Super-Size Bandwidth and Two-Way Video in the Classroom

Internet2 and its Impact on Video Conferencing

Connect. Any Way You Want.

Author:Alan GreenbergWainhouse Research

July, 2004

Executive Summary..........................................................................................................................................................2

Overview..........................................................................................................................................................................3Organization and Mission.............................................................................................................................................3Technology....................................................................................................................................................................3Applications.................................................................................................................................................................9

Inter-Organizational Partnerships and Distance Education..............................................................................................5

The Video ‘Stuff’ of Internet2.........................................................................................................................................6Internet2 Commons....................................................................................................................................................6The Megaconference................................................................................................................................................6-7ViDe and the Video Conferencing Cookbook..............................................................................................................7Beacon H.323 Test Software.....................................................................................................................................7-8IPv6.............................................................................................................................................................................8

Applications.....................................................................................................................................................................8Two-Way Video............................................................................................................................................................9Streaming Video...........................................................................................................................................................9Tele-Immersion.............................................................................................................................................................9Remote Instrumentation...............................................................................................................................................9

Conclusion — Super-Size Bandwidth’s Impact on Video in the Classroom...................................................................10About the Author........................................................................................................................................................10About Wainhouse Research.......................................................................................................................................10About Polycom, Inc.....................................................................................................................................................10

Appendix A – Acknowledgments...................................................................................................................................11

FiguresFigure 1 Abilene Network Design................................................................................................................................3Figure 2 Megaconference 2003 Global Map................................................................................................................7

Table of Contents

2

Executive Summary

For a time after it was founded in 1996,Internet2 (Internet2) was perceived amongmany as “that other network,” an exotic luxurythat was limited in value to all but a fewresearch institutions. This is evolving, howev-er, and Internet2 is now rapidly becoming a“mover and shaker” that is changing howresearchers collaborate and how educationalinstitutions reach out to one another.

Internet2 is a consortium of over 206 universi-ties (as of April 2004), as well as over 50 corpo-rate members, 40 affiliate members, and 40international ‘partners,’ many of which comefrom various consortia and governmental agen-cies. Its stated goals are threefold:• To enable a new generation of applications • To create leading edge research and educa-

tional network capabilities, and• To transfer technology and experience to the

“global production Internet”

The result is that Internet2 has become what itwas intended to be: a test environment forpushing the envelope on networking and net-worked computing technologies, while alsoserving as a home for creating new applicationsfor these technologies.

Internet2 backbones typically operate at 2.4Gbps (OC-48) to 10 Gbps (OC-192) capacitiestoday; its GigaPoPs (points of presence) provideregional high-performance aggregation points;and for member institutions, typically local cam-pus networks provide no less than 100 Mbps tothe desktop. At the backbone level, this is min-imally more than 1,000 times the bandwidth ofan E-1 line, and more than 19,000 times thebandwidth of a standard 128KB ISDN line. Thetechnique being used to provide ample band-width is through over-provisioning, so that noone ever runs out of bandwidth.

This creates almost endless possibilities forwhat can be delivered over or achieved usingthis amount of bandwidth. Advanced applica-tions include distributed computation, virtuallaboratories, digital libraries, distributed learn-ing, streaming video, remote instrumentation,and tele-immersion.

The focus of this white paper is on one specificapplication area, digital, two-way video in theclassroom. Video is an ideal application forInternet2: it ‘likes’ bandwidth, and it does notperform well when carried over the lower-speed, ‘under-provisioned’ public Internet.

In addition, Internet2 already has begun to pro-mote the increased use of video conferencing inthe classroom, specifically through initiativesand projects such as the K20 Initiative, Internet2Commons, the Megaconference, the ViDeCookbook, and early, test bed deployment of theITU-T H.350 standard.

Through the Sponsored Educational GroupParticipant (SEGP) program, some 30,000 K-12schools in 34 states in the U.S. alone alreadyare connected to Internet2.

This will only increase in time, even as Internet2also goes global through an increasing numberof inter-networking agreements.

Wainhouse Research believes the massivecapacity provided by Internet2 will have a domi-no effect on distance education, and in particu-lar on the use of two-way, interactive video inthe classroom. As educational organizationsseek to increase their partnerships, improvelearning in the classroom, and enhance tiesbetween their own communities and communi-ties elsewhere, there will be an exponentialincrease in use of two-way video. This trendmirrors the global economy in which we all live, and will help drive a greater ability for pub-lic and private institutions to discover applica-tions together.

Through the SponsoredEducational GroupParticipant (SEGP) pro-gram, some 30,000 K-12schools in 34 states in the U.S. alone already areconnected to Internet2.

3

Overview

For a time after it was founded in 1996,Internet2 (Internet2) was perceived amongmany as “that other network,” an exotic luxurythat was limited in value to all but a fewresearch institutions. This is evolving, howev-er, and Internet2 is now rapidly becoming a“mover and shaker” that is changing howresearchers collaborate and how educationalinstitutions reach out to one another. Contraryto popular opinion, Internet2 is funded primarilyby universities and private enterprise, thoughsome of its initiatives are government-fundedand the organization is working in parallel to(and often in partnership with) the government-funded Next Generation Initiative (NGI).

Organization and MissionIn reality, Internet2 is a consortium of over 206universities (as of April 2004), as well as over 50corporate members, 40 affiliate members, and40 international ‘partners,’ many of which comefrom various consortia and governmental agen-cies1. Its mission is no less lofty sounding thanthat of any other forward facing, futuristicgroup: To develop and deploy advanced net-work applications and technologies, accelerat-ing the creation of tomorrow’s Internet. Its stat-ed goals are threefold:• To enable a new generation of applications • To re-create leading edge research and edu-

cational network capabilities, and• To transfer technology and experience to the

“global production Internet” (in other words,the Internet we all know and love so well).

The result is that Internet2 has become what itwas intended to be: a test environment forpushing the envelope on networking and net-worked computing technologies, while alsoserving as a home for creating new applicationsfor these technologies. In many respects it hasbeen borne out of multiple forces which attimes conflict and at times coalesce into shared,mutually beneficial goals. And emulating thecreation of the original Internet, it is based on aunique partnership of governmental, education-al, and corporate entities.

TechnologyIn practical reality, Internet2 is a collection ofmultiple groups of organizations and institu-tions and networks that meet certain minimumcriteria for advanced network capabilities.Internet2 backbones typically operate at 2.4Gbps (OC-48) to 10 Gbps (OC-192) capacitiestoday; its GigaPoPs (points of presence) provideregional high-performance aggregation points;and for member institutions, typically local cam-pus networks provide no less than 100 Mbps tothe desktop. At the backbone level, this is min-imally more than 1,000 times the bandwidth ofan E-1 line, and more than 19,000 times thebandwidth of a standard 128KB ISDN line. Thepossibilities for what can be delivered over thisamount of bandwidth are endless.

To put the two Internets into perspective, theInternet, though powerful, could be describedas the equivalent of a local dirt path in compar-ison to what is possible with Internet2 – theequivalent of a national highway system(Internet2). One way in which Internet2 offerssuch a rich network is by providing a “super-sized” Meta network. The super backbone inthe U.S. is called the Abilene network: a collec-tion of 101 participants that are connected atthe full 10 Gbps. While those who connect toInternet2 may not all be directly connected toAbilene, they benefit from the bandwidthenabled by the network.

Internet2 is a predecessor to an even moreaudacious effort, National LambdaRail (NLR).Like Internet2, NLR is not a single network, buta rich set of facilities, capabilities, and servicesthat will support a set of multiple experimentaland production networks for the U.S. researchcommunity. At this point in the early designphase, NLR has at its foundation a dense wavedivision multiplexing (DWDM)-based nationaloptical footprint, being deployed on roughly10,000 route-miles of fiber. Using CiscoSystems’ 15808 optical electronic system with acapacity of 40 channels (wavelengths) per fiberpair, each wavelength will be able to supporttransmission of 10 Gbps. With four wave-lengths expected to be installed and availableon the first day of operation, that will providequadruple the network bandwidth of theAbilene network. Some aspects of NLR arecomplementary to Internet2 and it is worth men-tioning primarily because it will function toInternet2 much as Internet2 functions to theproduction Internet: a harbinger of things tocome. Details are at www.nationallamb-darail.org.

Returning to the topic of Internet2, thoughsuper-sized bandwidth is important, it is farmore valuable to think of Internet2 not only forits bandwidth, but also for the applications it isdriving and the inter-organizational partnershipsand change for which it is an agent.

Figure 1 Abilene Network Design

1 The actual number of institutions which are benefiting from Internet2 through affiliate memberships, as discussed later in this white paper, is even larger.

4

Overview (cont.)

ApplicationsWhat all this bandwidth makes possible aresome very advanced applications that makeStar Trek seem like yesterday’s news. Theseinclude separately or in combination:• Distributed computation• Virtual laboratories• Digital libraries (super-charged and made

more broadly available)• Distributed learning• Streaming video• Remote instrumentation• Tele-immersion

Some of these applications are described ingreater detail later in this document. Suffice tosay, each of these could be subject of a whitepaper unto itself; in this case our focus is ondigital, two-way video in the classroom.

Video is an ideal application for Internet2: it‘likes’ bandwidth, and it does not perform wellwhen carried over the lower-speed, non-Quality-of-Service enabled public Internet. Inaddition, Internet2 already has begun to pro-mote the increased use of video conferencing inthe classroom, specifically through initiativesand projects such as the K20 Initiative, Internet2Commons, the Megaconference, the ViDe cook-book, and early, test bed deployment of the ITU-T H.350 standard.

Internet2 already hasbegun to promote theincreased use of videoconferencing in the class-room, specifically throughinitiatives and projectssuch as the K20 Initiative,Internet2 Commons, theMegaconference, theViDe Cookbook, and early,test bed deployment ofthe ITU-T H.350 standard.

5

Inter-Organizational Partnerships and Distance Education

Wainhouse Research believes the massivecapacity provided by Internet2 will have a domi-no effect on distance education, and in particu-lar on the use of two-way, interactive video inthe classroom. The process by which this willtranspire is multi-faceted and complex, but putsimply, as universities continue to createGigaPoPs, they will be looking for regionalgroups interested in connecting to thoseGigaPoPs.

In effect, the universities behind the GigaPoPswill act as Internet2 ISPs for smaller universitiesand colleges, as well as some of the larger pri-mary and secondary school networks, such asvarious statewide networks or large metropoli-tan school networks.

The interest on the part of many organizationswill be to connect to universities and haveaccess to the wealth of information, applica-tions, and technologies available only throughInternet2; on the part of others, the interest willbe on two-way video and links with the broaderworld – as discussed later in this white paper.

Not so widely known is that Internet2 is alreadyreaching many organizations not specificallydescribed as one of its primary or affiliatedmembers. This is partly the result of the K20Initiative, meant to enlarge the reach ofInternet2 through promotion to K-12, smallercolleges and universities, and not-for-profitorganizations such as libraries and museums. Ineffect, the K20 Initiative enables smaller insti-tutions to join as connected, secondary mem-

bers at a fraction of the cost of full membership.According to Louis Fox, Director of Internet2 K20Initiative, some 30,000 K-12 schools in 34 statesin the U.S. alone already are connected toInternet2 via what is known as the SponsoredEducational Group Participant (SEGP) program.

The SEGP program supports the aggregation ofeducational organizations brought together in astatewide network or other wide-area networkwithin a state. Examples of these aggregationsare statewide K-12 networks, community col-lege networks, and similar collectives of educa-tional organizations. These networks areengaged in applications development and otherprojects designed to enhance their use ofadvanced networking infrastructure and ser-vices. Details about Internet2 K20 initiative andspecifics about its participants are available at:http://k20.internet2.edu/segp/currentpartici-pants.html

A state or regional education network interest-ed in connecting to Abilene must find anInternet2 primary member to sponsor the con-nection. The entire state network will then beconnected to the Abilene backbone via one ofthe 20 or so network GigaPoPs. Once connected,this state network will be known as an Internet2SEGP. All network traffic destined for anotherSEGP or Internet2 member is routed overAbilene at little or no cost to the SEGP. Thesponsoring Internet2 member generally coversthe cost of the SEGP traffic.

Every state in the U.S. has one or more statewide networks that are a combina-tion of higher education, governmental agen-cies, and primary and secondary school systems“riding” that network for video (and often data)services. We can only expect that the actualnumber of SEGP’s will continue to increase, pro-viding educational opportunities domesticallyand internationally.

This is because the next wave of Internet2growth will come from increasing connectionsfor collaboration around the world. Manynational (or regional) research network effortssimilar to Internet2 exist around the world, andInternet2 has signed memoranda of understand-ing (MOUs) with over 40 such organizations

with the objective of working together on bothconnectivity and applications. Internet2 doesnot specifically provide connectivity directly tothose countries, but many of them do in factconnect to the United States (often with supportfrom the U.S. National Science Foundation).Details on Internet2’s international activitiescan be found at www.internet2.edu/internation-al/ . For information about which networks con-nect to the United States, visit the StarTap Website at www.startap.net/.

Because Internet2 does not provide direct con-nectivity to international networks, it workshard (as if to compensate for this lack) by sup-porting peering between international networksand the Abilene network. Peering with Abileneis made possible via international connectionpoints (also referred to as exchange points)which facilitate peering among US and interna-tional research and education networks.Peering can also be done via international linkssuch as TransPAC or via other networks con-nected to Abilene, e.g., CALREN-2 or GEANT.The bottom line for organizations outside theU.S. is that it can cost surprisingly little to par-ticipate in a domestic or regional high-speednetwork, but the trick is to ensure that such anetwork is already involved in peering withInternet2. A visit to http://international.inter-net2.edu/index.cfm describes the essentials ofhow to harness the current international activi-ty in this space.

Wainhouse Researchbelieves the massivecapacity provided byInternet2 will have adomino effect on distanceeducation, and in particu-lar on the use of two-way,interactive video in theclassroom.

6

The Video ‘Stuff’ of Internet2

Internet2 CommonsInternet2 Commons is a concept that arose outof one of Internet2’s many working groups, inthis case the digital video working group. As anIP-centric network, the focus of Internet2 fromthe start (in terms of two-way video) has beenon the ITU-T H.323 standard. As a result, theCommons has become a focal point for IP-basedvideo conferencing. The goal is to foster two-way video between anyone in the broaderInternet2 membership, such as SEGP’s, as wellas sub-groups within the larger member institu-tions. Thus Internet2 Commons provides multi-way video conferencing services with capabili-ties such as simultaneous streaming, archiving,high-bandwidth support, continuous presencevideo layout options, and encryption. Hostedphysically (as are many video-related projectsand programs) by the Ohio SupercomputingCenter and Ohio State University, the Commonsis less an advanced network than a set of ser-vices meant to make two-way video (and otherapplications, such as Web conferencing,streaming, and archiving) available to groupsthat otherwise would not have access to thesecapabilities. It is available in two flavors:• The Prime Subscription, available for a $2,000

annual fee that makes available 300 MCUport hours, with additional time available ataffordable increments.

• The Alliance Subscription, available for a$1,000 annual fee that provides the CommonsMCU’s as a backup for a group’s own bridgingservice. It includes 100 MCU port hours andcalls for a reciprocal MCU agreement.

The Commons has led to some unique educa-tional initiatives that might not otherwise havetaken place, including a distributed nursinginformatics course that regularly brings togeth-er 11 sites with 6 instructors; the Network forEarthquake Engineering Simulation (NEES) thatconnects 15 to 20 sites some two to four timesa month; and the teaching of lesser taught lan-guages. When only one instructor of the NativeAmerican language Ojibway exists, as an exam-ple, Internet2 can foster cooperation betweenuniversities and lead to linguistics studentskeeping alive the language.

The Commons also serves as a way to delivertwo-way video capabilities in alternate ways.As an example, it has implemented desktop col-laboration technologies such as Polycom’sWebOffice™, the Virtual Rooms video confer-encing Schemes (VRVS — a web- and reflector-based, software MCU that arose for video con-ferencing and collaboration among the HighEnergy and Nuclear Physics communities),Wave Three Session Software, IMFirst, andother tools.

The MegaconferenceThe Megaconference events are the world’slargest H.323 video conferences, bringingtogether participants from locations literally all over the world. Held at various times afterFall Internet2 Member Meetings,Megaconference participants discuss their pro-grams from a content perspective and with aneye toward ‘Best Practices.’ These are nottechnical presentations; only end users whoactually use the technology in their own workare allowed to present.

The brainchild of Dr. Bob Dixon of Ohio StateUniversity and OARNet, the Megaconferencerepresents the ‘democratization’ of the videoconferencing community (somewhat in emula-tion of the traditional Internet — with an orga-nized disorganization that lends itself to spon-taneity, interactivity, and subsequent partner-ing). Each Megaconference — of which therehad been five as of fall 2003 — has consistedof a series of presentations, chats, and groupsessions meant to instruct, inform, and fosterpost-event contacts and partnerships. The mostrecent Megaconference included musical mini-concerts. Certain technical guidelines have

been established, such as the use of H.263video and 768 kbps connections, but these havenot limited the 200 or so participants who nowtake part in each event.

The following drawing demonstrates the complexity, sheer size, and global nature of theMegaconference. Listing the bridge locationsdoes not tell the entire story, as some of theidentified number of participants on each bridgemay have been connecting other bridges as well.

The Megaconferenceevents are the world’slargest H.323 video conferences, bringingtogether participants fromlocations literally all overthe world.

7

Events like the Megaconference have put pres-sure on some of the traditional ‘closed off’statewide networks that traditionally were‘islands’ unto themselves (as a result of legacytechnologies that lacked gateways or program-matic and funding decisions that limited open-ness to external networks). As Internet Protocolcontinues to grow, as an example, more andmore legacy networks (such as fiber-optic andATM networks) will find ways of connecting tothe Megaconference and to Internet2. Eventslike the Megaconference Junior (designed forprimary and secondary students around theworld -- http://megaconferencejr.cciu.org) andthe Keystone K-12 Conference (www.k12video-conf.org) will lead to increased exposure forInternet2 and video in the classroom.Megaconference details can be found athttp://www.megaconference.org.

ViDe and the video conferencing Cookbook ViDe (the Video Development Initiative) is a col-laboration of 13 universities and other membersof the international advanced networking com-munity. Though not specifically an Internet2program, its participants are members ofInternet2 and have close ties with the organiza-tion. The ViDe participants are working togeth-er to advance the state of networked digitalvideo and voice-over-IP (VoIP), sponsoring work-ing groups that are exploring such areas asvideo-on-demand, Quality of Service, SessionInitiation Protocol (SIP), and the MPEG-4 videostandard. Among other projects, ViDe providedan early test bed for H.350 directory services(see below), enabling vendors to see how theirearly implementations of H.350 deployed andenabled users to trial initial versions of thisimportant directory standard. The ViDe.Netenvironment allows users to communicate withother ViDe.Net sites around the globe using any

combination of IP telephone, wireless IP phone,desktop video conferencing, or classroom tech-nologies. Information about ViDe can be foundat http://www.vide.net.

ViDe also provides a popular video conferencingresource known as the “Cookbook.” Now on itsVersion 4.0, the Cookbook is an online tool forunderstanding the dynamics of H.323 video con-ferencing, containing tips, networking funda-mentals, best practices, and other valuableinformation. It can be found athttp://www.vide.net/cookbook.

H.350 Directory ServicesThe H.350 story is an excellent example of thesynergies that have developed betweenInternet2 members and video as an application(as represented by the Vide.Net). Ratified bythe ITU in September 2003, H.350 resulted froman Internet2 video middleware working group

Figure 2 Megaconference 2003 Global Map

8

The Video ‘Stuff’ of Internet2 (cont.)

(some of whose members were also membersof ViDe). The H.350 standard provides a uniformway to store and locate information related tovideo and VoIP in directories that are linkedseamlessly to enterprise directories. H.350uses LDAP (Lightweight Directory AccessProtocol) to store users’ voice, video and col-laborative multimedia information in a way thatintegrates with directory and identity manage-ment systems already in place at large enter-prises. Many of these enterprise systems alsouse LDAP; the new H.350 standard will enableenterprises to scale up video and VoIP opera-tions to very large-scale deployments. Accountconfiguration details, authentication and autho-rization are linked to the enterprise directoryusing LDAP. H.350 improves security by provid-ing standardized management and storage ofauthentication credentials.

The practical impact of H.350 cannot be over-stated when it comes to broadening the abilityto find other organizations interested in videoconferencing. H.350 enables networks to regis-ter their video users with other networks – ineffect creating super-directories of users.While we’re not likely to see one individualdirectory services provider anytime soon, weeventually will see numerous super-directoriessprouting up which will enable communities ofinterest to find one another more easily.Vendors are already embedding H.350 supportin their management and bridging softwareproducts; the IETF is likely to approve it as astandard in addition to the ITU; and it will playa role in further deployments of VoIP as well asvideo. All of this will go a long way to addressthe problem that was faced by adopters of videoconferencing in the past: whom are we going tocall?

Beacon H.323 Test SoftwareOther tools also have arisen out of Internet2’spush into video. An example of one is the H.323Beacon, a software tool that is the result of anInternet2 End-to-End Performance Project. TheH.323 Beacon is a client that loads onto serverslocated on Internet2, as well as on PC’s. Itallows users to create a kind of “fake” videoconference, in effect measuring, monitoring,and troubleshooting possible performance of acall even before it is established. Our analogy

at Wainhouse Research is that it is somewhatakin to a ‘ping’ or ‘TraceRoute’ command, only itprovides significantly more diagnostic informa-tion and features a superior user interface. Endusers, network engineers, and conference oper-ators can debug H.323 application performanceproblems from the network to the host (fromend-to-end) using the Beacon software.

Developed by Internet2 and the Ohio Internet2Technology Evaluation Center of the OhioSupercomputer Center, the H.323 Beacon wasintroduced to the advanced networking commu-nity in fall of 2003. After a rigorous test periodin conjunction with the Megaconference, theBeacon was selected for use by Internet2Commons and was recommended for perfor-mance troubleshooting by various vendors.Additional information is available at www.ite-cohio.org/beacon.

The H.323 Beacon is one example of the End-to-End Performance Initiative (E2EPI). This is aspecific Internet2 initiative designed to create apredictable and well-supported environment,making it possible for Internet2 campus networkusers to have routinely successful experiencesin their development and use of advancedInternet applications. The E2EPI is focused onimproving Quality of Service (QoS), understand-ing that ‘super-size’ bandwidth alone does notsolve all problems. As of this writing, E2EPI isdeveloping a distributed, scalable system tomonitor, test, and report end-to-end perfor-mance. The system, called the E2EPIPerformance Evaluation System (piPEs), will beable to indicate performance capabilities andlocate problems along the path between twocomputers connected by a backbone network,participating campuses, regional networks, andGigaPoPs. When piPEs is completely deployed,it should significantly improve the opportunityfor advanced Internet applications to operate atpeak performance. More information aboutE2EPI can be found athttp://e2epi.internet2.edu.

IPv6As version 4 of the Internet Protocol (IPv4) ages,vendors, research consortia, and advancedusers are preparing for the onset of the nextversion of the protocol: IPv6. With IPv4addresses reaching their quantitative limits,

and with the proliferation of wired and wirelessdevices, IPv6 will offer 128-bit Internet address-es (versus the 32-bit IPv4 addresses), which willdramatically increase the number of available IPaddresses. There are many other enhancementsto IPv6, including improved security and moreefficient packet handling. Why is this importantto video conferencing? For several reasons:first, because improved data transmission effi-ciency will ultimately result in improved videotransmission. Second, IPv6 will set the stagefor many new applications, including areas likewireless video conferencing. Many universi-ties, businesses, and government research labshave taken leading roles in identifying and solv-ing the challenges to improving IP. Internet2 isvery active in this area. More details are avail-able at http://ipv6.internet2.edu.

ApplicationsThus far this document has described some of the initiatives and programs of Internet2, and their potential impact on two-way video conferencing.

The fact is, two-way video is only one of theareas in which Internet2 is ‘pushing the enve-lope’ on digital video. Internet2’s video applica-tions encompass everything from IP video con-ferencing to on-demand content to remote con-trol of microscopes and other instrumentation.Internet2 Digital Video Initiative (http://dv.inter-net2.edu) is one effort involved in developing awide range of advanced digital video capabili-ties for the national research community. Overtime, many of these applications will filter theirway into other communities of interest whichparticipate in or benefit peripherally fromInternet2.

Internet2’s video applicationsencompass everything fromInternet-based video confer-encing to on-demand con-tent to remote control ofmicroscopes and otherinstrumentation.

9

Applications

Two-Way VideoAlready many K-12 practitioners throughout theworld are beginning to seek to network moreand more with one another or with institutionsof higher education and research – whetherthrough standard ISDN or Internet connections,or by taking advantage of access to Internet2.The examples are so numerous that those men-tioned here could be considered the tip of the iceberg.

California schools, as an example, connect overthe state’s non-profit Corporation for EducationNetwork Initiatives in California (CENIC)’sinfrastructure and through initiatives such asthe Digital California Project, a statewide initia-tive providing K-12 schools with access to thehigh-speed advanced services network, CalREN.This opens up both high-speed intra-state activ-ities, as well as access through CENIC’sInternet2 connections to the any other institu-tion on the globe. The Imperial County Office ofEducation, as an example, connects usingPolycom ViaVideo®, ViewStation®, VSX™ 7000,and MGC™-100 units for regular meetings, stu-dent activities, and even for international links,such as frequent connections to China. TheState of Oklahoma’s OneNet reaches 80% ofthe state’s K-12 schools; these schools usePolycom ViaVideo, ViewStation, VS 4000™, VSX 7000, and ClassStation™ packages to holdfrequent links with others throughout the U.S.,in Latin America, and Europe. While video con-ferencing traditionally was found in rural loca-tions, it increasingly is making its way intourban districts that are seeking to leverageinvestments in network and technology, as wellas their own educational resources. Many largemetropolitan districts are merging educational

interests with other community and businessinterests to “go international” in search of part-nerships for content, education, and business.

Streaming videoResearchChannel is pioneering new methods ofInternet-based distribution of HDTV and better-than-broadcast-quality video through ongoingcollaborative technology experiments. This is anonprofit consortium of leading research insti-tutions focused on building high-qualityInternet, cable, and satellite-based channels tofacilitate the communication of research infor-mation. http://www.researchchannel.com

Meanwhile, a Public Television: NextGeneration Interconnection Pilot, run by theUniversity of Wisconsin and Washington StateUniversity, is working on a thorny public televi-sion (PTV) challenge. Currently stations are con-nected by a one-way satellite system. The PBSstations at University of Wisconsin andWashington State University, along with a con-sortium of other university PTV stations, havebeen using Internet2 networks to begin devel-oping new applications for the next-generationinterconnection system. Using broadband IPvideo connections from their host universities,the PTV-Internet2 project members are testingstation-to-station, broadcast-quality (MPEG-2)video streaming, server-based broadcast video-on-demand, video segment search and fulfill-ment at MPEG-2 levels, and collaborative program editing. Information on this initiativecan be found athttp://ra.doit.wisc.edu:8888/vid.html.

Tele-ImmersionThe Tele-Immersion Project is exploring how toconduct realistic, life-sized, 3D tele-immersionto create what could be described as the most‘complete’ interactive experience imaginable. Itis a project of Advanced Network & Services,Brown University, the University of NorthCarolina, and the University of Pennsylvania.Project collaborators envision an Office of theFuture where computer vision, networking, andcomputer graphics techniques are used to cap-ture a dynamic 3D model of one office —including walls, furniture, objects, and people— and then re-create a virtual representation ina similarly equipped remote office. This con-

cept is not new, with companies like Telesuiteand others having attempted similar, thoughlower bandwidth, efforts at recreating the expe-rience of being together. Internet2 networksoffer a high performance test bed for thisresearch, which demands high bandwidth andlow latency to support real-time interactionbetween remote users and shared virtualobjects. The work being done on projects likeTele-Immersion will likely have an impact onfuture applications for video conferencing inmany other environments as well.

Remote InstrumentationWhile many have been expecting telemedicineto lead the way in remote operation of equip-ment, other interesting applications are underdevelopment using Internet2 as the test bed.North Carolina State University, for example, isrunning a project exploring real-time tele-oper-ation of the Tele-vator. This is a computerizedexcavation backhoe that can be remotely oper-ated over high-performance networks. Becauseof its size and potential criticality of operation(e.g., in hazardous rescue situations), the Tele-vator requires a high-level of sophisticated two-way feedback, including adequate depth ofvision provided via high-definition stereovision.Guaranteed QoS – such as network bandwidth,latency (delay) control, and jitter control – areessential to ensure the quality of the 3D image,audio, and equipment control channels. Visithttp://CARL.ce.ncsu.edu for more information.Other video-related projects are underway byvarious members of Internet2, exploring areaslike collaboration technologies, distributedlearning, and digital libraries.

The examples (of practition-ers throughout the worldseeking to network) are sonumerous that those men-tioned here could be consid-ered the tip of the iceberg.

10

Conclusion — Super-Size Bandwidth’s Impact on Video in the Classroom

About the AuthorAlan D. Greenberg is a Senior Analyst &Consultant at Wainhouse Research. As consul-tant, analyst, communicator, and strategist,Alan has worked in the telecommunications,video conferencing, software and services, andmultimedia arenas for 20 years, holding posi-tions with VTEL, Texas Instruments, and severalAustin, Texas-based startups, and consulting tomany organizations. At VTEL he conductedresearch into dozens of distance learning net-works, was product marketing manager for a setof turnkey classroom packages, and led a num-ber of educational and training initiatives andresearch into healthcare networks. Mostrecently he was primary author on the segmentreport, Video Communications ManagementSystems 2004, was lead analyst on theWainhouse Research WebMetrics research pro-gram, and authored the white papersNavigating the Sea of Research on Video con-ferencing-Based Distance Education and BestPractices in Live Content Acquisition forDistance Learning. He also has authoredreports on conferencing endpoints & bridges,streaming video, and voice/fax services. Hespecializes in primary end user research and isa trained focus group moderator and interview-er. Alan holds an M.A. from the University ofTexas at Austin and a B.A. from HampshireCollege, Amherst MA. He can be reached atagreenberg@wainhouse.com.

About Wainhouse ResearchWainhouse Research (www.wainhouse.com) isan independent market research firm that focus-es on critical issues in rich media communica-tions, video conferencing, teleconferencing, andstreaming media. The company conducts multi-client and custom research studies, consultswith end users on key implementation issues,publishes white papers and market statistics,and delivers public and private seminars as wellas speaker presentations at industry groupmeetings. Wainhouse Research publishesConferencing Markets & Strategies, a three-vol-ume study that details the current market trendsand major vendor strategies in the multimedianetworking markets, a variety of segmentreports, and the free newsletter, TheWainhouse Research Bulletin. Its subscriptioncontent service can be found at www.wrplat-inum.com.

About Polycom, Inc.Polycom develops, manufactures and markets afull range of classroom, corporate training, andmedical education solutions through its high-quality, affordable voice and video communica-tion endpoints, video management software,web conferencing software, multi-networkgateways, and multipoint conferencing and net-work access solutions. For additional informa-tion, visit www.polycom.com.

Conclusions — Swim in the Research without DrowningIt should be clear by now that the sum of super-size bandwidth, a term which this analyst used somewhat indiscriminately to connote the idea of whatInternet2 is all about, is going to be neither greater nor lesser than its parts. In other words, bandwidth itself is one great sum; but the parts themselves,the projects and programs underway to promote video conferencing technologies, are also great sums that add up to something larger than what seem atfirst glance.

Every time a Megaconference occurs, it has a viral impact on the sensibilities and outreach behavior of current video conferencing users. Every time astatewide network that is an SEGP enables some K-12 school to find and connect to another such organization across the country or across the oceans, itshows them that they can be connected to anyone on the globe in ways that far exceed text- or graphics-based, non-immediate mechanisms such as websites and e-mail. Projects like the Global Nomads links between students in the U.S. and Iraq, or Dr. Robert Ballard’s delivery of video directly from theTitanic over the public Internet and Internet2 in MPEG-2, are singular and unique today. These are likely to increase exponentially over time as greaterbandwidth and a greater ability to discover applications together lead private and public organizations to interact more and more using video.

While eRate grants in the United States may be funding equipment purchases, it is Internet2 that is fostering a network sensibility that can be said to be21st century. That sensibility will be like a steady drip that turns into a stream and then the equivalent of a fire hydrant. The flow will be less about thebandwidth, and more about what is riding that bandwidth: two-way interactive video and a host of other video-related technologies.

11

Appendix A – Acknowledgments

Super-Size Bandwidth and Two-Way Video inthe Classroom is based on this analyst’s opin-ions and the result of interviews with a handfulof individuals involved with Internet2.Wainhouse Research wishes to thank the fol-lowing individuals and their organizations, with-out whose support this document could nothave been written. We also refer readers to theInternet2 web site (www.internet2.org) for addi-tional information.

Bob Dixon (Robert S. Dixon, Ph D, PE)Chief Research EngineerOhio State University, Office of the CIOand Ohio Academic Resources Network(OARNet)Bob_Dixon@osu.edu

Jill GemmillAssistant Director, IT Academic ComputingInternet2 Applications Lead University of Alabama at BirminghamJGemmill@uab.edu

Jonathan TymanManager, Digital Video InitiativesInternet2tyman@internet2.edu

Greg WoodDirector of CommunicationsInternet2ghwood@internet2.edu

Rev. 7/04

www.ivci.com

Polycom Headquarters: 4750 Willow Road, Pleasanton, CA 94588 (T) 1.800.POLYCOM (765.9266) for North America only. For North America, Latin America and Caribbean (T) +1.925.924.6000, (F) +1.925.924.6100

Polycom EMEA: 270 Bath Road, Slough, Berkshire SL1 4DX, (T) +44 (0)1753 723000, (F) +44 (0)1753 723010

IVCi: 180 Adams Avenue | Hauppauge, NY 11788 | 1-800-224-7083 | www.ivci.com

©2004 Polycom, Inc. All rights reserved.

Polycom, the Polycom logo design, ViaVideo and ViewStation are registered trademarks and Polycom WebOffice, VS4000, VSX, MGC and ClassStation are trademarks of Polycom, Inc. All other trademarks are the property of their respective owners.