NII White Papers--Table of Contents NII White Papers--Chapter 7 NII White Papers--Chapter 9

The Electronic Universe: Network Delivery of Data, Science, and Discovery

Gregory Bothun , University of Oregon
Jim Elias , US West Communications
Randolph G. Foldvik, US West Communications
Oliver McBryan , University of Colorado

ABSTRACT

When technologists discuss the "information highway," their conversations usually revolve around high-end technologies such as "broadband" and "asynchronous transfer mode" (ATM). Unfortunately, this focus on the exclusive use of very high bandwidth solutions fails to reckon with the limitations confronting many learning institutions, and especially most of the K-12 environment, in gaining access to the "highway" or even exploring their options.

This paper describes a set of advanced networking tests that were conducted in Oregon and Colorado during 1994 and early 1995. High-speed ATM capabilities were effectively merged with readily available T1-based connectivity to deliver affordable information highway capabilities to K-12 educational institutions. Innovative collaborations between the university and K-12 level served to accelerate the introduction of the new technology to the K-12 level. A range of application capabilities were effectively delivered over the integrated ATM/T1-based infrastructure. Multimedia workstations were used for video to the desktop, desktop video teleconferencing, and access to multimedia servers via Mosaic and the World Wide Web.

The tests that were conducted, and the results and trends, are covered in detail in this paper. These findings are then used to suggest technology deployment models for the next 5 to 7 years.

STATEMENT OF THE PROBLEM

The concept of "network everywhere" is a goal that is achievable through cooperation between education, industry, telecom carriers, and the government. But why is this important? Its importance lies in a paradigm shift in how the individual interacts with and views the world. Almost 300 years ago, the philosopher John Amos Comenius penned these words: "It is lamentable, utterly unjust and insulting that while all men are admitted to God's theatre, all are not given the chance of looking at everything."

This is a description of an inclusive philosophy of education that has never really been implemented. Instead, education has focused intensely upon the exclusive theme of "educating each according to his or her needs." Ironically, K-12 education is now in a position of reviving this old idea through the use of new technology: the high-speed network. High-speed network connectivity will provide the K-12 classroom with a vast array of resources in all disciplines.

Is access to high-speed networking a reasonable goal for the K-12 classroom? Unfortunately, the high media profile of fiber optics, high bandwidth applications, and advanced communications technologies has distracted many institutions from exploring other options for access. This is the case with most of the K-12 environment. There may be other ways for them to deal with some of the limitations confronting them. This is particularly true in most of the western states, where low population densities and correspondingly high deployment costs prevail. These challenges created an opportunity to explore the limits and possibilities of defining a high-speed ATM trial network for the educational community that could be integrated with the existing geographically dispersed public network. This trial environment offered an opportunity to explore existing biases about fiber optic and high-speed networks, overcome the constraints of distance and dispersed population centers, and bridge the ever widening educational gap between technological "haves" and "have nots."

Education and lifelong learning are primary keys to societal success on both a personal and a national level. As our information-oriented society continues to move rapidly forward, tremendous pressures are exerted on schools at every level of learning. The rate of information acquisition in learning and scientific inquiry is astounding, making it extremely difficult for teachers and school districts to keep up. The problem is particularly difficult for both rural and urban schools with limited enrollments, diverse student populations, and constrained budgets. Many students in these types of schools are being taught advanced topics by minimally qualified teachers with outdated textbooks. These students are placed at a decided disadvantage if they advance to the college level and are forced to compete with students coming from schools with advanced curricula based on more current information. Worse yet, certain subjects may not even be taught in some schools due to lack of resources or expertise.

The trial activities described in this paper support the premise that these shortcomings can be addressed by means of a high-speed network to integrate communications transport at a variety of speeds and bandwidths and effectively partner university-level experts with teachers and students in the K-12 system. Such a network, both technological and human, can allow for a more efficient delivery of information and curricular material at both the university and K-12 levels.

Beginning in April 1994 and extending through March 1995, US West engaged in a set of technical trials of asynchronous transfer mode (ATM) technology in western Oregon and in Boulder, Colorado. Partnering with several leading universities and a number of other organizations, these trials explored issues surrounding the use of advanced networking technologies in combination with existing network services. Issues that were addressed included network platforms and architectures, along with an understanding of the types of applications that would use such an advanced mix of networking technology. Many planners assumed that high-end niche applications associated with such things as supercomputers would predominate. However, we soon realized that trial participants were placing a strong emphasis on the extension of advanced technologies to secondary schools. Even as that trend began to emerge, some felt that while extension of advanced technologies to secondary schools might be technically feasible, economic and social factors would make such applications unworkable.

Innovative work conducted separately by the University of Oregon and the University of Colorado proved the skeptics wrong. Advanced capabilities were in fact extended to a number of secondary schools in both Oregon and Colorado with encouraging results. This paper discusses the experiments that were performed, the trends observed, and subsequent plans made for follow-on activities. Insights gained are used to project a baseline of technologies that could be deployed over the next 5 to 7 years. We believe that these results should influence the deployment of advanced technology to secondary classrooms and could serve as a model for effective future cooperation between universities and K-12 schools.

BACKGROUND

US West announced a multiphase ATM strategy in October 1993.1 Key elements of this strategy included small, scalable ATM switches flexibly and economically deployed in a distributed architecture, as was done in the western Oregon and Boulder, Colorado, trials. The results were positive, and US West subsequently announced availability of an ATM-based Cell Relay Service offering in its 14-state region in January 1995.2

Experimentation in Oregon was conducted in conjunction with the Oregon Joint Graduate Schools of Engineering as part of "Project NERO" (Network for Education and Research in Oregon).3,4 Five widely dispersed graduate-level engineering schools (Oregon State University, Oregon Graduate Institute, University of Oregon, Portland State University, and Oregon Health Sciences University) and several state office buildings were linked together via a network of ATM switches and associated OC3c and DS3 lines in several major cities in western Oregon. In addition, connectivity was also extended to a teacher training workshop at school district headquarters in Springfield, Oregon. Experimentation in Boulder, Colorado,5 was conducted in conjunction with the University of Colorado6 and included three government laboratories, four industrial laboratories, a public library, four public schools, and even a private residence.

ANALYSIS OF PRIMARY TRENDS

LANs at All Locations

All major trial participants had an extensive imbedded base of legacy LANs. For most large organizations, the LAN environment has expanded rapidly over the past 10 years. For example, less than 10 years ago the campus of the University of Oregon in Eugene had only one Ethernet LAN serving about 50 host computers. Today the campus has more than 80 LANs spread across 70 buildings supporting 6,000 host computers. A complex infrastructure of LANs, bridges, and routers has evolved over the intervening years to support these networks, which have become an integral part of the University of Oregon environment.

The level of sophistication of these campus LANs continued to increase as the trial progressed. Typical campus configurations eventually included a variety of ATM switches providing gateways to remote locations and connectivity between central routers. Direct ATM connectivity was extended to workstations and hosts to allow them to communicate directly over ATM using native IP per RFC 15777 as well as to distributed ATM hubs supporting 10BASET and 100 Mbps Ethernet for devices using LAN emulation.8

Based on this high level of sophistication, one might make the mistaken assumption that LANs are therefore too sophisticated for school districts to deploy. However, we found that the trial school district and library locations had numerous LANs already in place. For example, the School District Headquarters in Springfield, Oregon, has an Ethernet LAN supporting a number of high-end workstations. A Macintosh lab in a school in Albany, Oregon, has a mix of LocalTalk and Ethernet LANs supporting 65 Macintosh devices. The four participating middle schools and high schools in Boulder, Colorado, all have Ethernet LANs supporting a mix of devices. Even the Boulder Public Library has an Ethernet LAN with a sophisticated set of routers and servers supporting local and remote access to online card catalogues and banks of CD-ROMs. While the existence of LANs may not be typical for all K-12 schools, their presence at our trial participant locations is an important indicator of future direction. Local area networks are quickly becoming a plug-and-play item that routinely comes with off-the-shelf workstations and can be installed by the local school administrator, the local library staff, the home user, or (in the case of the Issaquah school district near Seattle) by the students.

IP Networking and Mix-and-Match Telecom

In addition to the universal presence of local area networks, all trial participants accessed the network via routers and TCP/IP. Typical high-end configurations utilized a router to access the public wide area network via either a DS3-based (45 Mbps) or an OC3c-based (155 Mbps) ATM user network interface (UNI). The router often worked in conjunction with local ATM switches as previously discussed.

Existing distributed Internet infrastructures in both Oregon and Colorado are typically based on TCP/IP routers that are interconnected with T1 lines running at 1.5 Mbps. The university trial customers used ATM to construct new Internet infrastructures still based upon TCP/IP routers, but with 155 Mbps rather than 1.5 Mbps interconnectivity. Connectivity at 155 Mbps was then extended to a central campus hub where it was distributed in turn to a few individual desktop computers. By using ATM with routers and TCP/IP, the trial participants were able to gain a 100-fold increase in raw speed and aggregate carrying capacity without changing their basic protocol structure. One might assume that high-end transport options would preclude the participation of secondary schools; however, this did not prove to be the case. The use of IP-based networking provided the ability to cost effectively bring secondary schools into the networked environment.

Routers emerged as the universal anchors to enable the new ATM transport option without affecting existing applications and infrastructures. Routers support most types of network connectivity, providing the ability to transparently mix and match various types of networking options in different parts of the network. With the central dominance of IP-based routers, telecommunications transport became a transparent commodity to be mixed and matched based on cost-performance analyses for each individual location.

Using this extreme level of flexibility, the trial participants proceeded to deploy ATM in their backbone networks and their central locations where high aggregations of traffic were important, while extending trial connectivity to outlying schools and libraries via less expensive T1 or frame relay links. All of the secondary schools were in fact brought into the new environment via lower-cost T1-based telecommunications links in the existing copper-based public switched network. The universal use of IP-based routers to support all types of telecommunications transport allowed the creation of sophisticated "internets" employing the complete spectrum of telecommunications options from low speed to high speed. This is an extremely important trend to consider as we move forward into the next 5 to 7 years.

Multimedia Workstations

All trial participants made extensive use of multimedia workstations. These workstations provided the ability to receive audio/video broadcasts at the desktop, to engage in desktop video teleconferencing, and to access multimedia servers. Initial testing was primarily with high-end SGI or Sun workstations with multicast routing capabilities built into their UNIX kernels. One could easily assume that using high-end workstations would preclude the participation of secondary schools. However, this was not the case. In some situations the expedient option of loaning high-end workstations to the individual schools was adopted so that proof-of-concept work could proceed. As the trial proceeded, other less expensive terminals became available, such as X Window terminals and AV-capable Macintoshes. The use of multimedia workstations from the secondary schools to the university environments was an extremely important trend observed during the trial, along with the increasing availability of affordable multimedia-capable workstations.

Broadcast Video to the Desktop

The broadcast of educational video directly to the desktop was accomplished via M-bone (multicast backbone) capabilities.9,10,11 M-bone originated from an effort to multicast audio and video from Internet Engineering Task Force (IETF) meetings.12 By late 1993 M-bone was in use by several hundred researchers located at approximately 400 member sites worldwide. The M-bone is a virtual multicast13,14,15 network that is layered on top of portions of the physical Internet. The network is composed of islands that can directly support IP multicast (such as Ethernet LANs), linked by virtual point-to-point links called "tunnels." More recently, several vendors have supplied native IP multicast routing protocols. Multicast routing information was distributed using PIM (Protocol Independent Multicast).16,17 Desktop video delivery was initially accomplished via the M-bone tool "nv," with later migration to a newer video tool called "vic" that supported not only nv but also other video encodings and standards such as H.261 and JPEG.

M-bone has been routinely used for some time to broadcast various events to individuals' desks via the Internet. Examples include "NASA Select," the NASA in-house cable channel broadcast during space shuttle missions, and "Radio Free VAT," a community radio station. "Internet Talk Radio" is a variation on this technique to conserve network bandwidth by making prerecorded programs available as data files to be retrieved via ftp and then played back on the local workstation or other appropriate devices as desired. This can have a powerful impact on the inclusion of current, nearly up-to-the-minute information in curriculum development. Programming as varied as talks by Larry King and the "Geek of the Week" are made available via this technique.18

These broadcast services have been significantly expanded recently via the Internet Multicasting Service (IMS), which broadcasts live multicast channels on the Internet such as satellite-based programming like Monitor Radio and World Radio Network, network-based programming such as CBC News and SoundBytes, and live links into the National Press Club, the U.S. Congress, and the Kennedy Center for the Performing Arts.19

The Oregon universities used these capabilities to broadcast computer science colloquia out over the Internet. They found that bandwidth as low as 56 kbps was adequate for broadcast of the image of an instructor, while 500 to 750 kbps was required for proper transmission of text images with hand-pointing motions. They also extended experimentation to the actual teaching of an engineering statistics class to several registered students as an alternative to attendance, using specially equipped video classrooms.

Broadcast of various events via M-bone proved to be very popular at the four secondary schools in Colorado. Every time the space shuttle flew they tapped into the 24-hour-a-day M-bone broadcasts of the live space shuttle activities and monitored the shuttle flight with great interest. A window was brought up on the workstation screen on which the students could watch live pictures of the space shuttle as the mission proceeded. In November 1994 the students monitored the "Prince Charles Video Event" from Los Angeles via M-bone. In addition, movies precompressed via MPEG were broadcast out over the network. In December, an MPEG-compressed portion of a Star Trek movie was successfully broadcast out at 30 frames/s via schools' T1 connectivity and was then decompressed at the receiving workstations. Based on this success, it has been proposed that MPEG-compressed university courses could be delivered to desktop workstations at secondary schools with minimal T1-based access capabilities. Experiences in Colorado showed that only 128 kbps was generally needed for typical M-bone broadcasts, while full-frame 30 frame/s VHS quality video functioned very well via MPEG compression at T1 speeds. One of the more creative uses of M-bone broadcast capabilities was an "end of Boulder trial party" broadcast out over the Internet from a residence directly connected to an ATM switch. A 3-week-old baby at the party was given the distinction of being the "youngest person to ever do an Internet broadcast."

Desktop Video Teleconferencing

Closely related to the delivery of video to the desktop is the use of desktop workstations for video teleconferencing. Desktop video teleconferencing emerged as a strong trend in both the Oregon and the Boulder trials. M-bone proved itself to be extremely useful and versatile for desktop teleconferencing and collaboration, as did InPerson software from SGI and CU-See-Me software for Macintosh computers from Cornell University. Other similar software packages are also becoming more readily available on the market. In general, this type of software supports simultaneous conferencing with as many as four or five other people, along with the ability to share application and white-board windows.

These capabilities were used extensively and successfully in Colorado on a daily basis for four- or five-way conferences. T1-based connectivity was adequate for this type of conferencing, although more efficient compression by the associated software would be helpful. Dr. McBryan used the InPerson software to teach classes at the four Boulder secondary schools. He networked with the four public schools, each of which had an SGI Indy workstation in a classroom. The students then clustered around the Indy workstation while Dr. McBryan taught the class from his home. He taught classes on how to write World Wide Web (WWW) home pages so that the schools could come online with their own servers. All four schools subsequently came online with their own servers based on what they had learned in these online classes. This was an effective demonstration of cooperative work between universities and secondary schools via the network. The capability was so successful that the schools wanted to obtain a large-screen projection capability to make it easier for students to participate and to see the screen. It is significant to note again that this was done entirely with T1-based connectivity on the existing copper-based public infrastructure as well as fiber optics.

In October 1994 a similar experiment was conducted in Oregon at a demonstration of the "21st Century Classroom" for the ITEC Expo's "Computer Office and Automation Show" in the Portland Convention Center. A group of students from Grant High School in Portland participated in a live audio/video exchange with faculty in the Physics Department of the University of Oregon. During this exchange the students were shown how to use a software package called Mosaic and were then able to "sit in" on a physics lecture delivered at the University of Oregon. The students had full two-way audio/video capabilities. They could see the lecturer and the other students and could ask questions during the lecture.

Other Colorado applications included the ability for a student ill at home to interact with teachers and other students. Similar capabilities were used at the college level for off-hours consultation between instructor and students, with video consultation as an enhanced alternative to the current electronic mail procedure. All of these applications have direct value in delivering more timely and customized learning to students in dispersed locations.

Multimedia Servers, Mosaic, and the World Wide Web

The final trend is closely related to the use of desktop workstations for receipt of teleconferencing and video broadcasts. Desktop access to multimedia servers was an extremely important area of emphasis in both trials and was successfully extended to the secondary schools. The National Center for Supercomputing Applications (NCSA) at the University of Illinois developed an environment called "NCSA Mosaic"20,21 that enables wide-area network-based information discovery and retrieval using a blend of distributed hypermedia, hierarchical organization, and search functionality. NCSA has developed Mosaic client software for X Window Systems, Macintosh, and Microsoft Windows. NCSA Mosaic's communications support is provided courtesy of the CERN World Wide Web (WWW) project's common client library.22 No longer isolated within the academic community, the WWW has emerged as one of the Internet's most popular technologies. Commercial companies now advertise the availability of their "home pages," associated software is readily available, and commercial bookstores are stocked with numerous basic how-to books on Mosaic and the WWW.

As used in the Oregon and Boulder trials, the WWW servers were accessed via the Internet and controlled a set of self-paced courses that contained a mixture of video, audio, and text. A student would take a particular course via a multimedia workstation, with the course consisting of multiple windows on the screen, including full-motion film clips with audio, along with associated interactive text. The multimedia courseware is delivered over the WWW and is received and viewed using a WWW reader such as Netscape or Mosaic.

Examples of Technology Uses

The combination of IP networking capabilities, multimedia workstations, video to the desktop, video teleconferencing, and multimedia servers was used in creative and effective ways in both trials. During the summer of 1994 a teacher training workshop was held for Springfield (Oregon) School District science teachers on the use of multimedia X-terminals as network information retrieval machines. As previously mentioned, T1 connectivity was extended to the Springfield School District headquarters as part of the ATM trial in Oregon. Two X-terminals were located there, connected to the university hosts and used by the teachers to practice information access as well as the preparation of their own multimedia courseware.

During the 1994 fall and 1995 winter term at the University of Oregon, three physics classes were done entirely using Mosaic. The students in those classes were able to retrieve the class notes offline and hence were no longer compelled to incessantly take notes during the lecture. As judged by improved performance on exams, this situation seems to have led to better student comprehension of the material. The Mosaic browser allowed the students to personally annotate each lecture page and thus "take notes on the notes." The use of class newsgroups and e-mail allowed the students to work collaboratively on some class assignments as well as provide more useful feedback to the instructor than from the traditional lecture format. Normal lectures were still given, but the lecture material itself was in the form of Mosaic pages projected to the class on an LCD projector. In the winter 1995 term, in a course on alternative energy, students working together in self-organized teams successfully constructed WWW pages of their own in response to a homework assignment to create a virtual energy company and advertise it to the world. This effort and the course can be accessed at http://zebu.uoregon.edu/phys162.html.

In mid-November 1994, Dr. Bothun went to the Cerro-Tololo Interamerican Observatory in Chile to make observations with the telescopes located there. Even though the telescope and classroom were separated by 10,000 miles, successful interactive classroom sessions did occur and the class was able to follow along with the observations as the data were reduced and then placed on the WWW in quasi-real time. These data can be found at http://zebu.uoregon.edu/fp.html. The delivery of scientific data, in this case digital imaging from a telescope, to the K-12 classroom through the Mosaic interface is perhaps the most powerful educational application on today's Internet. As part of the NERO project, the University of Oregon is hoping to get its Pine Mountain Observatory (located at an elevation of 7,000 feet, 26 miles SE of Bend, Oregon) on a T1 link to the Internet for purposes of delivering access to a scientific instrument and data acquisition directly to a networked K-12 classroom. Another example of this kind of activity is provided by the recent space shuttle mission. Dr. Bothun was part of the scientific experiment on board the Endeavor and kept several WWW pages updated about mission progress and the scientific data being obtained. Digital optical photos of the targets that were imaged with the ultraviolet telescope were made available to the Internet audience so that interested followers of the mission could get some idea of the kind of object currently being imaged by the shuttle astronauts. This effort can be seen at http://zebu.uoregon.edu/uit.html.

As previously mentioned, a class on how to create WWW home pages was taught via teleconference to four Boulder, Colorado, secondary schools. Subsequent to receiving this remote training, each school set up its own WWW server on the Internet. The students then proceeded to set up their own WWW home pages, learned how to scan pictures into the computer, and then made their pictures available online via the Internet. The students used the network heavily as part of school projects. As part of an oceanography project, they used WWW "search engines" to research material and make personal contacts with researchers around the country. Individual students made personal contacts with professionals at Scripps Institute in La Jolla, California, and with researchers at Sea World. They also successfully accessed supercomputer centers at the National Center for Atmospheric Research and at NOAA. In addition, a teacher at one of the high schools developed a chemistry class for remote delivery to students at middle schools. In conjunction with this class, some of the high school students mentored middle school students via the network. It is precisely this kind of new, dynamic interaction between experts, information resources, and students at various levels that precipitates the kind of quantum leap the experience of learning can undergo.

Finally, a number of demonstrations were held in conjunction with efforts associated with the Council of Great City Schools, a nonprofit organization representing 50 of the nation's largest urban public school systems. These demonstrations included deployment of a technical model for a high-performance classroom at the Council's Annual Conference of Technology Directors in Denver, Colorado, in May 1994, and the design and deployment of a "21st Century Classroom" for a computer office and automation show at the Portland Convention Center in October 1994.

ANALYSIS AND FORECAST: A LOOK TO THE FUTURE

A number of follow-on projects are either proposed or under way as a result of these successful experimental activities. These follow-on activities, when taken in conjunction with the various experiments, are excellent examples to look to for help in forecasting networking trends for educational institutions over the next 5 to 7 years.

The schools in Boulder, Colorado, used their workstations and associated connectivity in an impressive and successful way. The teachers and students adapted very well to the advanced technology, perhaps even more successfully than anticipated. In 1993 the voters approved a bond issue to allocate funds for provision of a T1 line to every school in the Boulder Valley School District by early 1996. Within a year the schools will be running their own 51-school network, routinely doing the things that were on the cutting edge during the trial period. The ability of schools to actually know and experience the value of these networked services facilitates solicitation of taxpayer support for efforts of this type. When the community sees direct value, then the resources and synergies required to build these networks (both technological and human) emerge more easily.

The schools have a talented pool of teachers and students who will continue with the work begun during the trial. Dr. McBryan has continued to work with the schools and the public library and is currently testing delivery of highly compressed, 30 frame/s, full-frame, VHS-quality video to the schools. The work conducted in Boulder could serve as a model of cooperation between secondary schools and local universities, extending university expertise into the secondary schools and "bootstrapping" them into the information age.

In October 1994 the University of Colorado was awarded a Department of Commerce National Telecommunications and Information Administration grant to develop the Boulder Community Network (BCN). BCN collects information resources from all available sources in Boulder County and makes them available to the whole population through the World Wide Web. Many residents have access through work or school, including schools at the elementary and secondary levels. To widen access, information kiosks have been deployed at strategic locations throughout the county, including libraries and public buildings. The Senior Citizens Center and several similar institutions were provided with both computer platforms and network access. Over 50 community organizations now supply and update information on a regular basis and participate in BCN management and development. Commercial information resources are also available on BCN, including a countywide restaurant menu guide and shopping information. BCN may be accessed at http//bcn.boulder.co.us/.

Dr. McBryan was one of the founding members of BCN, is principal investigator on the NTIA grant, and has been BCN Technical Director from the outset. As a result the BCN information resources were rapidly made available to the T1-connected schools, resulting in interesting curriculum developments. As examples, Spanish classes are translating some BCN home pages into Spanish and a history class plans an online archive of tapings of early Boulder residents, both available to all residents via BCN. The marriage of community network and school Internet connectivity represents a new paradigm for outreach in education.

Looking forward into the future and taking a 5- to 7-year view, the Boulder participants see a number of key trends developing. In today's environment, students must leave their classrooms and go to a computer lab. In the future the computers must be located directly in the classrooms. CD-ROMs will be installed on every desktop, allowing individualized instruction programs to be developed with each student having his or her own programs and subjects. Even today schools can acquire a Performa 630 with a CD-ROM for $1,200 or less. Video broadcast and video on demand will mix freely with CD-ROM capabilities. All three media and their supporting network topologies will bring high-quality educational and current affairs materials into the classroom.

In October 1994 the University of Oregon received one of 100 NTIA grants from the U.S. Department of Commerce. The university in partnership with 15 other Lane County agencies will use the $525,000 grant to create the Lane Education Network. This high-speed computer network will connect the participating agencies' information systems in a seamless, fully accessible community network. The grant will finance demonstration projects, electronic classrooms, and distance-learning, off-campus classes. Agencies in the partnership include three local school districts. In one of those school districts, voters recently passed a $37.5 million bond levy, $3 million of which is designated for instructional technology.

In conjunction with these associated activities in Oregon, a cooperative effort is currently being proposed called "The On-line Universe: Network Delivery of Multimedia Resources in Science to the K-12 Classroom and the General Public." Building on the successful experiments outlined above, the project will focus in several areas. Multimedia educational materials will be developed in the areas of astronomy and the environmental sciences. These materials will be delivered via Internet connectivity at the T1 level between the University of Oregon, the Oregon Museum of Science and Industry (OMSI), and individual science classrooms at Grant High School in Portland and at Springfield and Thurston High Schools in Springfield, Oregon. All materials will be organized and accessible through the Mosaic interface. In addition, M-bone connectivity will be used at these sites for desktop conferencing and live interactions between students, university professors, K-12 teachers, and museum educational staff. The existing multimedia Internet classroom on the University of Oregon campus will be used to develop a new classroom at OMSI to serve as a training area for K-12 science teachers in the Eugene and Portland areas.

Along the same lines but on a larger geographic scale, the Council of Great City Schools is proposing an "Urban Learning Network" to link urban high schools in 14 states to deliver useful information and practical educational services to teachers and students in the real-world environments of urban community high schools. The proposed list of services and applications to be provided should sound familiar at this point: video on demand, interactive multimedia courses, distance instruction or teleconferencing, electronic field trips, and fully equipped smart classrooms.

These various activities show that electronic networks offer the possibility of forming learning communities of educators and learners who can share and develop courseware. The process can be highly interactive, and feedback from the students can be integrated into improving the content. Most importantly, a network of this type facilitates educational outreach activities between higher education, K-12 school systems, libraries, and science museums. These new, network-based partnerships will allow any expert's resources to be delivered to a larger audience with the aim of improving education through the delivery of real data in an interface that promotes learning as a discovery process. Such capabilities will place vast amounts of information at the fingertips of educators, students, and ordinary citizens.

SUMMARY AND RECOMMENDATIONS

This trial and our collective observations support the expectation that traditional barriers between student, teacher, and professional researcher will erode and be replaced by a learning community of individuals who are collectively interested in a particular subject. Herein lies the paradigm shift. The keeper of the knowledge will no longer be the individual K-12 teacher or the professor. Rather, students will now be able to access and interact with a diverse and distributed knowledge base. Professional researchers can make their real data available via interactive networks to this learning community, and science can be taught as a discovery process rather than a collection of facts "known" about the physical world. This approach duplicates what the professional scientist does and can go a long way toward improving science education.

These technological breakthroughs and educational links between universities and K-12 schools underscore the value of education at every level. The electronic superhighway will increasingly allow experts to share their research and foster an excitement for learning among interested students, and in this process they can also help K-12 teachers navigate quickly through the vast and often times confusing information highway. Thus, through this partnership, the "looking at everything" ideal becomes increasingly real with seemingly endless possibilities for exploration, dialogue, and learning.

Through the activities of this trial two unexpected insights were gained relative to the communications technologies: (1) the power and pervasiveness of TCP/IP as a common denominator for educational networking was dramatically underscored; and (2) the scalability, robustness, and aggregation strengths of ATM switching make it an ideal fabric for a public switched network to support educational and developing multimedia applications. Also, the fact that the IP networking of education communities is already in place and expandable, and that applications were deliverable at T1 bandwidths (allowing for the use of existing copper-based infrastructures), means that K-12 and learning communities can begin accelerated access to the information highway today, migrating to other exclusively fiber networks over time. Although DS3 and other higher speed network capabilities are often desirable, it is currently neither feasible nor financially possible to extend these capabilities universally to all schools. Until the value of these networks is comprehended, "affordability" remains a circular argument. The network demonstrated in this trial comes within the reach of anyone when properly scaled, and its incredible value can be experienced by learners and educators.

Society's challenge will be to provide schools and universities with the resources and support they need to have full and equal electronic access to information in this new era. Such a partnership extends also to business and industry, whose success ultimately depends on qualified personnel entering the work force. This trial highlighted the possibilities that will emerge when a learning community discovers the value that can be derived from the effective application of technology and communications networking. Communities of interest exist everywhere in human society. Each community must individually shape the vision and define the value that these networks can deliver. There is no "one size fits all solution." The trial demonstrated that the technologies are very flexible, with many of them already in place to deliver the "value" the users want. Value has a direct relationship to "affordability." Obtaining funding for these technological advancements and communications networks will require creativity from the communities that build them.

Notes

1. Wallace, B. 1993. "US WEST Plots ATM Course with Planned Expansion," Network World, October 18, pp. 28 and 31.

2. Greene, T. 1995. "US WEST Places Its ATM Services Cards on the Table," Network World, February 6, p. 23.

3. Owen, S. 1993. "Network for Engineering and Research in Oregon," proposal submitted to Education Division of National Aeronautics and Space Administration, September 23.

4. Foldvik, R., D. Meyer, and D. Taylor. 1995. "ATM Network Experimentation in the State of Oregon," Proceedings of the IEEE International Conference on Communications, June.

5. Foldvik, R., and O. McBryan. 1995. "Experiences with ATM: The Boulder Perspective," Proceedings of the IEEE International Conference on Communications, June.

6. Research supported in part by NSF Grand Challenges Application Group grant ASC-9217394 and by NASA HPCC Group grant NAG5-2218.

7. Laubach, M. 1994. "IETF Draft RFC: Classical IP and ARP Over ATM," RFC 1577, January.

8. "LAN Emulation Over ATM: Draft Specification," ATM Forum Contribution 94-0035.

9. Macedonia, M., and D. Brutzman. 1993. "MBONE, the Multicast Backbone," Naval Postgraduate School, December 17, taurus.cs.nps.navy.mil: pub/mbmg/mbone.hottopic.ps.

10. Macedonia, M., and D. Brutzman. 1994. "MBONE Provides Audio and Video Across the Internet," Computer 27(4):30-36.

11. Casner, S. 1993. "Frequently Asked Questions (FAQ) on the Multicast Backbone (MBONE)," May 6, available via anonymous ftp from venera.isi.edu:/mbone/faq.txt.

12. Casner, S., and S. Deering. 1992. "First IETF Internet Audiocast," ACM SIGCOMM Computer Communication Review, July, pp. 92-97.

13. Deering, S. 1989. "Host Extensions for IP Multicasting," RFC 1112, August.

14. Moy, J. 1993. "Multicast Extensions to OSPF," IETF draft, July.

15. Deering, S. 1988. "Multicast Routing in Internetworks and Extended LANs," Proceedings of the ACM SIGCOMM 1988.

16. Deering, S., D. Estrin, D. Farinacci, V. Jacobson, C. Liu, and L. Wei. 1994. "Protocol Independent Multicast (PIM): Motivation and Architecture," Internet draft, draft-ietf-idmr-pim-arch-00.txt, November.

17. Deering, S., D. Estrin, D. Farinacci, V. Jacobson, C. Liu, and L. Wei. 1995. "Protocol Independent Multicast (PIM): Protocol Specification," Internet draft, draft-ietf-idmr-pim-spec-01.ps, January 11.

18. Release 1.0. 1994. "The Internet Multicasting Service: Ted Turner, Watch Out!," 94(2):10.

19. Refer to http://town.hall.org/radio/live.html.

20. Andreessen, M. 1993. "NCSA Mosaic Technical Summary," May 8, marca@ncsa.uiuc.edu.

21. Valauskas, E. 1993. "One-Stop Internet Shopping: NCSA Mosaic on the Macintosh," ONLINE, September, pp. 99-101.

22. Powell, J. 1994. "Adventures with the World Wide Webs, Creating a Hypertext Library Information System," DATABASE, February, pp. 59-66.