1.0 Introduction
--back to the top--

The University of Missouri-Columbia (MU) is seeking funds to help establish a high speed connection to the vBNS network. This connection will support present and future applications requiring high bandwidth and/or bounded latency networking services. The projects described, herein, are ongoing efforts limited by the bandwidth available through the commodity Internet (CI). However, these applications are only the beginning. Many additional novel and innovative applications involving collaborative efforts both locally as well as nationally are anticipated in the near term future.

It is vital that MU be involved and positioned to support evolving applications requiring specialized networking services. The University of Missouri-Columbia is the only AAU Research I institution in the State of Missouri. In particular, as a AAU Research I institution, it is imperative to remain at the forefront of the technological frontiers in research and education. Being at the forefront insures the institutions' ability to attract and retain high quality faculty and students striving to push the frontiers of human knowledge now and into the 21st century. Failure to pursue this path will only relegate the institution to a secondary position with minimal ability to educate and nurture our researchers and students using the latest in scientific and educational technologies.

MU is dedicated to maintaining and improving its leadership role in higher education. With the funds sought with this grant, a step toward this goal will be achieved.

MU assures its continuing support upon the completion of this project. The high quality networking services established here is essential for the vitality of the institution in future years. In the early 1980's, MU established a connection to a computer network at a time when people were asking "why was the connection needed?" Presently no one even thinks of asking this question. We are now at a point in history when we have the opportunity to take the next step in high speed networking. History tells us that it is essential to take this step and to assure its continued growth and development. Only in this way can the institution remain in the mainstream of advances in science, engineering, technology, education, research and the advancement of human knowledge. A letter supporting this initiative at MU is provided by Chancellor Richard Wallace in the Appendix of this proposal.

As part of this effort, MU will be seeking continuing support financing as part of a "Mission Enhancement" initiative with the Missouri State Legislature. The continued support of this program is vital to the University's mission to provide high quality education in the state and the mission enhancement initiative is a way to insure that the funds needed are available.

MU is a charter member of the Internet 2 (I2) initiative and is actively involved in the startup and evolution of the I2 project. MU intends to remain actively involved in I2 and will be sharing knowledge and experiences with the other I2 member institutions. This will be an additional incentive for both the institution's researchers and their collaborators to pursue the latest technologies and techniques available. In addition, the high speed connection proposed here will only further enhance our participation as the I2 initiative expands. Our participation in Internet2 both demonstrates the University?s long-term commitment to high-performance communities and will provide a powerful organizational network to support our further efforts.

MU has been in contact with MCI concerning the vBNS connection from MU to MCI's Downer's Grove vBNS POP. Due to geography this is the most economical connection point we can make at this time. We are still discussing, as an active participant, the planning of the proposed Great Plains Gigapop for I2. If this gigapop materializes and is more cost effective, we will utilize the services of this gigapop instead.

In addition to the cost sharing efforts shown in the budget, there are several opportunities (described in section 3.2) for team efforts developing in our area. While none are far enough along to be included in this proposal, we are remaining open to such opportunities as they develop further. In addition, the trend in network availability has been for the prices to go down as time passes, and we are expecting this trend to continue, hopefully allowing us to purchase more bandwidth some time within the first year of the project.

This project will follow a formal evaluation plan to monitor and disseminate information on all aspects of the vBNS connection activity. This evaluation plan includes: (1) Advisory Board of regional and local experts, (2) Workshops and conferences organized by the MU vBNS groups, and (3) User Group evaluation mechanism. All of these avenues will be used to document and provide sources of information dissemination. We are also in the process of setting up a Web-accessible set of pages for our vBNS and I2 activities. This information will be readily available to the Internet community as a whole. The MU Home Page (http://www.missouri.edu) will be enlarged to incorporate links to the HPC activities at MU so this material can be readily found and utilized. Presently, our work with the I2 initiative is maintained on the Web (http://www.cecs.missouri.edu/i2). These pages will be enlarged to encompass our vBNS efforts as well. As part of the vBNS engineering and application group's activities, the high speed connection will be regularly monitored and evaluated for performance, acceptable use and utilization. These data will be reported to various groups both locally and peers throughout the network. In addition, any results and findings obtained by the research and application development activities will be disseminated through journal articles, conference publications, Web pages and the like. This activity will remain an active part of the HPC connection group's activities. Similarly, our experience will be shared with groups such as the I2 engineering and application groups as we participate with their efforts.

MOREnet's Network Operations Center has seven years experience in managing network infrastructure, and will be providing support to ensure that the connection to the vBNS is operating at peak efficiency.

This section describes the planned meritorious connection implementation strategy. Beginning with a description of the University of Missouri-Columbia?s current connection to the commodity Internet, we conclude with an overview of the planned connection to MCI?s Downer?s Grove vBNS point-of-presence.

MU has been involved with various national computer networks since 1983; (i.e., BITnet (1983), Internet (1987)). For a period of time, MU was the gateway for many academic institutions in the mid-West and Texas on the BITnet network. During the intervening years, MU has actively supported networking service to support the institution's connectivity to peers around the country. In the late 1980's, MU took the lead in facilitating network access for educational institutions (K-12 and higher education) throughout the State of Missouri. This leadership role resulted in the formation of a separate organization in the institution (MOREnet) to be focused on the state-wide connectivity issues. Presently, MU has a campus organization to provide network services locally and this organization interfaces with MOREnet to provide connectivity to the external networks.

The University of Missouri-Columbia, as a charter member of the Missouri Research and Education Network (MOREnet), has committed traffic rate to MOREnet of three (3) megabits per second. This service level has been in place since April, 1996. This service level will be upgraded to ten (10) megabits per second this summer via an ATM connection and can be easily reconfigured for additional service as the campus commodity Internet requirements grow.

Since its founding in 1990, MOREnet's growth has kept pace with the enormous growth of the Internet. Today, MOREnet is the premiere Network Service Provider serving Missouri's Non-Profit and Educational community, with over 400 dedicated connections and nearly 50 higher education members. As a charter member of MOREnet, the University of Missouri-Columbia has full access to MOREnet's expertise and support.

2.2 Planned Meritorious Connection Implementation
--back to the top--

It is expected the high performance connection will be implemented as a point-to-point facility over a DS-3 leased line. The meritorious connection will not be expected, nor provisioned for carrying any commodity Internet traffic. Figure 1 below shows the planned implementation.

Specific implementation notes are enumerated below.

1) All campus Virtual Circuits (VC) connect to the HPC Router. The interface from the HPC Router runs at OC-3 as does the ATM Switch connection to the Campus ATM Infrastructure. This router acts as the gateway for all off campus destinations.

2) The HPC Router has a Border Gateway Protocol (BGP-4) peer connection to the vBNS and receives route announcements of other meritorious network connections. Packets destined for the vBNS from qualified hosts/subnets are delivered at the full rate of the vBNS access facility.

3) The HPC Router has a VC to the MOREnet Commodity Router. Routes to vBNS locations are explicitly filtered out of routing exchanges between the HPC Router and the MOREnet Commodity Router.

4) Destinations unknown to the HPC router are presumed to be commodity Internet locations and use the MOREnet Commodity Router as default. Packets destined for the Commodity Internet are delivered at the rate contracted for Commodity access.

The University of Missouri-Columbia will work closely with both MCI and MOREnet regarding routing policies and implementation issues to insure that all in-force Acceptable Usage Policies are adhered to.

The initial implementation was specifically chosen to terminate in an ATM switch rather than an ATM enabled router. We believe this strategy offers significant flexibility for the future.

With this technique Virtual Circuits may be constructed between campus based host systems and systems located at offsite. For example, a PVC could be built between a location in the University of Missouri-Columbia?s Computer Engineering and Computer Science Department and a server at one of the National SuperComputer centers. Quality of Service parameters and the absence of level 3 IP routers between the client/server pair would offer the most consistent, tuneable, and guaranteed service delivery possible. This scenario will be pursued with respective organizational support personnel on an application specific basis.

The efforts outlined in this proposal involve a range of activities and services that position MU in the mainstream of the network infrastructure. As Missouri's only AAU Research I institution, we will lead the way for other educational institutions in the state as well as nationally to set the standards for quality and effectiveness in the search for knowledge. The networking services and applications will both utilize resources from afar as well as provide services to the community as a whole. In addition, our collaborations with peers around the country can only enhance the efforts of all and reduce unnecessary duplication of effort. It will lead to a synergistic "system" from which everyone will benefit.

MU is best positioned in the State of Missouri to lead this effort as well as support its sister institutions throughout the state. This includes outreach to both minority and women's institutions. MU has taken this lead in the past (e.g., MOREnet) and will continue to do so in the future. This proposal is a natural outgrowth of our taking the lead within the state of Missouri.

Meritorious applications will schedule QoS PVC's between MU and remote sites, both vBNS and others as available. The PVC QoS mechanism is attractive in that it is immediately available, however it lacks the flexibility of emerging techniques such as SVCs and RSVP. We believe Switched Virtual Circuits (SVC) are the clear goal for anywhere, anytime QoS. However, the ATM industry is still working towards delivery and interoperability of SVC services. As middleware, QoS services are developed for use in applications (such as RSVP and IPv6 QoS), access to ATM-based PVC and SVC services will be made available. Several faculty in the CECS Department (e.g., Springer and Joshi) are working on middleware for networked/distributed systems which address the QoS issue. We will integrate the results of their research into our QoS mechanisms. The CECS Department's dedicated ATM test facility (with an IBM 8260 switch) can be used to test these policies before being added. In addition, MU and MOREnet support staff will work with vendors and service providers to implement SVC support as opportunities prove themselves practical and reliable.

The decision to use DS-3 rather that OC-3c is based largely on connectivity cost. Although MU's cable plant supports higher bandwidth, the cost of this level of service is prohibitive, both for the fiscal guidelines of this program announcement and for ongoing service after the contract is complete. The most cost-effective link to vBNS/Internet is through MCI, with a vBNS node in Downer's Grove, Illinois. Working with the best quotes from vBNS Engineering, the yearly line charge for OC-3c service is some $600K, whereas the yearly charge for DS-3 service is $210K. Clearly the OC-3c cost is beyond the scope of the current proposal, and furthermore would require unacceptably large costs to continue this service, even if the program were funded at that level. In addition, we feel that DS-3 will provide us with a much better gauge of evolving requirements among the current research applications. It is clear that an OC-3c circuit presently would be underutilized much of the time. A DS-3 circuit will serve as a tool to better estimate the benefit of higher-bandwidth service. We will pursue with MCI (and others) the possibility of defining a new service normally rate-limited to DS-3 speeds, but carried over an OC-3c circuit. Applications requiring particularly high-bandwidth could then schedule Virtual Circuits not choked to the aggregate DS-3 rate (accomplished by ATM Traffic Shaping in the local ATM switch), and would incur an incremental, incidental charge based on usage. This mechanism is far more attractive than current conventional OC-3c service, which would be driven beyond DS-3 rates only intermittently. When bandwidth demands exceed capacity, additional bandwidth will be acquired to support the incremental needs of the institution.

Accordingly we propose to initially acquire a vBNS connection with DS-3 service provided by MCI. This DS-3 connection would be devoted, initially, to the meritorious applications. The cost of DS-3 service, after the contract completion, will be borne by the MU Campus Computing and funding at that level has been agreed to by the campus management, once the two year startup period is complete.

The University of Missouri-Columbia is enhancing its current engineering plan to upgrade significant sections of the campus network to take advantage of this increased connectivity. The fiber-optic upgrade of the campus backbone to an OC-3c, ATM-based network and in-building cable plants are being aggressively pursued so that users will see remote resources as an extension of their local computer. In addition, during the next calendar year, it is anticipated that parts of the backbone network will be upgraded to OC-12 to adequately support both the HPC applications as well as the intra-campus traffic.

The networking infrastructure at MU is evolving and is in a position to support any member of the MU research community with the ability to fully utilize high speed connections to the network. An ATM backbone network is being deployed to provide a wide range of networking capabilities from every office on the campus. This network is replacing a legacy token ring backbone network that has served the campus extremely well for many years. However, growth in usage and demand has reached a point that the bandwidth requirements no longer are sufficient to service the community adequately. The ATM backbone network at MU consists of four major switching hubs located at strategic points on the campus. These hubs are IBM 8260 ATM switches that can support a variety of connection types. Due to the demands for high speed networking in the Computer Engineering and Computer Science Department (CECS), an additional backbone connected ATM switch has been installed in the CECS Department which provides a direct OC3 connection to the campus? ATM backbone via an IBM 8285 ATM switch and IBM 8274 expansion box. Figure 2 (below) shows the evolving ATM backbone. The ATM switch labeled MOREnet provides the campus access to the vBNS and the commodity Internet, and the meritorious connections emanating from the MU campus will be provided a direct path through the campus infrastructure to the MOREnet switch. Thus, the projects described in section 3.1 only lack the appropriate high speed connection between MU and the vBNS network we need to use for our collaborations elsewhere.

The projects described in this section are indicative, but not exhaustive, of the types of activities that demand high bandwidth, low latency computer network access. Without such access, these projects would be extremely time consuming or infeasible. The projects described are not simply projects in the planning stages, but are ongoing and evolutionary projects that would be greatly enhanced by the availability of high speed connections between the collaborators at the University of Missouri-Columbia (MU) and their colleagues at other institutions and government laboratories. Immediate use of the high speed connections would occur as soon as the connection is established.

While each of the projects described involve applications being developed by various research teams and collaborations, there are some common themes in each of the projects that would permit many other constituencies to contribute to these projects both at MU and elsewhere. And, as with any evolving technology, there are many additional researchers and activities that would benefit from the early installation and deployment carried out by the early users of the technology. Among the themes in these projects is not only high speed network access, but also basic research into information organization, location and access, visualization techniques, collaborative research, educational delivery systems and techniques, and the development of the fundamental tools needed to facilitate the building of other applications. Thus, the described projects are only the beginning of the road that will eventually touch every member of the MU community as we move into the 21st century. It is imperative that we take this step and insure we will be positioned to remain among the forefront educational institutions in our ability to educate and train the citizens in the next century.

The projects described require the involvement of multi-disciplinary groups that cross traditional boundaries of departments and colleges. Most of the colleges on the MU campus are involved with these projects. As these projects evolve and other projects are initiated, the entire fabric of the institution will be touched by the abilities and capabilities that will result from the high speed connections existing on-campus as well as throughout the global networks. Likewise, there are many opportunities for joint collaboration to work on common problems faced by all of these projects, and the participants are poised to work independently or jointly when it is advantageous to do so. However, it is critical to these projects and to the mission of the institution that we have the ability to carry out these projects and can utilize the connection to the vBNS network to interact with and collaborate with our peers located throughout the nation.

Project Leaders: Su-Shing Chen, G. Springer, A. Joshi, J. Keller and X. Zhuang, CECS

Through the advent of high performance computing and communications networks, networked computer systems have emerged to be the dominant paradigm, and information has become widely available in digital formats, creating the phenomenon of information glut or information overload. Computational intelligence will be useful to reduce the information overload and to build the next-generation, more efficient, networked information systems.

The MINIS Project intends to explore this new frontier of intelligent networked information systems by integrating computational intelligence methodologies and networked information. The project is the research and development of a "vBNS-based knowledge network" - a collaborative information environment for the learning, managing, reasoning, and visualizing of a broad spectrum of networked multimedia information.

We will develop networked information technology research by extending computational intelligence methodologies from a single computer platform to a large-scale, intelligent, networked information system. We have done extensive research in fuzzy logic, neural networks, and cognitive maps. The main task is to combine and integrate these computationally intelligent algorithms, and to extend them further to distributed computational intelligence in the general framework of networked information systems. The MINIS Project will include a large number of heterogeneous information and knowledge bases managed by a host of intelligent agents distributed over a large computer and communications network. Intelligent agents may be humans, humans interacting with computers, humans working with computer support, and computer systems performing tasks with or without human intervention. Intelligent, networked information systems are evolved from several disjoint technologies such as databases, computational intelligence, natural language processing, and computer networks. It is easy to imagine the benefits of powerful knowledge base systems (e.g., medical diagnostic systems) with efficient access to several large information bases, and of large databases (e.g. airline reservation systems) with added intelligence (e.g. scheduling and re-scheduling travel reservations considering individual's preferences). The MINIS Project will be considerably more powerful than such extrapolations of existing system concepts.

The design and evolution of the MINIS Project will require the extension and integration of several technologies (databases, computational intelligence, computer networks, etc.). Database systems will supply information management techniques, particularly for distributed databases, as well as efficient implementation techniques for information bases. Computational intelligence will contribute distributed problem solving, knowledge representation and reasoning techniques. The vBNS network connection will provide the necessary underlying communications and interconnections bandwidth for the "knowledge network".

In order to provide a viable test bed for the MINIS knowledge network, our collaboration with other MU research groups will provide the necessary large collections of digital patient records (with names and identities removed), on-line stock market data, and medical images to be tested in a distributed, network-based environment. Our collaborators include the Department of Healthcare Information Systems, College of Business and Public Administrations, and College of Veterinary Medicine. There are also several ongoing collaborative projects between the MINIS group and the University of Illinois at Urbana-Champaign, University of Florida at Gainesville, and University of Maryland at College Park. We need the vBNS connection to communicate, for example, 3-D rapid prototyping design files, interactive video and audio files, and networked knowledge reasoning. Technical issues such as network data volumes, latency and scalability in addition to the issues of computational intelligence and user interaction can be actively measured in the MINIS test bed.

Project Leaders: K. Palaniappan, CECS; J. Engel, Geological Sciences; C. Fulcher, Center for Agricultural Resources; T. Haithcoat, Geo. Resource Center; D. Diamond, MoRAP

The MU Regional Validation Center will consist of ground station hardware for receiving both GOES-8 and GOES-9 environmental satellite GVAR data using real-time direct readout systems. The system consists of a 12' Paraclipse satellite dish, with Integrated Feed/Downconverters, ISA bus-based BPSK receiver, ISA bus-based Bit Frame Synchronizer and Frame Formatter. Processing of the GVAR data stream signal at approximately one gigabyte per hour is accomplished using a Quorum GVAR Data Server based on an Intel Advanced ZE Minitower 166 MHz Pentium computer.

This high volume data stream will be shared via high speed networks such as 100BaseT and ATM with other research groups on campus and in the vicinity including the Departments of Computer Engineering and Computer Science, Electrical Engineering, Soil and Atmospheric Science, Center for Agricultural Resources Environmental Systems (CARES) in the Department of Agricultural Economics, Geographic Resources Center (GRC) in the Department of Geography, Geology, Missouri Resources Assessment Project (MORAP), Agronomy, and others. The real-time GOES satellite data will also be shared with collaborators on joint projects at other institutions such as NASA Goddard Space Flight Center, University of Wisconsin-Madison, University of Washington-Seattle, University of Arizona-Tucson, and the University of Tennessee.

Instrumentation in terms of GOES-8 and GOES-9 satellite GVAR data receiving systems, supercomputer class computational servers and visualization platforms will be used to explore new computational techniques for: (1) processing real-time remote sensing data, (2) distributed client-server visualization of extremely large remote sensing datasets (such as digital terrain elevation data) with automatic level-of-detail management using multi-resolution and graph observations with model data, (4) information visualization of multidimensional time varying datasets, (5) data exploration and learning tasks applied to inter-comparison of remote sensing data, and (6) ground and in-situ based validation of automatic cloud height and cloud motion vector estimation for input to numerical weather modeling and improving short range forecasting. Many of these analysis algorithms require access to supercomputers at NASA GSFC and JPL for timely processing of real-time data.

Science areas that can be investigated include: (1) impact of urbanization on watersheds, (2) improving short term forecasting of severe weather such as tornadoes and flooding potential using MM5 assimilated satellite observations, (3) predicting surface skin temperature and heat stress index using MM5 model forecasts with assimilated satellite observations for application to energy demand modeling by power utility companies and livestock management (such as effect on milk production) by agribusinesses, (4) correlating field measurements of plant health and crop stress with GOES remote sensing data similar to the AVHRR-based NDVI ratio at the state and regional scale, (5) linking remote sensing observations with economic impact and decision support tools for application to commodity pricing, etc., (6) combining multiresolution, multi-source datasets with automatic pixel classification for natural resource assessment and conservation, (7) estimating soil moisture at the regional scale for assessing environmental deterioration due to soil erosion and flooding, and (8) intelligent management of state transportation facilities through integration with real-time weather data.

The MU Regional Validation Center will be an ideal test bed for developing innovative teaching and educational curricula related to remote sensing and real-time satellite data handling in the environmental and computational science and engineering fields. The Regional Validation Center test bed can also be used as a training center in real-time satellite data processing by state and national agencies and consortia such as MORAP.

The campus wide ATM backbone and vBNS connection will provide fast network access to the real-time satellite data acquired by the direct-readout data system. In addition, the real-time data will also be routinely made available for wide area access using Web protocols via the Internet. All of these activities require high bandwidth and low latency capabilities. Without such capabilities, this project would not be feasible.

Project Leader: Su-Shing Chen, CECS

An Object-Oriented (OO) framework for digital libraries, human end-user collaboration and their integration has been proposed in our earlier papers. The information space of digital libraries is integrated with the digital shared workspace of human end-user collaboration. A digital shared workspace is where coordination of computers and collaboration of human end-users take place for a variety of networked applications and services. In a digital shared workspace which spans the group memory space of many computers over a wide area network, shared multimedia objects will be annotated, applied, created, consumed, indexed, and stored. The proposed OO framework of digital libraries and human end-user collaboration depends on an Information Encapsulation Principle extending the Data Encapsulation Principle of object-oriented programming to information access and utilization. Information is captured as information objects and collaborated information is represented as shared objects. In this framework, information access - query, search, and retrieval - is represented by well-defined operations on the object hierarchies in the information space and the digital shared workspace. There will be three alternative kinds of operations: complex objects, functional mappings and intelligent agent architectures. Information utilization, including collaboration, is also represented by the three kinds of operations.

The research develops a common content-based indexing scheme of information objects in digital libraries and shared objects in digital shared workspaces. Content-based indexing in digital libraries is an important ongoing research issue. Although there are many results, it is still a very difficult issue because of the complexity and scalability of information content. Our content-based indexing relies on conceptual graphs of Meta information. These graphs sufficiently represent the information content of objects in digital libraries and shared workspace. Our framework provides a formalism to control, emulate, manage, model, and simulate the information access and utilization process and the human end-user behavior in the digital shared workspace. It helps to understand human collaborative learning and learner-system-information interaction, and serve as a means of validating evaluation field data of human learning.

The collaborative learning environment is called the "Classroom of the 21st Century" in the Department of Computer Engineering and Computer Science. It develops a large-scale networked information system for project-oriented undergraduate education in computer science and engineering with the state-of-the-art facility (e.g., video-conferencing, white boards, DCOM and CORBA-based courseware, hypertext-based learning tools, annotation and evaluation software tools). The vision of the College of Engineering is to develop project-oriented undergraduate engineering education for ABET and CSAB accreditation in the 21st Century. The responsibility of the department is to develop the component in computer science and engineering. Due to the unique role of our department in modern computer-based engineering education, what is developed in our project-oriented undergraduate courses will be highly relevant to other engineering departments. Project-oriented courses will provide research/education training for our undergraduate students in computer science and computer engineering programs. This will also provide integrated research/education opportunities for our faculty members, because their research will be directly connected to their teaching. Project-oriented collaborative courseware will be indexed, managed and stored in a networked repository of our digital library of educational materials. Students will become not only users, but also developers of the "Classroom of the 21st Century" collaborative learning system.

Although much of the digital library development and high speed network service is local to MU, our collaboration with Bruce Schatz and H. Chen of the NSF/DARPA/NASA Digital Library Group at NCSA will be essential as a source of experience and expertise to be exchanged. As the project evolves, the necessity of vBNS access to NCSA repositories and collaborators will be critical. The objective is to interactively access and utilize their Digital Engineering Library at the University of Illinois by MU engineering students.

Project Leaders: K. Palaniappan, X. Zhuang, S. Chen, and A. Joshi, CECS

The CISS (Collaborative Information Spread Sheet) Project is concerned with the accessing (query, search, retrieval), application, communication, computation, indexing, storing and visualization of geo-spatial information for collaborative research, research training and education activities in the scientific community. A test bed using NASA EOS image data will be established among three locations: University of Missouri-Columbia, NASA Goddard and JPL.

The U.S. Global Change Program over the next decade (beginning in 1997) will produce an enormous volume of observational and processed data each day. NASA's component of the U.S. Global Change Program is Mission to Planet Earth utilizing satellite remote sensing instruments that will produce several terabytes of data per day. The need to develop automatic algorithms that take advantage of parallel supercomputers and distributed processing, using heterogeneous distributed remote sensing data archives, with new interactive visualization techniques for combining 2-D and 3-D displays is essential to make effective use of the anticipated terabyte data volumes from the Global Change Program.

The collaboration with NASA Goddard requires accessing, processing and analyzing meteorological data from the GOES-8 and GOES-9 geo-stationary satellites (2GB/hour). Access to this real time data is needed at the MU to do processing and analysis for extracting cloud structure and motion information. The data is currently not being archived at Goddard and is overwritten approximately every 24-hours with new satellite information. Consequently, the full GOES datasets need to be transferred to the MU using a high speed network in a timely fashion during the short duration that the full resolution data is available on the GOES server (camille.gsfc.nasa.gov) at NASA Goddard. When the MU Regional Validation Center is fully established (see project 3.2.2 above), there will be bi-directional flow of data between Goddard and MU as the regional sectors of interest are shared (i.e., East Coast and hurricane alley at Goddard and Great Plains at MU).

The goal of the CISS Project is the research and development of a collaborative computing and communications environment for learning, managing, reasoning, and visualizing a broad spectrum of networked geo-spatial and remote sensing information. A multi-window architecture will be developed using OpenGL on Silicon Graphics workstations extending previous work in developing the IISS (Interactive Image SpreadSheet) Environment to our proposed CISS Environment for the networked visualization of geo-spatial information space. Instead of the usual database indexing and single screen visualization of geo-spatial information, our CISS supports heterogeneous database management, collaborative multi-window visualization and collaborative work indexing. A collaborative immersive workbench (e.g., the Fakespace Workbench) will be used to prototype an innovative scientific visualization environment by extending the cellular multi-window display architecture of the IISS to a cellular collaborative Virtual Reality (VR) architecture.

Supporting interactive collaborative VR sessions across three sites in different parts of the country will require very high bandwidth. The model datasets are typically hundreds to thousands of megabytes in size. These will need to be replicated and shared during a collaborative session. Furthermore, real-time video and audio streams need to be managed during the collaboration.

Project Leaders: Dale Musser and James Laffey, Center for Technology Innovations in Education

The ISJS is a network-based software system for supporting new forms of teaching and learning that emphasize field experiences, inquiry and reflection. The ISJS has many capabilities, but a good starting point is to think of it as an electronic tool for keeping and sharing a reflective journal containing multimedia entries and for real-time and asynchronous communication and collaboration. The ISJS system utilizes a web-based client, a server for processing journal system transactions, and a database system that supports the storage of media objects and index information. In the Fall 1997 semester, the ISJS will support 600 students and faculty and is expected to grow to over 1200 users in the next three years.

Although ISJS can be utilized in a range of learning communities, it has been implemented in the MU College of Education (COE) to support the students who are enrolled in the Undergraduate Teacher Education Program. These students use the tool to chronicle their learning experiences, which extend from the classroom at the University to field experiences (observations, aiding, and student teaching) at partner K-12 schools. The journal editor allows students to create multimedia documents that contain text, images, sounds, URL's, and when network speeds support it, video. These documents illustrate student experiences, capture their thoughts and observations regarding these experiences, and document questions and concerns that the student may have. Once a journal entry has been created, it is submitted via the client to the server for storage within the journal system database. Other members of the journaling community can access the journal entries and respond to them through an append mechanism. Appends, like journal entries, are multimedia documents. ISJS allows the community of students, teachers, and mentors to share experiences and receive expert advice. In the past, students, while engaged in field experiences, were isolated from each other and from their teachers and mentors. ISJS removes that isolation and creates the ability for the members of the community to communicate and collaborate with each other.

In addition to the ability to journal, ISJS provides access to the web, custom information bases, real-time chat, newsgroups, e-mail, and multimedia profiles for each of the journal authors. The journal author profile provides information about an author, access to their web page, and a context for chatting with or e-mailing the author. With a high bandwidth connection the real-time chat function could be developed to support multiple media types such as images, audio, and video.

The user experience is extremely important in the design and operation of ISJS. The system needs to support storage and retrieval of multimedia documents in a timely fashion. A student is encouraged to create media rich journal entries to capture and articulate their experiences. But, network speeds can discourage them from doing so because of data transfer rates. What sets the ISJS software apart from other Internet applications is its support for two-way communication of multimedia information and asynchronous transfer of multiple documents. The multi-threaded nature of both the client and the server connections allows the client to simultaneously upload and download multiple, multi-media documents. This requires a high bandwidth. Experience has shown that real-time communications capabilities, such as those provided by the ISJS chat feature, are useful in bringing groups of learners together to engage in a dialogue to share information and discuss problems and their solutions. The limiting factor in providing media-rich, real-time experiences is the speed of the network connection. As a result, an important component of this work is to investigate the quality improvement of the user experience when working in an unconstrained networking environment.

A collaboration has been formed with Dr. Keith Hall from the Ohio State University who is investigating the use of the Internet as an asynchronous tool for offering courses beyond the campus. ISJS will be one of the tools that his team will explore beginning the Fall 1997. In addition, the opportunity to use ISJS in the OSU teacher education program is being explored. The goal is to eventually provide ISJS to a range of education and research institutions. Given the needs of the system, the first institutions will be those that have access to high speed networks.

As a community of learners, teachers, and mentors engage in the journaling process, they build an information base that can be useful to members of other similar communities. For example, the content of the journal system at MU could be very useful to students in other teacher education programs. A student in the teacher education program at Ohio State could access the MU system to learn what it is like to student teach or aid in a K-12 classroom. The student at Ohio State could converse with a student at MU to gain a deeper, and more personal, understanding of the experience. The problems, solutions, and insights of learning to become a teacher are embedded in the journal entries and their appends in the journal system.

The development of the ISJS software has not only addressed innovations in how technology can be applied to support teaching and learning in new ways, but advances the technologies themselves. These advances address methods of media storage, architectures and protocols utilized by the server, techniques for locating documents and media objects that match user needs and interests, and methods of creating multimedia documents. Improved network speeds will allow the ISJS research efforts to push the limits of what can be delivered to the user and therefore push advances in all of the technologies underlying the development of the Interactive Shared Journal System.

Project Leaders: Gordon K. Springer, CECS; J. Forrester, DNA Core Facility; T. Patrick, Medical Informatics Group

Genetic databases (DNA and protein) continue to grow at an ever increasing rate. In addition the number of databases, some of which contain information not available from other sources, is also increasing. Due to the volume of data, researchers cannot, in many cases, peruse one or more genome databases easily or efficiently when attempting to follow a research lead. The problem is two-fold. First, the location, content and access to the various genome databases may be difficult for a researcher to keep track of, especially when involved a discipline specific line of thought. Second, maintaining a local copy of all databases requires disk storage space that is becoming hard for many to afford. Retrieving and reformatting files is also a tedious and time consuming for researchers.

In making decisions about what sequences to include in an analysis, the researcher needs to have ready access to the actual sequence data and the data must be in a format that can be input to a particular analysis program. Thus, widely distributed access to standard genome databases as well as specialized databases (e.g., gene mapping databases, specialized disease databases such as AIDS, etc) must be available and accessible in nearly real-time fashion. These sequences need to be "imported", formatted and input into an analysis program without the researcher having to actually manipulate the data themselves. That is, an analysis program may be directed to retrieve, reformat and use a given sequence from a remote genome database as if the data were contained in a local repository. A method of making the data available from a few sites that appear to be local and with built in format conversion would eliminate the need to maintain redundant copies of the databases at numerous sites. This method gives rise to the concept of a "virtual local database" wherein the physical location of the data is not known by the user. However, the data can be viewed, processed or utilized as if it were physically housed locally. We propose to develop such a capability so that the enormous wealth of genome data is readily available to the researcher regardless of its physical location. In order to accomplish this goal, access to low latency network bandwidth is required.

In prior HPCC work, a prototype system was developed to integrate a wide collection of biomedical analysis and research tools in a seamless fashion. This prototype system incorporated analysis tools that ran at the PSC, SDSC, and NIH as well as locally. This system automatically combines information sources based on the data paths between them. A data path exists between two sources of information when it is possible to extract data from one information source and then to use that data, under a possibly null transformation, to drive retrieval of information from the other information source. We call this approach to combining information sources data path integration. All of the information sources are located via location independent identifiers whose instances can be found on the network dynamically.

By extending our work with data path integration and the evolving efforts to provide location independent access to data, the concept of the "virtual local database" can become a reality and provide researchers with enormously powerful tools with which to pursue their scientific investigations. Utilizing our collaborative effort with the biomedical group at PSC, we can begin to provide the environment and the techniques needed to locate and access both the data repositories and extremely powerful computational tools to analyze the data from the researcher's own desktop.

With the growth of the Internet in the last few years, this work has been severely limited due to the ever-increasing delays encountered on the network. The network traversal time to make simple analysis requests at PSC has increased by several orders of magnitude. This has severely hindered the natural interactive behavior of the system as well as the time needed to transfer analysis results (amounting to many megabytes) back to the waiting researcher. With a vBNS connection, this limitation can be eliminated.

Project Leader: Anupam Joshi, CECS

The evolution of the Internet into the Global Information Infrastructure (GII) is impacting many institutions of life in general, and the way we view computing in particular. The realization of rapid multidisciplinary problem solving or prototyping is the new grand challenge for Computational Science and Engineering (CS&E). The scientific computing software of tomorrow, developed with such applications in mind, will use collaborating software systems and agent based techniques to build systems from software components which run on heterogeneous, networked platforms. It will allow wholesale reuse of legacy software and provide a natural approach to parallel and distributed problem solving.

The advent of fast networks (vBNS, I2, NGI), and distributed computing infrastructure like the CORBA/DCE/NCA/Servelets etc. are making the concept "The Network is the Computer" a reality. The "network" interconnects heterogeneous computational units (ranging from workstations to massively parallel machines). An infrastructure needs to be developed to allow scientific computing applications to use resources and services from many sources spread across the network. The system, however, needs to be network and evolution transparent to the user. In other words, the user must be presented an abstraction of the underlying networked infrastructure as a single "meta-computer", which can be accessed using a Problem Solving Environment (PSE).

In some sense, a primitive version of the system we described above can be said to exist today. Several software libraries for scientific computing are available, such as Netlib, Lapack/ScaLapack etc. There are even some attempts to make such systems accessible over the web, such as Web //ELLPACK(from Purdue) and NetSolve(from UTK/ORNL). Software such as GAMS and PYTHIA exists which enables users to chose the right software for their problem. The user can locate and download the appropriate software components and their installation and use instructions, install the software, compile and (possibly) port it, and learn how to invoke it appropriately. These are clearly not tasks for the uninitiated, such as most application scientists are wont to be. The systems today are the equivalent of programming ENIAC, which required direct manipulation of connecting wires. Research is needed to provide systems that will abstract away the detail of the underlying networked system from the user, who should be able to interact with this system in the application domain. Our long-term research focus is to create PSEs to address this task.

This effort involves a collaborative effort with Rice's PSE group at Purdue and will build on our prior work in creating the PYTHIA advisory systems and SciAgents collaborative agents system for PDEs. The issues we address solve only some of the problems associated with realizing the vision of networked scientific computing. However, it is a starting point and would not be feasible without high-speed access to a range of computational resources and agents distributed throughout the network.

To close this section on applications, it is appropriate to note that many researchers have been working independently on their respective projects. However, the need to acquire access to high speed networking services has brought these people together to work toward a common goal. Even as these projects were being described for this proposal, other researchers have begun to develop other collaborations and projects. For example, it has recently been brought to our attention that a joint project involving at least four colleges on the MU campus, Washington University in Saint Louis and the Monsanto Corporation, also in Saint Louis, is currently being defined. This project has interactive video and distance learning components requiring significant networking bandwidth services. Although this project is not described here, it only points out the fact that there is an enormous latent demand for networking services capable of supporting the kinds of applications and collaborations that cannot be considered using the lower bandwidth abilities of the Internet. MU is positioned and ready to move forward and be actively involved in this exciting pursuit.