Computer Animation
August, 1999

1. Principal Investigator: Gary B. Wilcox

2. Associated Investigators: John Leckenby, Francesca Talenti, Wayne Danielson

3. Technical Support Contact: Charles Soto

4. Project Abstract:

   The Animation/Interactive project will investigate new applications of computer animation in a multimedia Web environment in terms of both content creation and consumption. Following the trajectory of the price-performance curve, computer animation projects can increasingly migrate to more widely available Intel platforms. The Web provides an ideal environment for the delivery of computer generated animation as part of larger entertainment, advertising or educational content.

5. Equipment Status:

   The Digital Interactive Studio has the following equipment installed:

   • 9 Dual Pentium II/300, 512MB RAM, 4GB HD, Iomega Jaz & ZIP, 20" color monitor, flatbed color scanner, 6x9" Wacom drawing tablet, speakers
   • 1 Single Pentium II/300, 256MB RAM, 2GB HD, Iomega Jaz & ZIP, 20" color monitor, flatbed color scanner, 6x9" Wacom drawing tablet, speakers
   • One system has an Intel Smart Video Recorder III video capture card and an 800x600 dpi NTSC scan converter. This allows the digitizing and taping of media to/from an SVHS video tape recorder.
   • One system is attached to a 1024x768 dpi color LCD projector for demonstration purposes.
   • Two Quad-Pentium Pro/200 servers have been set up to support the lab. They provide basic NT file services, as well as HTTP and FTP services. One system is to be configured as a test system (with access restricted to the lab workstation IP range), while the other is for actual production/web content serving.

6. Research Progress:

   The arrival of the equipment the week before classes presented several technical challenges for the tech team as well and the investigators. Most of these problems were solved throughout the semester, but the workstations will be re-built this summer.

   In the area of content creation, Professor Talenti worked with 18 upper-division students in the course Digital Animation during the Spring. Skills developed were: working knowledge of 2-D and 3-D animation; concepts of movement, design, key frames, cycles, layering, lighting, editing, and sound. Programs used: Macromedia Director 6.0 for 2-D animation; Infini-D 4.0 for 3-D animation. Students each completed one flip-book; one 2-D cycle; one 2-D animation using a scanned image; one 3-D tutorial; and a final project, with sound.

   In the area of consumption, Professor Leckenby's Interactive Advertising Graduate Seminar focused on the following five goals:

   • To Develop an Understanding of IMPORTANT Issues in Interactive Advertising.
   • To Get a Sense of What Other "Thinkers" are Thinking about Interactive Advertising.
   • To Develop Ideas about How New Media Fit with Old Media in Advertising.
   • To Achieve Understanding of the Socio-Cultural Impact of Advertising in Computer-mediated environments.
   • To Create Knowledge about the "How to" of Interactive Advertising.

   As a part of research projects of the Seminar, each seminar participant created a Web site dealing with the issue of web audience measurement. The main objective of this project is to provide a knowledge base to those who wish to understand how to MEASURE the "audience size" of Internet Advertising and Marketing Communication for their company (profit-making enterprise) or organization (such as non-profit). This is currently one of the most perplexing and pressing issues facing Interactive Advertising planners.

   It is essentially to be a tutorial and "how to do it" with respect to audience measurement on the Web which has become an increasingly complex task. The sites explore the current measurement systems available and include such information. These Web sites aim to serve as a vehicle to express some sense of what is needed to develop our understanding of measuring advertising audiences on the Internet. This project is published at URL:

   http://uts.cc.utexas.edu/~admedium/ADV391K_Spring1998_wam.html

   In addition to the above projects, Professor Wayne Danielson worked with a class of students on the redesign of the journalism department web site. Its value is that it provides a mechanism for the department to keep its web pages updated and innovative. At the same time, it gives students valuable experience in web page organization, management and design (http://www.utexas.edu/coc/journalism/).

   The purpose in the spring was to offer a good picture of what the department does. In the fall, Professor Danielson will move toward offering more supplementary materials for instruction across the department and some actual units of on-line instruction. They will also increase the amount of multimedia material-- in particular, they will feature a section called "soundbytes" to display brief audio and video clips from the work of broadcast journalism students. The Intel facility played a key role in the project.

Previous Status Report

Back to Top

Digital Video Telecommunications Research and Teaching
August, 1999

1. Principal Investigator: Al Bovik

2. Associated Investigator: Brian Evans

3. Project Progress:

   We have developed an innovative approach for achieving high quality, low bitrate video communication over computer networks and over wireless communications channels: foveated video compression. In this strategy, digital video is resampled at a nonuniform spatial density, reflecting the human eye's own foveated retinal sampling. In this way, video/image compression is multiplied by factors of 5-10. We have successfully implemented such an algorithm on H.263 / MPEG-I / MPEG-II standard-compliant video streams. Several demonstrations of the foveated compressed video algorithm can be seen at: http://pineapple.ece.utexas.edu/class/Video/demo.html, for both H.263 and MPEG-II.

4. Publications:

   [1] S. Lee and A.C. Bovik, "Rate control for foveated MPEG/H.263 video," IEEE International Conference on Image Processing, Chicago, Illinois, October 4-7, 1998.

   [2] S. Lee, M.S. Pattichis and A.C. Bovik, "Foveated image/video quality assessment in curvilinear coordinates," International Workshop on Very Low Bitrate Video Coding, Urbana, Illinois, October 8-9, 1998.

   [3] S. Lee, M.S. Pattichis and A.C. Bovik, "Foveated image/video quality assessment and compression gain," IEEE Transactions on Image Processing, to appear

   [4] S. Lee and A.C. Bovik, "Optimal rate control for real-time, low bitrate foveated video coding," IEEE Transactions on Image Processing, to appear.

   [5] S. Lee and A.C. Bovik, "Traffic smoothing for variable bit rate video transmission," IEEE Transactions on Circuits and Systems for Video Technology, to appear.

   [6] J. Kim, A.C. Bovik and B.L. Evans, "Generalized predictive binary shape coding using polygonal approximation," Signal Processing: Image Communication, to appear.

   [7] T.D. Kite, B.L. Evans and A.C. Bovik, "Modeling and quality assessment of halftoning by error diffusion," IEEE Transactions on Image Processing, to appear.

   [8] D. Craievich, B.S. Barnett and A.C. Bovik, "A stereo visual pattern image coding system," Image and Vision Computing, to appear.

   [9] S. Lee and A.C. Bovik, "Very low bit rate foveated video coding for H.263," IEEE International Conference on Acoustics, Speech, and Signal Processing, Phoenix, Arizona, March 15-19, 1999.

   
   Biographical Sketches

   Alan C. Bovik is currently the General Dynamics Endowed Fellow and Professor in the Department of Electrical and Computer Engineering, the Department of Computer Sciences, and the Biomedical Engineering Program at the University of Texas at Austin, where he is the Associate Director of the Center for Vision and Image Sciences. During the Spring of 1992, he held a visiting position in the Division of Applied Sciences, Harvard University, Cambridge, Massachusetts. His current research interests include digital video, wavelets, and computational aspects of biological visual perception. He has published over 200 technical articles in these areas and holds U.S. patents for the image and video compression algorithms VPIC and VPISC.
   Dr. Bovik is a winner of the University of Texas Engineering Foundation Halliburton Award and a two-time Honorable Mention winner of the international Pattern Recognition Society Award for Outstanding Contribution (1988 and 1993). He is a Fellow of the IEEE and has been involved in numerous professional society activities, including: Board of Governors, IEEE Signal Processing Society; Editor-in-Chief, IEEE Transactions on Image Processing (present); and Founding General Chairman, First IEEE International Conference on Image Processing, Austin, Texas, November 1994.

   
   Brian L. Evans is an Assistant Professor in the Department of Electrical and Computer Engineering at The University of Texas at Austin, and the Associate Director of the Laboratory for Vision Systems within the Center for Vision and Image Sciences. His research interests include real-time software, embedded systems, heterogeneous systems, image and video processing systems, system-level design, symbolic computation, and computer-aided design. He has developed numerous computer-aided design tools to prototype and test research ideas. His B.S.E.E.C.S. (1987) degree is from the Rose- Hulman Institute of Technology, and his M.S.E.E. (1988) and Ph.D.E.E. (1993) degrees are from the Georgia Institute of Technology. From 1993 to 1996, he was a post-doctoral researcher at the University of California at Berkeley with the Ptolemy Project. Ptolemy is a research project and software environment focused on design methodology for signal processing, communications, and controls systems. He is the recipient of a 1997 NSF CAREER Award.

Previous Status Report

Back to Top

Multimedia Delivery Systems
August, 1999

1. Principal Investigator: Harrick M. Vin

2. Associated Investigators: Jayadev Misra

3. Summary of Project's Progress since April 1999

   During the past year, our research has focused on two problems: (1) design of networks and operating systems for supporting next-generation distributed applications, and (2) development of tools for enhancing distance learning and self-paced education. To support next-generation of distributed applications, we have developed (1) resource management techniques that meet the performance requirements of applications, and (2) a low-overhead fault-tolerance framework for cooperative distributed applications. This has led to the design of an integrated file system, packet scheduling algorithms and congestion control protocols for integrated services networks, operating system mechanisms for end-stations, an extensible toolkit for providing transparent low-overhead fault-tolerance to parallel and distributed applications, and insights for designing next-generation networks and operating systems.

   We have developed and released the following software:

   • InfoWeave Educational Application

     This application contains an authoring component, a document delivery component, and a user/group management utility. The authoring component allows an author to publish a previously authored html document as a customizable InfoWeave document. The document delivery component allows users to customize their views of the document through the use of custom note attachments and dynamic link embedding. Finally, the user/group management utility allows the system administrator to create and maintain users and groups. Together, these components enable a student to generate personalized views of each document, and thereby facilitate distance learning and self-paced education. This application can be accessed via http://www.cs.utexas.edu/users/dmcl/software.html.

   • QLinux

     QLinux is an enhanced Linux kernel for multimedia computing; it provides quality of service (QoS) guarantees to applications. QLinux, based on the Linux 2.2.x kernel, combines some of the latest innovations in operating systems research. It includes (1) Hierarchical Start Time Fair Queuing (H-SFQ) CPU scheduler, (2) Hierarchical Start Time Fair Queuing (H-SFQ) network packet scheduler, (3) Lazy receiver processing (LRP) network subsystem, and (4) Cello disk scheduling algorithm. The H-SFQ CPU scheduler enables fair allocation of cpu bandwidth to individual applications and application classes. The H-SFQ packet scheduler provides rate guarantees and fair allocation of network bandwidth to packets from individual flows as well as flow aggregates (classes). Lazy receiver processing enables accurate charging of TCP/UDP protocol processing overhead (including interrupt processing) to the appropriate process. The Cello disk scheduler supports multiple application classes such as interactive best-effort, throughput-intensive best effort and soft real-time and fairly allocates disk bandwidth to these classes. The software can be accessed via http://www.cs.utexas.edu/users/dmcl/software.html.

4. Related Publications [From April 1999 to August 1999]

   [1] S. Rao, L. Alvisi, and H.M. Vin, ``The Cost of Recovery in Message Logging Protocols,'' IEEE Transactions on Knowledge and Data Engineering (to appear), 1999

   [2]  P. Shenoy and H.M. Vin. ``Efficient Striping Techniques for Multimedia File Servers,'' Performance Evaluation Journal (to appear), 1999

   [3]  P. Shenoy and H.M. Vin. ``Efficient Support for Interactive Operations in Multi-resolution Video Servers,'' ACM Multimedia Systems Journal (to appear), 1999

   [4]  P. Shenoy and H.M. Vin, ``Failure Recovery Algorithms for Multimedia Servers,'' ACM Multimedia Systems Journal (to appear), 1999

   [5]  R. Tewari, H.M. Vin, A. Dan, and D. Sitaram, ``Resource-based Caching for Web Servers,'' ACM Multimedia Systems Journal (to appear), 1999

   [6]  P. Shenoy, P. Goyal, and H.M. Vin, ``Architectural Considerations for Next Generation File Systems,'' In Proceedings of ACM Multimedia'99 (to appear), October 1999

   [7]  J. Sahni, P. Goyal, and H.M. Vin, ``Scheduling CBR Flows: FIFO or Fair Queuing?,'' In Proceedings of the 9th International Workshop on Network and Operating Systems Support for Digital Audio and Video, pp. 13-27, June 1999

   [8]  S. Rao, L. Alvisi, and H.M. Vin, ``Egida: An Extensible Toolkit For Low-overhead Fault-Tolerance,'' In Proceedings of IEEE International Conference on fault-Tolerant Computing (FTCS), pp. 48-55, June 1999

   [9]  R. Tewari, M. Dahlin, H.M. Vin, and J. Kay, ``Design Considerations for Distributed Caching on the Internet,'' In Proceedings of the International Conference on Distributed Computing and Systems (ICDCS), pp. 273-284, May 1999

5. Related Grants
   • LARIAT: Strategic partnership between Dell Computer Corporation and UT-Austin, 2 grants totaling $50,000, 1999-2000
     
   • Novell at UTAustin Program, $183,800, 1999-2000
     
   • Texas Higher Education Coordinating Board, Advance Technology Program, ATP-443, $132,749, 1/1/98 - 12/31/99
     
   • IBM University Partnership Program, $70,000, 9/1/97-8/31/99
     
   • Dell, Equipment grant (quad-processor servers), 1998
     
   • Lucent Bell Laboratories, Equipment grant (video compression hardware), 1998
     
   • AT&T Foundation, $20,000, 1997
     

6. Related Web Sites

   Distributed Multimedia Computing Laboratory URL:
   http://www.cs.utexas.edu/users/dmcl

Previous Status Report

Back to Top

Virtual Museum Project
August, 1999

1. Principal Investigator:
   

   Timothy Rowe

2. Technical Support Contacts:
   

   Ms. Rhonda Baird
   Dr. Richard Ketcham

3. Project Overview:

   The Virtual Museum Project aims to train faculty and students in advanced instructional technologies, by developing multimedia and novel visualizations highlighting the special collections of UT¼s many museums and galleries. A network of highly skilled technical staff already in place, has begun to train students, staff, and faculty as they assist them in developing state-of-the-art instructional multimedia. Our instructional materials are now being distributed via Internet and CD-ROM, and as interactive museum and gallery exhibits. The Virtual Museums Project will break new ground in multimedia development in a highly visible way, while training strategically positioned groups of faculty and students.

   Our training focus is the development of premiere instructional materials that utilize the vast holdings of the Archer M. Huntington Art Gallery, the Texas Memorial Museum, and the Harry Ransom Humanities Research Center. The grand technical challenge for the Virtual Museum Project is in digitizing in great detail the many delicate, complex 3-D objects from art and natural history collections. We will use everything from simple scanners to high resolution X-ray CT scanning, laser surface scanning, video digitization, and other techniques will be used to build highly detailed models of the gems of UT¼s special collections.

4. Equipment Status:

   All equipment received so far is in operation and has contributed directly to the publication of two new Web sites, one educational CD-ROM, and one museum exhibit. Several computers have undergone major upgrades, and we have begun operating a volumetric 3-D rendering program on our quad- processor server.

5. Progress to Date:

   We have developed a Web site entitled "The Digital Morphology Group" that serves advanced 3-D visualizations on significant fossils, corals, and modern skeletal materials from several natural history museum collections. Most of the imagery and animations on this site are based on 3-D datasets generated by our high-resolution X-ray CT scanner, and these data are already fueling secondary research by scientists in Australia, Argentina, France, Germany, and in the US.

   We have also rebuilt the Web site for the University of Texas High-Resolution X-ray CT Facility, a facility of unique power for digitizing in 3-D complex natural materials. The Web site demonstrates basic CT technology, along with a CT image folio of such diverse objects as natural diamond-bearing rocks, meteorites, and numerous fossils.

   In the Fall, The Texas Memorial Museum mounted a special exhibit that included a multimedia presentation for its dinosaur gallery. We CT scanned the world¼s oldest dinosaur ‚ Herrerasaurus ‚ and developed an education multimedia exhibit that demonstrated the skull being CT scanned, along with 3-D models of the skull and other novel animations that showed details of its internal and external structure. We believe that this special exhibit at the Texas Memorial Museum was the first exhibit ever held in which the audience could inspect a unique and ancient fossil from the inside out.

   We continue an industrial collaboration with Bio-Imaging Research, Inc., of Lincolnshire, Illinois, who manufactured our CT scanner, to exploit the application of this powerful new technology across a broad spectrum of natural materials and applications. Our collaboration involves hardware refinement, software refinement and development, and the development of novel visualization techniques.

   The following publications and Web sites document this work:

   • The University of Texas High-Resolution X-ray CT Facility
     http://www.ctlab.geo.utexas.edu/

   • The Digital Morphology Group
     http://www.ctlab.geo.utexas.edu/dmg/index.html

   • Examples of 3-D Animations
     http://www.ctlab.geo.utexas.edu/imfoframes/imfoani.html

   • Bio-Imaging Research, Inc.
     http://www.bio-imaging.com/

   CD-ROMs:

   Rowe, T., K. Kishi, J. Merck, Jr., and M. Colbert, 1998. The Age of Dinosaurs. Educational Interactive Multimedia on CD-ROM for Macintosh and PC computers. W. H. Freeman & Co. ISBN 0-7167-3378-1.

   Publications:

   Rowe, T., J. Kappelman, W. D. Carlson, R. A. Ketcham, and C. Denison, 1997. High-Resolution Computed Tomography: a breakthrough technology for Earth scientists. Geotimes, 42: 23-27.

Previous Status Report

Back to Top

Multimedia Navigation Systems
August, 1999

1. Principal Investigator: Russell Pinkston

2. Associated Investigators:
   
   

   Fritz Schwentker
   Yacov Sharir
   Darlene Wiley

   Graduate and Technical Assistants:

   Tyson Breaux (Office of Computing Technologies)
   Anderson Mills (Engineering/Acoustics)
   Larisa Montanaro (Music/Voice)

3. Technical Support Contact: Russell Pinkston

4. Current Project Status and Summary:
   

   VR Project (Mills/Pinkston/Sharir)

   The University of Alberta Minimum Reality Toolkit (MRTK) was installed and configured on both SGI Onyx and Intel Pentium/NT systems. A stereo Head Mounted Display (General Reality CyberEye) and magnetic tracking system (Polhemus Fastrak) were installed, calibrated, and incorporated into the CASA VR System. A virtual world "Dancing with the Virtual Dervish" (Gromala, Novak, Sharir), original created at the Banff Centre for the Arts in Canada, was successfully mounted on the Onyx. Work continues on incorporating a glove controller into the MRTK and mounting the Dervish World on NT systems.

   Vocal Arts Lab Project (Montanaro/Wiley)

   Over the past six months, the Vocal Arts Lab at the Center for Advanced Studies in the Arts (CASA) continued working on a web site devoted to the teaching of voice to beginning students.

   The lab has also been used by individual students to examine their singing voice. These techniques of analysis were addressed in a Vocal Pedagogy course that was taught in the lab and will also be addressed in a course this fall that deals with the use of technology in the study of vocal pedagogy.

5. Additional Funding Obtained

   Professor Pinkston received a CASA Summer Fellowship, which supported the research described above. He also has a Guggenheim Fellowship in Music Composition for 1999-2000, which will be devoted primarily to the creation a new VR work on the CASA VR system.

   Professor Wiley and her assistant, Larisa Montanaro, have received a grant from CASA for the spring 2000 to complete a CD-ROM incorporating aspects of the existing Vocal Pedagogy web site with more advanced topics and macros.

   Microsoft has donated $60,000 to CASA to support faculty research through a program called "CASAbatical." This year, the program will provide one semester leaves to professors Schwentker and Wiley, who will pursue research related to their Intel D5 projects at CASA.

6. Related Links:

   Main CASA Web Page: http://casa.pac.utexas.edu
   Electronic Music Studio Page: http://www.utexas.edu/cofa/music/ems
   Professor Schwentker's Project: http://wwwvms.utexas.edu/~fritzs/CasaMidi.htm
   Vocal Pedagogy Web Page: http://www.utexas.edu/cofa/music/voice
   

Previous Status Report

Back to Top

Distributed Visualization
August, 1999

1. Principal Investigator: Chandrajit Bajaj

2. Equipment:

   Six Dual 400 Mhz Intel Zeon Processors PC with 5 12 Mbytes of RAM and Intergraph Intense 3D Pro 3410T Graphic adapters

   • 16 Mbytes of video RAM
   • no Geometry acceleration
   • 9.1 Gybyte Ultra-2 SCSI hard drive

   Six Dual 550 Mhz Intel Zeon Processors PCs with 512 Mbytes of RAM and Intergraph Intense 3D Wildcat 4000 Graphic adapters

   • 16 Mbytes of video frame buffer
   • 64 Mbytes of texture memory
   • 2 Gflops of geometry accelleration allowing 3.4 Million Triangles/second
   • 9.1 Gybyte Ultra-2 SCSI hard drive

   These machines are intended for two experiments (subprojects A and B) in distributed visualization.

3. Current Project Status:

   Sub-project A is peer-to-peer collaborative visualization on the multimedia desktop . Our current collaboration architecture defines an extensible environment for developing collaborative visualization applications that support large scale geographically dispersed interaction amongst multiple participants and systems. The architecture is connectionless, scalable, supports multi-group federation, has an adaptable data location model, supports flexible applications coupling, , multiple coordination strategies, dynamic downloading of executable code and is event driven. Interoperable implementations exist in C, C++. and Java.
   

   Sub-project A	Sub-project B

   Sub-project B is on multi-pipe, multi-projector distributed visualization on a panoramic power wall. The cluster of Intel machines shall be connected using a high speed switch (fast, low latency). Three of the machines shall initially serve as display or graphics servers and drive a set up of three electrohome projectors in a rear projection panoramic power wall. The research challenges include dynamic resolutions visualization with guaranteed display frame rates, scalability across diverse network bandwith, disk transfer rates and inter-processor communication latency.

4. Grants related to the use of the Intel machines include:

1. "Modeling and Visualization with algebraic surfaces and splines"; National Science Foundation
2. "Data Intensive Visulization"; National Aeronautics and Space Administration
3. "Data and Display Intensive Visualization"; Sandia National Labratories

Leveraging from other projects we have obtained the following necessary equipment utilizing funds from NSF/KDI, and DOE-ASCI grants:	Additionally we plan to obtain the following:
(1.) 6 Extron 112+ (2.) 3 flat panel displays (3.) Altinex 8 x 4 RGBHV matrix switcher (4.) 3 electrohome 9500projectors with fast phosphor (5.) 3 extended range stereo emitters and 3 active LCD stereo glasses (6.) an 8 port KVM Switch allowing a single control point to an 8 machine cluster.	(1.) An additional cluster of 6 Merced based Intel machines(as they become available.) (2.) an 8 port Myrinet switch (3.) 8 Myrinet NICS for the Intel PCs (4.) 4 Top of the line 21" monitors or flat panel digital displays

Previous Status Report

Back to Top

Internet City
August, 1999

1. Principal Investigator: Dr. David Maidment (bio) (Center for Research in Water Resources)
   

2. Associated Investigators: Dr. Ken Foote (The Department of Geography) and Dr. Barbara Parmenter (Community and Regional Planning Program in The School of Architecture)
   

3. Technical Support Contact: Ty Lehman lehman@mail.utexas.edu 512-471-3111
   

4. Project Summary:
   

   We are creating and currently serving our www and ftp site on our Intel equipment. We are working with the city of Austin to serve some of the data on our ftp site as well as work with Texas natural resource and conservation commission (tnrcc) and the LCRA on web mapping projects. We are working with ESRI to beta test Arc Info 8.0, which will help us create the next generation of web based map sharing.

5. Related Publications:
   

   Center for Research in Water Resources and CRWR ftp site

Previous Status Report

Back to Top

Virtual Laboratories and Examinations
August, 1999

1. Principal Investigator: John Kappelman, Professor, Department of Anthropology, The University of Texas at Austin

2. Associated Investigators: None

3. Technical Support Contact: John Kappelman, Greg Weiner

4. Project Summary:

   The VExams© program is in the beta-testing phase. It was used in Anthropology 301 during the fall of 1998, and is currently being used in Anthropology 348. It is anticipated that it will be used in the Austin Independent School District during the next academic year as well as in several courses at UT Austin.

   I have given presentations on the VExams© program to the following groups:

   • 1999, 23 March. "The VExams© Program." Discovery Learning Group, UT Austin
   • 1999, 14 March. "The VExams© Program." AISD Science Director
   • 1999, 28 February. "The VExams© Program." Center for Instructional Technologies, UT Austin
   • 1998, 4 December. "An Interactive Computer-Based Multimedia Program for Examinations." Invited session participant, "Anthropology Courseware: Teaching with Electronic Media: National meeting of the American Anthropological Association, Philadelphia, PA
   • 1998, 27 October. "An Interactive Computer-Based Multimedia Program for Examinations." National Science Foundation, Division of Undergraduate Education.
   • 1998, 16 October. "Testing on the Fly: A Program for Delivering Virtual Multimedia Examinations." Department of Anthropology, The University of California Los Angeles, CA.
   • 1998, 1 October. "Virtual Multimedia Exams for the Geosciences." Department of Geology and Geophysics, University of Utah, Salt Lake City, Utah.
   • 1998, 30 September. "Lasers, hi res x-ray CT, hard copy output and animations for virtual laboratories, multimedia exams and research." Society of Vertebrate Paleontology, Snowbird, Utah.

5. Equipment Status:

   We were supposed to take delivery of the second shipment of computers for this grant in December 1998 but unfortunately delivery has been delayed and we have still not received these computers. We have therefore not been able to integrate the Intel computers into our teaching laboratory. We were also informed that the computers for this second shipment of computers would not include monitors.

   The Dean of Liberal Arts at UT Austin has generously agreed to supply the six monitors for the second installment.

6. Related publications

   • In press. A computer program for delivering virtual multimedia examinations across secure networks and the web. American Journal of Physical Anthropology (J. Kappelman, A. Gordon, D. Johnson, T. Ryan, R. Scott, and E. Seiffert. Abstract.
   • 1999. Virtual multimedia labs and exams in physical anthropology. Current ACITS 32, pp.3-5.
   • 1998. An interactive computer-based multimedia program for examinations. In American Anthropological Association , Abstracts of the 97th Meeting, Philadelphia, PA, p. 254. Abstract.
   • 1998. Virtual Multimedia Laboratories and Examinations in Physical Anthropology. In Computer Enhanced Learning: 100 Courses at 50 of America's Most Wired Colleges
   • Electronic publication: http://www.wfu.edu/~womack/Vignettes/Socsci/KappelmanRev..html
   • 1998. Lasers, hi res x-ray CT, hard copy output and animations for virtual laboratories, multimedia exams and research. Journal of Vertebrate Paleontology. 18 (supp. to No.3):54A-55A.

7. Related Web Sites:
   

   http://www.dla.utexas.edu/depts/anthro/kappelman/kapp.html
   Virtual laboratories in physical anthropology

Previous Status Report

Back to Top

ADVANCED MODELING & SIMULATION

---

Computational Fluid Dynamics and HPC

(or HYDRA: A Prototype PC Cluster System for Flow and Transport Simulation)

August, 1999

1. Principal Investigator: Graham F. Carey

2. Associate Investigators: W. Barth, R. McLay, A. Ardelea, B. Kirk

3. Project Summary:

   During the past six months work in the CFD Lab on the cluster research project has continued in two main areas: 1) the solution of coupled viscous flow and heat transfer problems including surface tension effects; 2) the numerical modeling of patterns in chemically reacting systems. The work has involved both performance studies and phenomenological investigations in both cases. In the coupled flow and transport problem we have been investigating the effects of thermocapillary surface tension in a low gravity environment and have carried out performance studies to investigate scaling as the number of processors increases. Results of the scaling studies are indicated in one of the web figures provided. In the case of the pattern studies, we have been investigating the structures of the solutions to nonlinear coupled reactive transport problems and carrying out comparisons with experimental studies performed by Harry Swinney and his associates in the Center for Nonlinear Dynamics. In this case we made parallel simulations on high resolution grids to explore both organized and labyrinthine pattern formations. A sample pattern is also provided.

4. Related Publications:

   G. F. Carey, H. Swinney, R. McLay, S. van Hook, A. Pehlivanov, G. Bicken and W. Barth, "Thermocapillary Surface Tension Driven Flows", presented at AIAA 37th Aerospace Sciences Meeting and Exhibit, Reno, NV, Jan 11-14, 1999 and submitted to AIAA J. of Thermophysics and Heat Transfer, March 1999.

   G. F. Carey, R. McLay, G. Bicken, S. Swift, W. Barth, A. Pehlivanov and A. Ardelea, "Parallel Distributed Solution of Viscous Flow with Heat and species Transfer", to be presented at High Performance Computing '99, part of ASTC 99, San Diego, CA, April 11-15.

   G. F. Carey, R. McLay, G. Bicken, B. Barth, "Parallel Finite Element Solution of 3D Rayleigh-Benard-Marangoni Flows, Submitted to Int'l J. Numer. Meths. in Fluids, 1998.

5. Project website:

   http://www.cfdlab.ae.utexas.edu/intel_beowulf

Previous Status Report

Back to Top

Deep Submicron Process and Device Modeling and Analysis
August, 1999

1. Principal Investigator: Al F. Tasch

2. Associated Investigators: Sanjay Banerjee

3. Research Progress:

   We have completed the development of rigorous, physically based damage models for Ge and In implants into Si. These models are based on the original Kinetic Accumulative Damage Model (KADM) approach developed by Tian et al. in the UT-Austin ion implant modeling program. These new models provide the required as-implanted impurity profiles and damage profiles, which can be used as inputs for transient enhanced diffusion simulation, and for the simulation of multiple implants. These models have explicit dependence on all of the major implant parameters and have been extensively validated by a large range of experimental SIMS and RBS data for both single and multiple implants. The models have been implemented in UT-MARLOWE Version 5.0 and have been proactively transferred to at least 20 individuals in Intel at Hillsboro, Oregon and Santa Clara, California.

   We have completed the development of TOMCAT, a new physically based Monte Carlo ion implantation simulator which is capable of simulating 2-D implants into arbitrary topography with multiple layers, both crystalline and amorphous. Emphasis has been placed on computational efficiency, and improvements over previous simulators ranging from 1-3 orders of magnitude in CPU time have been observed. The accuracy of the simulation is observed not to be compromised by the algorithms for improved computational efficiency. This contract funded the development of the computationally efficient damage models in TOMCAT.

   The new simulator has been implemented in UT-MARLOWE Version 5.0 and has been proactively transferred to at least 20 individuals in Intel at Hillsboro, Oregon and Santa Clara, California.

   New Features of UT-MARLOWE 5.0

   1. Portability across all major UNIX platforms (AIX, Solaris, Linux, HP-UX, SGI, DEC), as well as Windows NT.
   2. Improved 2-D capability (TOMCAT).
   3. Physics models for germanium and indium implants.
   4. Dose-splitting method of statistical enhancement of impurity and defect profiles.
   5. 3-D defect output.
   6. User-defined oxide density.
   7. Improved defect placement model.
   8. Experimentally verified low-energy scattering model for sub-keV implant energies.

   Major Features of TOMCAT

   1. Native 2-D simulator - does not rely on postprocessors.
   2. Capable of simulating implants into arbitrary topography.
   3. Simple user-interface for topography generation.
   4. Import/export capability for standard process simulator meshes.
   5. Uses physics model from UT-MARLOWE.
   6. Topography-aware lateral trajectory replication for greatly enhanced computational efficiency.
   7. Improved deterministic propagation algorithm for improved computational efficiency.
   8. Use of stochastic MC algorithms for ion propagation in amorphous materials for greatly improved computational efficiency.
   9. Dose-splitting method of statistical enhancement of impurity and defect profiles.
   10. Multiple implant capability is supported in 2-D. Arbitrary sequences of implants can be performed with accurate modeling of implant history on current implant.
   11. Both dopant and defect information are available. Although the supported KP model is simple, it is nevertheless quite accurate in modeling the dose dependence of implants, and reasonably accurate in modeling defect concentrations. User-defined +n net defect density is available.

   It should be kept in mind that all of the models developed and implemented in UT-MARLOWE 5.0 and TOMCAT have been extensively verified with experimental data in almost every case. This includes both the ultra-low and MeV energy ranges, for all implant species available (B, BF2, As, P, Ge, In, Si).

   We have proactively transferred User's manuals, model information, and source code (for UT-MARLOWE Version 5.0) for those models to at least 20 individuals in Intel at Hillsboro, Oregon and Santa Clara, California

4. Relevant Publications:

   The documents in which these results are described are:

   • "Physically Based Models for Indium and Germanium Ion Implants into Silicon," Y. Chen, B. Obradovic, M. Morris, G. Wang, G. Balamurugan, D. Li, A.F. Tasch, D. Kamenitsa, W. McCoy, S. Baumann, R. Bleier, D. Sieloff, D. Dyer, and P. Zeitzoff, Proceedings of the Fifth International Symposium on Process Physics and Modeling in Semiconductor Device Manufacturing, Spring Meeting of the Electrochemical Society, Seattle, May 2-6, 1999.

   • "Monte Carlo Simulation of Heavy Species (Indium and Germanium) Ion Implantation into Silicon," Y. Chen, B. Obradovic, M. Morris, G. Wang, G. Balamurugan, D. Li, A.F. Tasch, D. Kamenitsa, W. McCoy, S. Baumann, R. Bleier, D. Sieloff, D. Dyer, and P. Zeitzoff, IEEE Journal on TCAD, http://www.ieee.org/products/online/journal/tcad/accepted/chen-feb99/final.pdf.

   • "UT-MARLOWE 5.0 with TOMCAT," B. Obradovic, G. Wang, Y. Chen, D. Li, C. Snell, and A.F. Tasch, User's Manual, April 20, 1999, available directly from Al F. Tasch, University of Texas at Austin.

   • "Monte Carlo Simulation of Ion Implantation into Topographically Complex Structures," B.J. Obradovic, G. Balamurugan, G. Wang, Y. Chen, and A.F. Tasch, Technical Digest of the IEEE 1998 IEDM, pp. 513-516, San Francisco, CA, December 6-9, 1998.

Previous Status Report

Back to Top

Reservoir Simulation
August, 1999

1. Principal Investigator: Dr. Kamy Sepehrnoori

2. Associated Investigator: Dr. Gary A. Pope

3. Technical Support Contact: Jason Abate, abate@ticam.utexas.edu, 512-471-6947

4. Project Abstract:

   Accurate and efficient reservoir simulation is of critical importance to the economic well-being of the United States. Future oil and gas production in this country is dependent upon improved oil recovery, reservoir simulation and reservoir characterization technologies. Intense computer simulation is essential for effective reservoir management. Parallel reservoir simulators have the potential to solve larger and more realistic problems than previously possible. Our research involves development of new physical and chemical models, accurate numerical methods, new reservoir description techniques and their implementation in parallel environments. In particular, we investigate porting of reservoir simulators to a distributed memory cluster of Intel processors. By performing this task, the feasibility of using a cluster of Intel processors for solving large-scale reservoir simulations will be assessed.

5. Equipment status:

   At this time all equipment anticipated for this project has been received and assembled. This equipment includes the following:

   • Quantity 16 - Single processor Intel Pentium-II 300Mhz with MMX technology, 256 MB of memory, and 4 GB of disk.
   • Quantity 8 - Dual processor 400 MHz Intel Pentium II Xeon, 512 MB of memory and 9 GB of disk
   • Quantity 3 - Intel Ether Express PRO/100 LAN Adapter, 5 pack.
   • Quantity 3 - Intel 510T Switch.
   • Quantity 2 - Ultrascan 20T monitor.
   • Quantity 8 - 1000HS monitor

   We have ordered an additional 128 MB of memory for each of the original cluster machines, bringing the total to 384 MB per processor.

   The machines are currently setup as two separate clusters, one of the 16 300 MHz Pentium IIs and one of the 8 dual Xeon machines. Each cluster is connected by one of the 510T switches, and both switches are connected to a file server/development machine. We hope to connect the two switches with a 2.2 Gbps switch interconnect, which would allow us to utilize all of the machines as a single cluster, and make simulation runs with 750,000 gridblocks.

   Each machine runs RedHat Linux 5.1, which provides the standard development tools and utilities, including the GNU C/C++ compilers. Additionally, Portland Group Fortran 77 and Fortran 90 compilers are being used. All parallel communication is done with MPICH, the implementation of the standard Message Passing Interface developed by Argonne National Laboratory. We also installed PBS, the Portable Batch System to handle batch queueing of parallel jobs.

   A summary of the current hardware status and individual machine software configuration may be found at http://topeka.cpge.utexas.edu/status.html.

6. Research Progress:

   After installing the software, we have run a number of tests and benchmarks to veryify the machines were working correctly. Following that, the IPARS (Integrated Parallel Accurate Reservoir Simulator) framework [1,2,3] was ported to Linux, which required a few minor modifications to the framework. IPARS provides support for input/output, memory management, domain decomposition and message passing between the processors. The simulator is a new, fully-implicit EOS (equation of state) compositional model which uses PETSc, the Portable, Extensible Toolkit for Scientific Computation for solving large, sparse linear systems which arise during the simulation. The simulator is a module which was developed under the IPARS framework.

   Once the simulator results were verified through comparison with results from other machines, several test problems were run to study the range of simulation possible with a cluster of PCs. A modified Society of Petroleum Engineers (SPE) 5th comparative solution problem [4] which uses three-phase flow throughout the simulation was used to benchmark the cluster. A paper describing this work has been submitted for publication [5], and was also presented at the 1999 SIAM Conference on Mathematical and Computational Issues in the Geosciences, March 24-26, San Antonio.

   We have also run performance tests on the Xeon machines, and found them to exhibit excellent performance. They perform nearly twice as fast as the 300 MHz Pentium IIs, mainly due to the faster clock speed and the 100 MHz system bus. They are very competitive with the 160 Mhz P2SC nodes in the IBM SP we use at the Maui High-Performance Computing Center. The table below shows the execution time in seconds for one of our benchmark tests on each of these platforms:

   Number of Processors	300 MHz Pentium II	400 MHz Xeon	160 MHz P2SC
   1	1112	645	554
   2	483	281	242
   4	245	143	126
   8	133	74	66

   We have recently tested the Xeon cluster in dual-processor mode under MPI using the shared memory capabilities of MPICH. We have found that the dual-processor machines perform well, although there is some penalty for sharing a single memory bus and network interface between two processors.

   We have recently begun a comparison study of cluster networking performance using a variety of available technologies. We will be comparing the performance of 100 Mbps Ethernet, Myrinet and Scali. Our goal is to determine how significant of an effect the use of higher bandwidth, lower latency networking has on real-world reservoir simulation problems, and whether they are worth the additional expense. Initial experiments with Myrinet suggest that it significantly improves the parallel performance past 16 processors:

   

   Additionally, we will be serving as a beta test site, evaluating GigaNet's networking hardware under Linux.

   Work has continued on development of a new Fortran 90-based framework for reservoir simulation, named STAMPEDE (the Simulation Toolbox for Advanced Modelling for Parallel Experimentation, Design and Execution), and a new simulator which uses this framework, GPAS (the General Purpose Adaptive Simulator). All development work is being done locally on our clusters, which have proven to be a very efficient, convenient resource for developing parallel programs. We expect to make the first test runs of GPAS in the coming months, which will be followed by benchmarking the simulator on both of our clusters.

   This work has demonstrated that clusters of PCs are a viable platform for making large-scale simulation runs, that they perform competitively with traditional parallel machines (especially the latest PC hardware), and that they make very good development platforms for parallel simulation codes.

   Further details of these results will be reported in upcoming papers and presentations and our cluster website. The current draft of the paper Parallel Compositional Reservoir Simulation on a Cluster of PCs can be found here.

References:

• Peng Wang, Ivan Yotov, Mary Wheeler, Todd Arbogast, Clint Dawson, Manish Parashar and Kamy Sepehrnoori, A New Generation EOS Compositional Reservoir Simulation: Part I - Formulation and Discretization, Paper SPE 37979 presented at the 1997 SPE Reservoir Simulation Symposium, San Antonio, June 8-11, 1997

• Manish Parashar, John A. Wheeler, Gary Pope, Kefei Wang and Peng Wang, A New Generation EOS Compositional Reservoir Simulation: Part II - Framework and Multiprocessing, Paper SPE 37977 presented at the 1997 SPE Reservoir Simulation Symposium, San Antonio, June 8-11, 1997

• Gary Pope, Kamy Sepehrnoori, Mary Wheeler, and Tom Morgan. Advanced computational technology initiative: New generation framework for petroleum reservoir simulation first annual report. Technical report, Argonne National Laboratory and The University of Texas at Austin, 1996.

• J. E. Killough and C. A. Kossack. Fifth Comparative Simulation Project: Evaluation of Miscible Flood Simulators, February 1-4, 1987, Paper SPE 16000 Presented at the Ninth SPE Symposium on Reservoir Simulation, San Antonio, TX

• Jason Abate, Peng Wang and Kamy Sepehrnoori, Parallel Compositional Reservoir Simulation on a Cluster of PCs, Submitted to Communications in Numerical Methods in Engineering, December, 1999.

---

Previous Status Report

Back to Top

Real-Time Process Control
August, 1999

1. Principal Investigator: T. F. Edgar, edgar@mail.utexas.edu, (512) 475-9201

2. Associated Investigators: Prof. J. Qin, Dept. of Chemical Engineering

3. Technical Support Contact: Tyler Soderstrom, 471-1046

4. PI Assistant: Maria Cruz, m.cruz@cc.utexas.edu, (512) 475-9239

5. Project Abstract:

   This project deals with the application of PC's for real-time modeling and control of chemical processes. The goal of the research is to demonstrate the use of Intel PC's in a real-time, computationally demanding environment where advanced optimization and control algorithms are employed. The experimental applications under study include a batch distillation column, a reactive distillation column (currently under construction) and a rapid thermal chemical vapor deposition system. In addition, we are using PC's off-line to carry out optimization and estimation with chemical dynamic simulators such as Dynaplus from Aspentech.

6. Equipment Status:

   So far we have received one dual Pentium PC and three single Pentiums. All computers are operational.

7. Research Progress.

   The motivation of the subproject on dynamic data reconciliation is to minimize the effort of reconciling dynamic measurements from chemical processes while making the results more reliable and faster using nonlinear programming and parallel computing. With improvements in process modeling and simulation and computer technology, many process simulation packages for chemical plants and refineries have been introduced into the market. A leading commercial simulator, (HYSYS) has been chosen for this research. Modeling and simulation with HYSYS is simple and straightforward. The accessibility to different property packages, and graphical properties make modeling easier for the user. HYSYS also enables interaction with other programs, such as MATLAB. HYSYS supports a programming environment for both steady state and dynamic cases. It is capable of changing the value of process variables such as set points to introduce a step change into the system. This is very important for data reconciliation, since during the optimization step, the process model is initialized using different sets of values.

   The proposed data reconciliation algorithm consists of four steps. First, a process is chosen to model in HYSYS. To be able to verify the performance of the algorithm, a process with a known model will be modeled. Second, a data reconciliation algorithm is developed for this application. Next, a sequential optimization algorithm combined with HYSYS using Object Linking and Embedding (OLE). Finally, applicability of parallel processing using Intel processors for the developed data reconciliation algorithm will be investigated.

   Before modeling or writing any code for a given process, dynamic data reconciliation with a moving horizon window will be developed in MATLAB, which provides the means for using different toolboxes to execute various necessary operations. Then the sequential optimization algorithm will be implemented into the program, and the process model will be treated as a black box model. This is an important step to check the correctness and applicability of the algorithm. It also allows determination of the CPU time taken for each run before and after HYSYS is used. Because of the OLE application with HYSYS, there will be some communication delay.

   For the optimization routine, a sequential-modular optimization algorithm with successive quadratic programming (SQP), generalized reduced algorithm (GRG) or, successive linearized programming (SLP) will be used. A sequential modular optimization can be modified to provide separability of the algorithm. In such an algorithm about 90% of the computational time is required for the function evaluations. To take care of this problem some of the steps of the optimization problem can be solved simultaneously and in parallel. These steps include (1) approximation of the Hessian matrix using a quasi-Newton method like BFGS; (2) calculation of numerical gradients; (3) line searching strategies. Different optimization algorithms are available for the minimization of constrained objective functions. The applicability of these techniques will be investigated.

   The application of parallel computing into a nonlinear optimization algorithm will allow a robust and reliable solution to the data reconciliation problem. It may not only increase the CPU time, but also may give the opportunity to perform online application. The most important issue to be addressed is the separation of the algorithm into subproblems, which can be compiled independently and simultaneously. Typically the time associated with function calculation is 90% of the total optimization time; therefore, one of the easiest ways to implement parallel processing into nonlinear optimization is for the calculation of the various function values for the numerical derivatives of the gradient necessary for Newton-type methods. This might be a good starting point for the application of the parallel processing into the data reconciliation problem. Further possibilities and ways of application of parallel processing will be investigated, possibly linking several PC¼s together or using a PC cluster.

   Because of some difficulties with MATLAB, the data reconciliation algorithm will be converted in a Visual C++ compiler. Some of the optimization software such as GRG2 and SQP are available in FORTRAN and C languages. However, OLE applicability of C++ makes it more efficient than FORTRAN.

   
   
8. Publications.

   Dr. Edgar recently gave two named lectures on high performance information technology, which are to be published by the respective universities:

   • "Process Engineering in the 21st Century: The Impact of Information Technology," Phillips Lecture, Department of Chemical Engineering, Oklahoma State University
   • "Information Technology and Chemical Engineering Education: Evolution or Revolution?", Chemical Academy Lecture, Department of Chemical Engineering, University of Missouri ‚ Rolla

   Accompanying slides are available for the Rolla lecture. Download the PDF version or the Powerpoint version.

   (Visit Adobe to get a FREE copy of the Adobe Acrobat Reader needed to view the PDF documents.)

9. Related Web Sites
   • TWMCC (Texas-Wisconsin Modeling and Control Consortium)
     
   • Dr. Thomas F. Edgar's Homepage
     
   • Automatic Process Control (ChE 360) Homepage
     
   • Process Modeling and Control: A Vision of the Future
     (Visit Adobe to get a FREE copy of the Adobe Acrobat Reader needed to view this document.)

Previous Status Report

Back to Top

Subsurface Modeling
August, 1999

1. Principal Investigator: Mary F. Wheeler

2. Associated Investigators: Steve Bryant, Clint Dawson, Todd Arbogast, and Joe Eaton

3. Technical Support Contact: Chris McGraw

4. Project Abstract:

   Develop parallel and accurate simulators for modeling subsurface and surface flow.

5. Equipment Status:

   16 Intel machine cluster fully operational for over 18 months.

6. Research Progress:

   Application of Parssim, a flow and reactive transport simulator that was ported to the cluster in the past year, has focused on a study of diagenesis. Recently the strong coupling between cementation (conversion of the originally deposited aragonite into calcite) and permeability (the capacity of the porous medium for fluid flow) has been included in the model. Simulations for the coupled model agree with the petrographical characterization of examples of this depositional environment. This brings closer the exciting possibility of predicting reservoir quality in advance of drilling and production.

   The IPARS new generation framework for parallel reservoir simulation continues to expand the range of physical models and geologic domains that can be studied, and the Intel cluster is a primary development platform for this project. Recent advances include the introduction of a two phase model for air/water flow. This model provides an advantage over many such models used for environmental studies which do not account for the compressibility of the air phase. A new flow model that uses the discontinuous Galerkin method is also being developed and implemented on this platform.

   Both these simulators are being used in an important extension of research on the influence on flow patterns of spatial correlation of reservoir permeability. Previous studies have been limited to 2D flows because larger problems cannot be run in a reasonable amount of time on a workstation. In contrast, parallel processing with domain decomposition makes large 3D simulations feasible.

7. Related Web Site: Please visit the scientific animations — hydro B1, hydro B2, & hydro C — at http://king.ticam.utexas.edu/CCV/gallery/movies/movies.html

Previous Status Report

Back to Top

Wireless & High Performance Telecommunication Networks
August, 1999

1. Principal Investigator: Gustavo de Veciana

2. Associated Investigators:
   Takis Konstantopuolos, San-qi Li, Ed Powers, Guanghan Xu

3. Technical Support Contact:
   Gustavo de Veciana

4. Project Abstract:
   

   This project is intended to support the core of telecommunications research faculty within the Electrical and Computer Engineering Department. The equipment will support computer intensive research on physical layer (satellite,wireless) transmission issues, measurement-based channel and traffic modeling, and protocols for traffic management in high-speed networks.

5. Equipment Status:

   All the equipment has been delivered and is currently being used.

6. Research Progress.

   Ongoing research benefiting from the donated equipment include:
   1. Development of SMAQ, a web-based tool for statistical traffic matching and performance analysis of queues.

      This is a measurement-based tool which integrates traffic modeling and queueing analysis. The tool can be used in a variety of network design areas. For instance, it can be used as a traffic generator to generate traffic traces for network testing. It also can be used to find numerical solutions of the queue length and loss rate performance for transport of prioritized traffic subject to congestion control. Several application modules were built into the tool for evaluating statistical multiplexing, and performing buffer dimensioning and link bandwidth allocation. Other examples include the evaluation of transient performance during congestion and the modeling of wireless channel dynamics. A Web-based version of the SMAQ tool will be made available at http://www.apollo.ece.utexas.edu on trial basis.

      References:

      ``SMAQ: A Measurement-Based Tool for Traffic Modeling and Queueing Analysis: Part I - Design Methodologies and Software Architecture and Part II - Network Applications'', authored by S. Q. Li, S. K. Park and D. Arifler. IEEE Communications Magazine, August 1998.

      The challenge in telecommunications research has shifted towards finding effective and pragmatic methods to manage heterogeneous networks of tremendous complexity. Currently there is no effective mechanism to manage packet switched networks although the telecommunications infrastructure is migrating to packet switching technology. Traditional network management solutions fail to provide a dynamic environment in which new services can be added on-demand. This is mainly due to the fact that most network management solutions are static and proprietary. In our research, we are investigating different technologies which might create an open, dynamic management environment. We are focusing on the Java technology, especially Java Dynamic Management Kit (JDMK), its interoperation with the most widely deployed SNMP for internet management, and its ability to push or pull active code over the network. We are also investigating the Common Information Model (CIM) to model, and eXtensible Markup Language (XML) to represent management data. We are current in the preparation stage to connect our PCs to a Cisco ATM switch for on-line performance monitoring and networking resource management based on the JDMK/SNMP platform.

   2. Stability, Performance and Scalability of Traffic Routing for Communication Networks

      We are investigating approaches to routing traffic on large-scale communication networks. In particular we have focused on the interaction between flow/congestion control and routing and their impact on overall performance. Our theoretical models and network level simulations suggest interesting stability and performance issues will need to be addressed in designing the next generation of networks to meet quality of service specifications.

      Our current research focuses on new routing paradigms for the proposed differentiated services framework, and as well as generic multi-service network systems. We believe that routing mechanisms that differentiate about traffic types can present significant advantages when there are geographic or temporal inhomogeneities in the network traffic loads on the networks.

      References:

      G. de Veciana, T.-J. Lee, and T. Konstantopoulos. ``Stability and Performance Analysis of Networks Supporting Services with Rate Control-- Could the Internet be unstable?" in Proc. 18th IEEE Infocom'99, New York, NY, March 1999.

      S.H. Rhee and T. Konstantopoulos. ``Virtual path capacity allocation: user-optimal equilibrium," in Proc. 18th IEEE Infocom'99, New York, NY, March 1999.

   3. Vector Channel Propagation Study for Wireless Communications

      The donated Intel personal computers have been used to support our important project of Vector Channel Propagation Study for Wireless Communications. Wireless communications is perhaps one of the most important sector of the electronics industry. Although the growth of cellular subscribers is explosive nowadays, wireless communications still has several chief difficulties, including fast channel fading, capacity and coverage limitation. Smart antenna technology has been proposed to overcome those difficulties. Unfortunately, we do not have adequate understanding of the so-called vector channel propagation characteristics that are relevant to a smart antenna system. We have built several smart antenna testbeds to perform experimental studies of the vector channel propagation. Due to a huge amount of data collected with the testbed, it is important to have a fast personal computers with a large amount of DRAM and hard disk capacities. The donated Intel personal computers fit in this data processing role perfectly. They have been extensively used to process the real data and build a realistic vector channel propagation model.

      References:

      M. Torlak, S. Kim, and Guanghan Xu, "A capacity measure for space-division-multiple-access channels," Proc. IEEE Asilomar Conf. On Signals, Systems, and Computers, Nov. 1-4, 1998, Pacific Grove, CA.

      A. Kavak, W. Yang and Guanghan Xu, "Characterization of Fast Fading Wireless Vector Channels," Proc. IEEE Asilomar Conf. On Signals, Systems, and Computers, Nov. 1-4, 1998, Pacific Grove, CA.

   4. An investigation of RF High Power Amplifier Linearization.

      To maximize RF high power amplifier output used in communication systems, it is necessary to operate the amplifier close to saturation. Such operation, however, results in nonlinear amplitude and phase distortion, which in turn results in an increased bit error rate. Using the computer provided by Intel's Technology for Education 2000 Grant, we have designed, tested, and demonstrated a new predistorter which effectively serves to linearize the RF high power amplifier. This results in considerably enhanced performance in terms of reduction of: (a) bit error rate, (b) adjacent channel interference and spectral spreading, and (c) total degradation.

      References:

      Chi-Hao Cheng and Edward J. Powers, "Optimal Third-Order Volterra Kernel Estimation Algorithms for Nonlinear Communication System Under PSK and QAM Inputs," Proceedings of the Eighth IEEE Digital Signal Processing Workshop, Bryce, Utah, August 9-12, 1998, 4 pages (CD ROM).

      In-Seung Park and Edward J. Powers, "Compensation of Nonlinear Distortion in OFDM Systems Using a New Predistorter," Proceedings of the Ninth IEEE International Symposium on Personal, Indoor, and Mobile Radio Communications, Boston, Massachusetts, September 8-11, 1998, 5 pages (CD ROM).

      Chi-Hao Cheng and Edward J. Powers, "Fifth-Order Volterra Kernel Estimation for a Nonlinear Communication Channel with PSK and QAM Inputs," Proceedings of the 9th IEEE Signal Processing Workshop on Statistical Signal and Array Processing, Portland, Oregon, September 14-16, 1998, 4 pages (CD ROM).

      Chi-Hao Cheng and Edward J. Powers, "Intermodulation Distortion Estimation with Fifth-Order Volterra System," Proceedings of the IASTED International Conference on Signal and Image Processing, Las Vegas, Nevada, October 27-31, 1998, 4 pages.

   Industrial Interactions ---

   Southwestern Bell Corp. / Technology Resources Inc, TI Raytheon

   Related Funded Projects ---

   NSF Grant - Adaptive Resource Management for IP/ATM Hybrid Switching Systems (PI: San-qi Li)

   NSF Career Award - Analysis and Design of Hierarchical Source Routing and Embedded ATM Netowrks (PI: G. de Veciana)

   Soutwestern Bell Corp./ Technolgy Resources Inc - Traffic Control and Management (PIs: S.-Q. Li and G. de Veciana)

   National Science Foundation, Office of Naval Research, provide additional support for Prof. G. Xu's research.

   DoD Joint Services Electronics Program, AFOSR F-49620-95-C-0045. (PIs: E. Powers)

Previous Status Report

Back to Top

Distributed Simulation
August, 1999

1. Principal Investigator: Dr. Craig M. Chase, ECE Dept, ENS 436 (512) 471-7457 chase@ece.utexas.edu

2. Associated Investigator: Dr. Suzanne Barber, ECE Dept, ENS 240 (512) 471-6152 barber@ece.utexas.edu

3. Technical Support Contact: Dr. Craig M. Chase, ECE Dept, ENS 436 (512) 471-7457 chase@ece.utexas.edu

4. Project Abstract:

   The goal of this project is to research next-generation distributed simulation technology. Our principle foci are:

   a) Support for System Area Networks (SANs) and cluster computing. Tandem's ServerNet system is the interconnection fabric for our system. This 50MByte/second, switched network provides less than 10 microsecond hardware latency and support for shared-memory operations. This combination of features and performance should allow us to dramatically reduce simulation overhead in our prototype.

   b) Integrating multi-resolution and multi-domain simulations. Isolated, single-purpose simulation systems are no longer acceptable for manufacturing, process control or other simulations. Regrettably, there are no established architectures or universal object models for sharing the results from different simulations. We are exploring integration and multi-resolution strategies for manufacturing and design simulation. The goal of this work is to permit the impact of a design change upon the manufacture and production process to be immediately apparent.

5. Equipment Status:

   All year-one equipment received, installed and operational with the following exceptions:

   a) 20" monitor was never received (3 of 4 monitors were received)

   b) A disk failed in one of the dual-processor desktop workstations. This disk has recently been replaced and the machine has just been brought back on line at the time this report is being written.

6. Research Progress:

   Two independent clusters are up. The "stable" cluster is made up of the complete set of six quad-processor servers. This 24-processor farm is in nearly continuous use as simulation engine. Most of the research being supported by this cluster is in the area of computer design. This research is the target of a recently awarded, three-year, National Science Foundation grant "Performance Impact of Contemporary Programming Paradigms and Workloads" (with Dr. Lizy K. John) in the amount of $350,000.

   The experimental cluster is in the process of being converted over to the Intel/Compaq/Microsoft "Virtual Interface Architecture" specification. VI is a software architecture for very high-speed, very low-latency system area interconnect. We have recently (in July, 1998) entered into a research agreement with Tandem (a Compaq company) that provides access to the Tandem VI software development kit. Benchmarking of the "ServerNet I VI" has recently begun and preliminary results will be posted to the project web page (http://www.ece.utexas.edu/projects/distsim/) shortly. In conjunction with the benchmarking study, a distributed object model (analogous to CORBA or Microsoft's DCOM) from Xerox PARC is being ported to the experimental cluster. We are optimistic that we'll be able to demonstrate (by end August '98) that the combination of Tandem's ServerNet, with the VI architecture and the Intel cluster permits remote method invocation on distributed objects up to 100 times faster than current commercial CORBA or DCOM solutions.

Previous Status Report

Back to Top

Computational Finance Lab
August, 1999

1. Principal Investigator: Patrick Jaillet

2. Co-Principal Investigator: Stathis Tompaidis

3. Active Associated Investigators:
   Stefano Grazioli, Prabhudev Konana, Ehud Ronn, Sheridan Titman (Faculty)
   Sergey Tsyplakov, Fei Zou, Zeigham Khokher, Richardson (PhD Students)

4. Project Summary:

   The computational finance lab - UTC 1.110 - created with the help of the Intel grant, continues to generate various research activities and educational purposes in the area of computational/mathematical finance. A latest addition to the special software and hardware available in the lab is a live financial data-feed from Bridge (installed early this spring).

   As explained in the previous report, research continues to be very active in the area of pricing of commercial mortgages and problems in optimal capital structure (requiring finding solutions to 4 dimension PDEs), of determining the effectiveness of momentum strategies (using very large databases and simulation), matching pursuit applications to hedging large portfolios (heavy calculations), and pricing/hedging of complex derivatives in energy.

   Two new research projects have also started in the past two months on the pricing and hedging of derivatives in the presence of jumps, and on an equilibrium model for capital structure.

Previous Status Report

Back to Top

Electronic Commerce Networks
August, 1999

1. Project PI: Andrew Whinston

2. Research progress summary:

   We have been working on prototyping a traffic pricing system for the virtual private network (VPN) since last year. Based on a VPN testbed constructed with the computers donated by Intel a pilot VPN traffic pricing system named VTPE has been developed. The prototype operates on a scalable VPN and with reconfigurable parameters in accordance with experimentation needs. Its traffic pricing kernel system provide flexibility to allow testing different network resource scheduling techniques, such as round robin, and pricing strategies on a real-time basis. Its experimental user interface allows configuring the system easily and is able to dynamically display observations in curves. Right now, experiments have turned out many valuable results in the effectiveness of traffic pricing under different conditions, which are helpful for us to find the optimal pricing strategy and policies.

3. Related paper:
   • Lin, Zhangxi, Peng Si Ow, Dale Stahl, and Andrew B. Whinston, "Exploring Traffic Pricing for the Virtual Private Network", submitted to WITS'99.

4. Related URL:

   Intel Project Homepage: http://crec05.bus.utexas.edu/intel/

Previous Status Report

Back to Top

Enterprise Systems Management
August, 1999

1. Principal Investigator: Larry Leibrock, Patrick Jaillet

2. Associated Investigators: N/A

3. Technical Support Contact: Larry Leibrock

4. Status Report:

   Over the recent quarter a significant level of progress has been made in several inter-related projects at the Business School.

   Most of these efforts are under NDA's. (Non-Disclosure-Agreements).

   • These Intel systems have now supported deployment of over 700 notebooks and 500 desktop systems.

   • The Business School is the only academic entity in Microsoft's Office 2000, Windows 2000, Exchange 6 and Polar Early Deployment Programs. We are also in SAP's academic partnership program.

   • Enterprise Digital Certificate Servers - Public Key Infrastructures -

     We created a pilot test-bed of two X509 certificate servers that provide a repository for both public and private keys.

     Student teams researched three areas-

     1. Total cost of ownership
     2. Legal aspects of the key ceremony and
     3. A proposed deployment framework for certificate servers

   • Intel Servers - Three Tier Data Models

     A Team created and validated of an industry hierarchical data storage model for three tier data servers in a fast-Ethernet client/server environment.

   • Windows 2000 - Intel Server Business Model

     An inter-disciplinary team created and validated a TCO model that assessed both the costs and benefits of Windows 2000 in an enterprise setting.

   • Firewall Total Cost of Ownership Model

     A team tested an Intel-based firewall server in three differing policy scenarios and conducted intrusion comparative analysis.

   • Information Security Knowledge Base

     A student/faculty team continues work on the information security archive located at http://niim.bus.utexas.edu

   • Office 2000 Rapid Deployment

     We have completed testing and deployment for Office 2000. All Office 2000 collaborative servers are hosted on Intel servers.

   • Directory Project

     We are designing and testing LDAP directories on Intel enterprise servers. Directory replication models are being planned.

   • Thin Client Test bed

     We have deployed three thin client test models for then clients.

   • Advanced Encryption Standard

     (AES)

     Performance testing and models for the AES on both Pentium II and Pentium III has completed and the paper was accepted by National Institute of Standards and Technology.

   • A High-School Outreach Project (AMIGOS)

     The Business School is completing the 2nd high school project in Laredo, Texas.

     The project is on an Intel server at http://amigos.bus.utexas.edu

   • Performance Testing for Intel Servers

     Both AIM and BlueCurve performance metrics are in final completion.

   Related Projects:

   Enterprise W2K Dell/Business School Project

   The Business School/Dell Spring Summer internship involved 12 MBA students working with Microsoft Solutions Framework in planning for Microsoft 2000. Detailed reports, statements of work were completed and the project was a success for both UT-Business School and Dell. The project is now complete. We have now moved to Windows 2000 RC1 and are planning a Q1 deployment on 4 - 8 way SMP servers.

Previous Status Report

Back to Top