuration of the network and generating Nagios configuration data.
The presentation will provide an overview of the architecture that has been developed and highlight with comparisons to the process of managing Nagios configurations manually.

3.  Bruce Morgan <>   Backbone Measurement and Security
Network security is important, especially when the capability of Research and Education networks reaches multi gigabit throughput. There are numerous tools available to monitor activity and provide both an early warning system to infections and monitor malicious activity on the network. With the growth of malicious activity on the network it is important that we cooperate to contain that activity.

4.  John M Hicks <> & T.Ikeda      Observatory: Measurement Infrastructure
Measurement data collected from critical points along the network path provides valuable information for engineers and researcher to debug problems and improve application performance.  The Abilene Observatory is a successful program from Internet2 that supports the collection and dissemination of network data associated with the Abilene Network in the United States.  This talk will focus on the deploy of an Observatory like infrastructure in the U.S and Japan to support Transpacific measurements across TOPS-TransPAC2 and other links that connect APAN to U.S. and other network resources.  This work is supported by a collaborative effort between APAN-JP and the TOPS-TransPAC2 project from Indiana

5.  Chris Myers <>     Eduroam Australia Project experience in location independent wireless networking with international collaboration with TERENA
EduRoam allows roving researchers to log-in, with their usual "user name/password",  to a wireless networks at participating campuses around Australia and the World and gain access to resources at their home institution. The program, which is based on inter-institutional trust, originated in the Netherlands and is rapidly spreading across the UK, Europe and now Australia.
The Australian R&E community has in braced EduRoam as the preferred Method for sharing their wireless infrastructure, we will be presenting our experiences in rolling out the service and how this service my be of benefit to the wider APAN community.

6.  Kei Hiraki <> etc.      Challenge to Internet Land Speed Record at SC2004
Authors: Kei Hiraki, Mary Inaba, Makoto Nakamura, Junji Tamatsukuri, Nao Aoshima, Ryo Nishimura ,(The University of Tokyo) Akira Kato, Seiichi Yamamoto (WIDE) Ryutaro Kurusu, Masakazu Sakamoto, Yuki Furukawa, Yukichi Ikuta, (Fujitsu Computer Technologies LTD)
One of the key component to construct a single-server Data Reservoir for 10 Gbps internet is a high-speed TCP. We proposed "optimization of TCP by inter-layer coodination", and did experiments to evaluate the feasibility on actual Long Fat pipe Network at SC2004. We used about 31,248 Km 10 Gbps network from CERN to Pittsburgh, USA via Tokyo using WIDE, APAN/JGN2, IEEAF/Tyco, CA *net4, Abilene and SURFnet. As a result of the experiment, we measured 7.21 Gbps single-stream TCP data transfer, which is recognised as Internet Land Speed Record by Internet2. In the presentation, we discuss on technical and non-technical issues to realize high-speed TCP experiment at SC2004.

7.   John Hicks  <>    FTP Performance
As international circuits are migrated to speeds of 10 Gigabits per second, the distances and bandwidth available provide interesting challenges to traditional layer4 stacks. This talk will survey some of the work being done on TCP at various institutions throughout the world, with a focus on the results obtained during the Supercomputing 2004 bandwidth challenge.

8.  Osamu Tatebe <>          Belle/Gfarm Grid Experiment at SC04
This talk introduces the Gfarm Grid software, and describes Belle/Gfarm Grid experiment at SC2004. In this experiment, we demonstrated (1) online distributed data analysis, and (2) large-scale data analysis using four sites; KEK, AIST, SDSC, and SC2004 conference hall in Japan and US.  The experiment (1) would store the online raw data generated by KEKB accelerator to a Gfarm file system that is a shared file system consisting of AIST and SDSC clusters.  The stored raw data would be analyzed in parallel.  The snapshot of histogram would be displayed every 5 seconds at SC2004.  The experiment (2) would try to analyze terabyte-scale data in US using a cluster in Japan.  During this experiment, the large-scale data would be transferred from US to Japan.  We will report the performance of data transfer using the JGN2 10Gbps international network.

9.  Andrew Howard <> AARNet and ResearchChannel uncompressed High Definition video over IP experiences at
 SC04 and other ResearchChannel activities

Over the past six months AARNet (Australia's Academic and Research Network) and ResearchChannel have collaborated on a number of projects to develop and operate compressed and uncompressed High Definition video conferencing using the SXTransport, Internet 2 and NLR networks.
The SuperComputing 2004 event provided a unique opportunity to operate an uncompressed High Definition video conference between Canberra (Australia), Seattle (USA) and Pittsburgh (USA) for the duration of the conference.
This presentation provides a status update on the current AARNet and ResearchChannel collaborations and covers the steps required to construct the network, systems and production elements required to run a demonstration with a 1.4Gb bidirectional data flow.

10. Masaki Hirabaru <>  etc.    e-VLBI Deployments with Research Internet
We report our recent e-VLBI experiments done over APAN networks and collaborators. Its focus is on high-performance data tranfer in this kind of long distant and high volume science applications. Our talk also includes emerging e-VLBI collaborations like with EU (JIVE), China (Urumuqi), and Russia (Lebedev Institute).

Lunch Break

Resource Allocation Sub-Committee
Chair: Kiyoshi Igarashi (NICT)

11. Naoyuki Fujita <>       High Performance Remote File System
Abstract :
Using an HPC-server, we must manage huge data that is almost |impossible to copy to remote computer over the Internet. For example, on our organization research field: CFD, 273MB/s and 30GB/file are needed.  Even if copy operation is possible, we must manage multiple replicas to keep data coherency.  So we developed high performance remote file system named SRFS: Shared Rapid File System on Ether.  The file system already enchmarked 56-90 Mbps on Fast-Ethernet and 600 Mbps on Gigabit-Ethernet without TCP/IP tuning on Japanese domestic network.  In this presentation, I'll show the result of SC2004 BWC that used the Pacific Ocean link. 

12. Katsushi Kouyama <>   The LAC/UIC experiences through JGN2/APAN during SC04
 UDT(UDP-based Data Transport Protocol) is an application level transport protocol used in distributed data intensive applications.
This new protocol motivated by the emergence of wide area high-speed optical network, in which TCP is often found to fail to utilize the
abundant bandwidth. UDT demonstrated good efficiency and fairness characteristics in high performance computing applications where a small
number of bulk sources share theabundant bandwidth. It combines both rate and window control and uses bandwidth estimation to determine the
control parameters automatically. In this presentation, we show the outline of UDT protocol and the results of bandwidth challenge through JGN2/APAN at SC 2004.

13. Jin Tanaka <> & Chris Robb        Lessons Learned from SC2004 Bandwidth Challenge
With the growth of APAN, we ToykoXP NOC require advanced operation for supporting multiple high performance experiments. In supercomputing 2004, we were realized difficulty of high-speed network operation over inter-domains. This talk will propose the necessity for high-speed network measurement and cooperation of NOCs among international networks.

14. Chris Myers <>     GrangeNet services overview
In GrangeNet (GRid And Next GEneration Network) we have been developing value add services for our network to deploy turn key solutions to enhance R&E collaboration within e research and middleware. We will be presenting some of our projects including our Distributed Data Centre Mass Storage (DDCMS) Prototype project which uses ISCSI to deliver storage to researchers.

15. Kiyoshi Igarashi <>   Network updates by NICT
In April 2004, the National Institute of Information and Communications Technology (NICT: president: NAGAO Makoto) launched the JGN II as a succesor of the JGN. A new 10G Japan-US Link from Tokyo to Chikago(StarLight) was established on 1 August, 2004 as a part of JGN II. Japan - Korea link as APII IPv6 R&D testbed project (APEC-TEL) has been used for promoting a development of broadband application in a variety fileds, network engineering, tele-medicine, e-learning and so on. An advanced network between Japan and China is under development.

16. Xing Li <> IPv6 project updates
The presentation will provide 6noc related updates

17. Ma Yan <>                  CERNET2 updates
CERNET2 is the largest next-generation Internet network backbone which is part of the demonstration project CNGI and the only nationwide
academic network; also it is one of the world's largest native IPv6 backbones. In December 25, 2004, CERNET2 backbone, which covers 20
cities and connects about 100 universities are in operation.

18. Jaehwa Lee <>, Y.Hattori & Pensri     APAN Servers Relocation part2
APAN servers had been upgraded and relocated for supporting the new Secretariat in Thailand and stability of APAN Web service. APAN Web servers had been relocated to three locations - Thailand, Korea and Japan. The design of the relocation, updating of Web servers and the method of the Web contents synchronization will be described.

19. Jun Matsukata <>              SINET updates
Though it is planned to change the carriers of the circuits in many sections of SINET and SuperSINET by April 2005, the overall topology will remain unchanged. SINET and SuperSINET has four transpacific OC48 circuits to New York which are used to connect research and educational networks in U.S.A. and Canada as well as Europe. We are going to replace the four OC48's by a single OC192, and to install a new OC48 circuit to Los Angeles by April 2005. We are also working on the upgrade of the circuit to Thailand.

20. Chris Robb <>            HOPI deployment
With the emergence of lambda-based networks over the past year, the Internet2 HOPI project is aimed at investigating the synergies between traditional packet based networks and lambda networks. The HOPI project will be in an investigative phase over the next year as Internet2 evaluates different products and signaling paradigms. This talk will highlight the current status of the HOPI project and the future roadmap.