Information Technology: 21st Century Revolution
Large Scale Networking
left right
Overview
LSN networking infrastructure support
Fundamental networking research
STAR TAP
Wireless standards
Active networks
Tolerant networking
Quorum: End-to-end mission success
Sensor Information Technology (SensIT)
Very high speed networking
DOE R&D
Global Observation Information Network (GOIN)
Load balancing
LSN applications R&D
NGI: Next Generation Internet
Thrust 1 accomplishments and plans
NSF
DARPA's SuperNet
NIST
Thrust 2 accomplishments and plans
Performance measurement and improvement
NSF applications
DARPA applications
NIH applications
DOE applications
NASA applications
NIST applications
New Starts: SII program FY 2000-2001


Overview


In education, in health care, in science, and in business, astonishing advances in networking technologies are driving fundamental changes in society, dramatically influencing the way Americans pursue their everyday activities. From instantaneous stock market quotations to online art auctions, from Web-based U.S. Postal Zip Code lookups to nationwide law enforcement databases, it is easy to see the significant impact of Federal agency support for Large Scale Networking (LSN) on today's evolving environment. LSN has become a pivotal force, generating critical advances in technologies that have been quickly adopted by the governmental, academic, and commercial sectors.

Federal LSN R&D includes:

  • Traditional networking research to support agency mission requirements
  • The Next Generation Internet (NGI) initiative
  • Research in Scalable Information Infrastructure (SII)

LSN programs explore long range fundamental networking research issues and transition developing LSN products into tools to support agency missions. Continuing the Federally supported R&D responsible for the core technologies that made the Internet and Internet applications possible, LSN focuses on technologies needed by the Federal agencies, infrastructure to support agency networking, and networking applications development.

Since its inception in 1998, the NGI initiative has been a primary focus of LSN, building on the LSN base programs to provide the networking research, testbeds, and applications needed to assure the scalability, reliability, and services required by the Internet over the next decade. In FY 2000, with the construction of its testbeds largely completed, NGI's original three goals were refocused as two thrusts that emphasize improved network performance and functionality and revolutionary networking applications.

In its 1999 report, "Information Technology Research: Investing in Our Future," the PITAC warned that Federal support for IT R&D was seriously inadequate. Research programs intended to maintain the flow of new IT ideas and train the next generation of researchers were funding only a small fraction of the necessary research. The PITAC recommended a significant new program and increased funding for IT R&D and an expanded Federal role in networking R&D that includes interoperability and usability.

Federal agencies responded to this challenge with a proposed new program in IT R&D. A major component is SII, whose research goal is to develop tools and techniques enabling the Internet to grow (scale) while transparently supporting user demands. An integral part of LSN, SII R&D complements the base LSN and the NGI efforts. SII research will focus on deeply networked systems, anytime, anywhere connectivity, and network modeling and simulation.

The following section of this report describes the FY 2000 accomplishments and FY 2001 plans for base LSN, NGI, and SII R&D.

LSN teams

The LSN Coordinating Group (LSNCG)-formerly the LSN Working Group (LSNWG)-coordinates multi-agency Federal networking R&D programs. Four teams report to the LSNCG to assist in this task and to help implement advanced networking technologies:

The Joint Engineering Team (JET) coordinates the network architecture, connectivity, exchange points, and cooperation among Federal agency networks (FedNets) and other high performance research networks, and provides close coordination of connectivity, interoperability, and services among Government, academia, and industry to improve end-to-end user performance and avoid duplication of resources and efforts. The JET also coordinates international connectivity and interoperability.

The JET cooperates with the academic community's gigabits per second points of presence (Gigapops), the Abilene network (a consortium among Qwest, Cisco, Nortel, and Indiana University), and the University Corporation for Advanced Internet Development's (UCAID's) Internet2 (I2). During FY 2000, the JET is developing and coordinating implementation of the Next Generation Internet Exchange Points (NGIXs), where connectivity and peering of the FedNets take place (page 89).

To date, NSF has made High Performance Connection (HPC) awards for NGI connectivity to 167 universities and four high performance computing centers, with other NGI sites supported by DARPA's SuperNet (15), NASA (5), DOE (5), and NLM (1). The JET continues to coordinate the connection of additional sites to the NGI testbeds, helping to implement multicast on the NGI testbed backbone and provide metrics to improve performance.

The JET supports cooperation among the agencies and with Abilene to provide improved, lower-cost services to geographically challenging areas such as Hawaii and Alaska. In FY 2000, the JET is coordinating an increase in research connectivity to Hawaii and Alaska.

The JET coordinates connectivity requirements for applications demonstrations, using NGI testbeds at supercomputing conferences and other demonstrations. The JET and other LSNCG Teams contributed to LSN's "Bridging the Gap Workshop" held August 11-12, 1999, at NASA's Ames Research center at Moffett Field, California. This workshop brought together networking researchers and NGI applications developers to promote increased awareness and cooperation, providing roadmaps and schedules for developing and implementing emerging technologies.

The Networking Research Team (NRT) coordinates agency networking research programs, shares networking research information among Federal agencies, and supports NGI Thrust 1 activities. It provides

 

outreach to end users by disseminating networking research information and coordinating activities among applications developers and end users.

The NRT is active in developing agency workshops on middleware. In cooperation with the other LSN Teams, the NRT organized the "Bridging the Gap Workshop," which identified three key areas of networking research-quality of service, multicast, and Internet security-as the highest priorities for achievable technology advances within the next one to three years. In December 1999, NRT coordinated an NGI research/Principal Investigator (PI) meeting with researchers from DARPA, NSF, and NIST to disseminate research results and increase collaboration.

The High Performance Networking Applications Team (HPNAT) coordinates Federal R&D to maintain and extend U.S. technological leadership in high performance networking applications, encouraging research that employs advanced networking technologies, services, and performance to support leading-edge applications. Advances will lead to new and more capable network applications that support Federal agency missions and build the foundation for the evolving national information infrastructure.

The HPNAT promotes cooperation in large-scale networking applications development among Federal agencies, Government laboratories, academia, and industry, and organizes information dissemination activities including technology demonstrations, workshops, and seminars. The HPNAT supports NGI Thrust 2 by helping organize NGI demonstrations at conferences such as SC99 in Portland, Oregon (page 99). The HPNAT coordinated the identification and presentation of 15 application case studies (see http://www.nren.nasa.gov/case.html) at the "Bridging the Gap Workshop." They were selected to represent nationally significant needs and broad application areas for the technologies examined in the workshop.

The Internet Security Team (IST) facilitates testing of and experimentation with emerging advanced network security technologies. It provides the LSNCG with feedback and direction for NGI research in network security by serving as a forum for the exchange of security requirements and needs and current and emerging security technologies. The IST encourages development and use of Internet security testbeds, helping LSN agencies and the JET to implement these testbeds and publicize testbed activities to national and international security research communities. In January 2000, the IST began planning for an ongoing series of "Workshops on Public Key Infrastructure (PKI) for Advanced Network Technologies," with the first held in April 2000. Workshop objectives are to provide a roadmap for developing and implementing PKI at the point of human-computer interface and to implement a multiagency/commercial testbed for developing Internet security standards.



LSN networking and
infrastructure support




Research networks

Agencies' base LSN R&D activities include providing the networking infrastructure for scientific research that addresses the mission requirements of participating agencies and developing enabling technologies and applications to expand the global-scale capabilities of the Internet. LSN supports research networks such as NASA's Research and Education Network (NREN), DOE's Energy Sciences Network (ESnet), and others, and network management such as at NSF's National Laboratory for Applied Network Research (NLANR). LSN infrastructure support includes implementing evolutionary leading-edge technologies for Federal agency use, including coordination of:

  • Multimodal network testbeds
  • Satellite interconnects
  • Wireless technology development including standards and testing
  • Network attached devices
  • Network attached services
  • Software objects
  • Data sets
LSN coordinates R&D initiatives with other communities such as the Interagency Engineering Task Force+ (IETF+).

Measurement and
network analysis
To meet the measurement and network analysis challenges posed by the rapid growth and expansion of the Internet, NLANR is building and operating a network analysis infrastructure with a primary focus on NSF High Performance Connections sites. R&D areas include active performance measurement, data collection from network entities via Simple Network Management Protocol (SNMP), data related to stabilities and status of Internet routing, and passive header trace data.



Fundamental networking
research


NSF's networking research focuses on the fundamental science and technology needed to facilitate efficient, high-speed information transfer through networks and distributed systems. Funded projects range from network design and performance evaluation to middleware and software frameworks supporting applications on networks and distributed systems. They also address how networks and distributed systems interact with underlying communications technologies. Research areas include agent-based networks, high-speed networks, multicast, multimedia applications, multiple access protocols, network architectures, network design, network management, network security, network systems, object-oriented frameworks for networks, optical networks, performance evaluation, protocols, quality of service (QoS), resource management, traffic control, and wireless and mobile networks.


NSF's cooperative agreement with MCI WorldCom to provide very high performance backbone networking services (vBNS) to universities ended March 31, 2000. NSF has negotiated and completed a no-cost extension of these services.

Special networking
research projects

NSF special projects support larger and in some cases multidisciplinary efforts, specialized hardware and software research, and networks for networking systems research, and help develop research agendas and enhance community development. Accomplishments include production and distribution of a gigabit asynchronous transfer mode (ATM) switch kit and Internet Protocol version 6 (IPv6) development and source distribution. Current and anticipated projects may include all-optical networking, gigabit networking, communications, control theory, databases, devices, distributed systems, operating systems, research priorities in wireless and mobile communications and networking, software, and signal processing. Proof-of-concept demonstrations of novel networking ideas may range from laboratory experiments to national collaborations.

Advanced Networking
Infrastructure (ANI)

The ANI program supports research, development, implementation, and testing of high performance network testbeds and related technologies to support the distributed IT goals of the U.S. research and education community. ANI makes the High Performance Connections awards to colleges and universities, stimulating improvements in campus research networking infrastructure and encouraging development of high performance applications requiring advanced networking services and speeds. The ANI program supports NLANR, which provides technical, engineering, and traffic analysis support for NSF High Performance Connections sites and high performance network service providers (HPNSPs) such as the NSF/MCI WorldCom vBNS.

In FY 1999, the high performance network increased its speed by activating an OC-48 (2.4 Gbps) research link between Los Angeles and San Francisco to serve the CalREN-2 university research and education network. The OC-48 upgrade allows for experimentation with multicast, IPv6, and QoS classes.

NSF's cooperative agreement with MCI WorldCom to provide very high performance Backbone Network Services (vBNS) to universities ended March 31, 2000. NSF has worked with MCI WorldCom, its vBNS+ service, and Abilene to effect a smooth transition for the universities being served.

Internet technologies
program

The Internet technologies program focus areas include complex network monitoring and problem detection and resolution mechanisms; automated and advanced network tools, networked applications tools, and network-based middleware; and usable and widely deployable networking applications that promote collaborative research and information sharing. Current research includes:

  • Prototype testing and evaluation of wireless instrumentation for ecological research at remote field locations . In conjunction with NSF's long-term ecological research (LTER) project,the goal is to develop remote sensing devices for retrieving field data, from the frozen lakes of Wisconsin to rugged mountain terrain in Puerto Rico.
  • Surveying the digital future . This project is studying the impact of personal computing and Internet technologies on society and families. It will address the PITAC's call for research on socioeconomic issues.



STAR TAP


NSF's ANI program coordinates international networking activities that connect the U.S. research and education community with academic counterparts and resources around the world. ANI established the Science, Technology, And Research Transit Access Point (STAR TAPSM) in Chicago to interconnect the vBNS with international advanced networks to support high performance applications and develop new networking technologies. STAR TAP is managed by the University of Illinois at Chicago's EVL and operated by Ameritech Advanced Data Systems.

STAR TAP facilitates the long-term interconnection and interoperability of advanced international networking to support applications, performance measuring, and technology evaluations, anchoring the international vBNS connections program and allowing collaboration with the NGI initiative and the I2 community. STAR TAP has connections with:

Federal agency NGI backbone networks:

  • DoD's Defense Research and Education Network (DREN) (155 Mbps)
  • DOE's ESnet (155 Mbps)
  • NASA's Integrated Services Network (NISN) and NREN (155 Mbps)
  • NSF's vBNS (155 Mbps)

Other U.S. networks:

  • Abilene/University Corporation for Advanced Internet Development (UCAID)
  • vBNS (after March 31, 2000)

International networks:

  • APAN (Asia-Pacific: Japan, Korea, Australia, Singapore [second connection] and Thailand, 70 Mbps)
  • CA*Net (Canada, 45 Mbps)
  • CERN, IUCC (Israel, 45 Mbps)
  • MirNET (Russia, 6 Mbps)
  • NORDUnet (Denmark, Finland, Iceland, Norway, Sweden, 45 Mbps)
  • RENATER2 (France, 45 Mbps)
  • SingAREN (Singapore, 14 Mbps)
  • SURFnet (The Netherlands, 155 Mbps)
  • TANet2 (Taiwan, 15 Mbps)
  • TransPAC (Japan, Korea, Singapore, Australia, 35 Mbps)

Several other countries in Asia and South America have expressed their intention to connect to STAR TAP. Activities now focus on improving performance with higher level services.

STAR TAP provides an integrated approach to the management, scheduling, consumption, and performance measuring of geographically distributed network, computing, storage, and display resources-a collection of resources called the international Grid (iGrid). The iGrid, with the support of STAR TAP, demonstrated international collaborations at SC99. In FY 2000, STAR TAP is moving to a distributed architecture with points of connectivity on both coasts and in Chicago. International STAR TAP participants provide U.S. transit.


Schematic map of the Science, Technology, And Research Transit Access Point (STAR TAP) located in Chicago, Illinois. STAR TAP facilitates the long-term interconnection and interoperability of advanced international networking to support applications, performance measuring, and technology evaluations, anchoring the international vBNS connections program and allowing collaboration with the NGI initiative and the I2 community.



Wireless standards

Today's new markets make extensive use of low-cost, distributed, embedded devices whose architecture supports mobile and self-configuring environments with a combination of wired and wireless technologies. Further development of these markets is hindered by a lack of widely accepted standards, however. NIST is working with standards organizations on issues such as connecting portable wireless devices to traditional networks, service discovery, self-configuration and dynamic resource sharing, mobile code and data, and the software infrastructure needed to create and manage pervasive services and applications.



Active networks


DARPA's active networks program is creating a flexible and extensible networking platform to accommodate the rapid evolution and deployment of networking technologies and to support the increasingly sophisticated services demanded by defense applications. The "active nets" architecture is based on a highly dynamic runtime environment that supports finely tuned control of network services. The packet itself is the basis for describing, provisioning, or tailoring resources to meet delivery and management requirements.

Active networks goals include quantifiable improvement in network services such as fault tolerance, multitiered mobile security, dynamic access control, andaudio/video synchronization and full-rate video over multicast. Some of the challenges in meeting these goals are the defining composite protocols; efficient, secure, and survivable "smart packet" processing; new strategies for routing and service provisioning in large networks with overlapping topologies and mobility requirements; and upgrading services to keep pace with network complexity.



Tolerant networking


DARPA's tolerant networking R&D will develop technologies to support continued operation during attacks on networks, addressing vulnerabilities and issues expected to arise in DoD's emerging network-centric warfare vision. Tolerant networking technologies will strengthen networks by introducing fault- tolerant defenses that emphasize integrity and availability and developing technologies to mitigate potential vulnerabilities in the dynamic creation and management of mission-driven coalitions. Examples are thwarting denial-of-service attacks by constraining an attacker's resources and enabling secure collaboration within dynamically established mission-specific coalitions while minimizing potential threats from increased exposure or compromised partners.

Wireless information technology and networks

The coming generation of terrestrial wireless and satellite communications technologies promises a revolutionary leap forward in information accessibility and an increase in economic yield. In just 20 years, wireless personal communications services have gone from concept to connecting more than 300 million subscribers globally, or 40 percent of the world's installed wired telephone base. Current trends point to a revolution in wireless mobile IT within the next couple of decades that will provide complete freedom of location to the individual. Future-generation wireless systems R&D will enable the U.S. to compete successfully in a worldwide wireless market expected to grow to 700 million subscriber units by 2002.

Research topics include not just bandwidth but management to accommodate a wide mix of traffic types, power limitations, and scalability. Potential applications include crisis management, navigation (for example, in cars or through large buildings), telemedicine (not only for physicians, but for on-site paramedics), and learning, especially in remote locations. Current activities include:

 
  • Research at Rice University on a seamless communication platform that will function in environments as different as high-speed indoor wireless and conventional cellular systems

  • Columbia University research on a comprehensive QoS architecture for adaptive real-time services in mobile ad hoc networks, wireless flow management, and admission control algorithms that allocate bandwidth to adaptive real-time flows

  • University of Southern California research on new concepts of directed diffusion and behavior networks to address the boundary problem between applications and networks

  • University of Arizona research to develop a wireless architecture that uses "fine-grain configurability" and "fast coordinated adaptation" as cornerstones to provide QoS for applications in diverse and dynamically changing environments


Fault-tolerant networks


Denial of service attacks-which take advantage of high-cost protocol checks such as authentication in order to consume resources--can easily disrupt and cripple network operation by rapidly consuming all available network resources, rendering the network inaccessible to legitimate users. Fault-tolerant networks ensure continued availability and graceful degradation of the network infrastructure under partially successful attacks, maximizing network capacity for legitimate users.

Dynamic coalition
management
Traditional system designs have "central nervous systems" that, if attacked, can react by completely disabling the system. Corrupted or malicious components can also lead to a malfunction of the entire system. Dynamic coalition management enables secure collaboration within dynamically established mission-specific coalitions while minimizing potential threats from increased exposure or compromised partners. R&D in this area will develop the capability to dynamically manage and validate operational policy configurations across multiple theaters, securely manage information dissemination within large groups, and augment existing PKI technologies to accommodate rapid revocation and cross-certification.



Quorum: End-to-end
mission success


Defense applications require seamless interoperability, distribution over multiple nodes, and information sharing to support rapidly organized joint and coalition missions. Such systems will be composed of rapidly evolving COTS components deployed in highly dynamic hostile environments. But today's commercial technology emphasizes functional interoperability with virtually no assurance of or control over mission-critical properties such as timeliness, precision, reliability, and security and acceptable tradeoffs among them.

Advances in networking and computing have spurred research and development in distributed computing, ranging from high performance workstation clusters to wide area information retrieval and collaboration environments. These approaches typically emphasize either integrated solutions, which are tightly bound to particular operating systems and platforms, or overlays that support portability at the expense of performance. No single approach exhibits scalability from local to wide area environments and none can provide the assured service and adaptivity necessary to support mission-critical Defense applications in the dynamic, shared, heterogeneous, wide area environments now emerging.

DARPA's Quorum program is developing technologies for an operating system to support global distributed computing for mission-critical applications. The Quorum program premise is that end-to-end QoS management across middleware, operating systems, and networks is the key to providing applications with the end-to-end assurance needed to guarantee mission success in highly dynamic, unpredictable networked environments; scalability over heterogeneous resources whose performance may span a range of several orders of magnitude; and evolvability to keep pace with technological advances. The Quorum program is structured as three interrelated technology development tasks and a fourth integration and demonstration task.

QoS architecture Technologies include methods to specify application-level QoS constraints and allowable tradeoffs; protocols and assurance bounds for QoS negotiation; algorithms for mapping application-level expectations to individual resource and service constraints; instrumentation and analysis methods for dynamically monitoring delivered QoS; and protocols for providing feedback to applications and for triggering adaptation or renegotiation if necessary.

Translucent system
layers
In distributed computing, functional transparency describes a situation in which implementation decisions that have an impact on mission-critical properties are visible to the user, yet frustratingly uncontrollable. R&D in this area will develop translucent services that preserve the benefits of functional transparency while remaining dynamically responsive to QoS constraints imposed by higher layers or feedback from lower layers or the environment, adapting their behaviors through selection or specialization of alternative implementations, policies, or mechanisms. Specific layers addressed include virtual machine APIs, distributed shared memory, distributed objects, communication services, and operating systems.

Adaptive resource
manager
Quorum will develop the resource management technologies to dynamically discover, allocate, and schedule resources from a global heterogeneous pool to an application in accordance with its negotiated QoS constraints. Technologies being developed include methods for characterizing resource capabilities and mapping them to application requirements and profiles; technologies for collecting and maintaining a consistent global view of resource status; and dynamic resource allocation algorithms that yield near-optimal performance and support adaptation in response to workload demands, failures, information warfare attacks, or crisis modes.

Integration, demonstration,
and validation
Realizing the Quorum vision requires coordinated development of constituent technologies, their integration into complete reference implementations of successively greater capability, and their evaluation and demonstration for realistic defense applications. The principal DoD testbed is the Navy's 21st Century Surface Combatant program, which is exploring architectures and technologies for the next generation Aegis combat control system.



Sensor information
technology (SensIT)


Microfabricated sensors are the interface between the physical world and the information world of the future. Delivering massive amounts of data cheaply and in real time, these sensors will be a crucial part of decision-making in the battlefield, surveillance with minimal manpower, and maintenance of equipment. DARPA's sensor information technology (SensIT) program is dedicated to maximizing the useful information that a network of thousands of sensors can deliver to key decision-making points in a timely manner. SensIT's mission is to develop all necessary software for a networked system of inexpensive, pervasive platforms that combine multiple microsensors, embedded processors, positioning ability, and wireless communication. The resulting technology will perform as if a supercomputer were miniaturized and distributed into the environment, with each node computing and collaborating to "see" into its sensor region.

Research challenges include developing large-scale sensor networks employing algorithms for self-assembly of highly dynamic ad hoc networks with minimal latency and maximal survivability, nanocryptography for security, easy to use querying and tasking languages, operating environments, distributed asynchronous algorithms for collaborative signal processing, and internetworked fixed and mobile sensors on humans, microrobots, andvehicles. (DARPA also supports software research in this area in the SDP PCA [page 113].)


Very high-speed
networking


NSA's very high speed networking program provides a high performance research network infrastructure characterized by multigigabit per second trunking speeds and will ultimately support sustained multigigabits per second data flows. In FY 2000, NSA will build on earlier experimentation with optical networking and networking protocols to:

  • Delineate alternate network management solutions to reduce protocol layers
  • Demonstrate optical multicasting by completing some of the agency's optical crossbar network and connecting it to ATDnet MONET equipment to create an all-optical internetwork
  • Extend its transparent optical network to North Carolina in collaboration with MCNC and universities in the Raleigh-Durham area and connect to North Carolina's advanced education network and Internet2 to further investigate long distance issues
  • Complete functionality/viability experiments begun in FY 1999

Extending earlier work in signaling, routing, addressing, and multicasting to meet NSA's addressing and bandwidth needs, in FY 2001 the agency will conduct research in congestion control and multidomain network management. NSA-supported researchers will examine a peer relationship among network management centers that exchange information in a controlled manner to enable end-to-end monitoring and fault isolation of network connections, helping address an environment in which there is little sharing of network management information among vendors, since such information can be used for competitive advantage.



DOE R&D


DOE's R&D addresses the need for a grid architecture that integrates networking, middleware, and applications to support wide area, data-intensive, and collaborative computing to link users with DOE's experimental and computation facilities. DOE has funded 12 projects to develop advanced network architectures and components, advanced middleware services, and advanced network monitoring tools and services with a goal of providing the performance guarantees required by applications.


Next generation
Internet Protocol
(Ipv6)

One of the biggest long-term growth and scalability problems for the Internet is the lack of sufficient address space to globally address all systems with the current Internet Protocol version 4 (IPv4). A new version, IPv6, has been adopted as the next generation (IPng) network layer protocol standard. In August 1999, DOE's ESnet requested and was assigned the first production IPv6 addressing prefix by the American Registry for Internet Numbers (ARIN) and is using it to provide IPv6 services to ESnet users.

ESnet, which provides high-speed connectivity to thousands of scientific researchers at more than 30 DOE sites, has established a production IPv6 network initiative called the 6REN to encourage research and education networks worldwide to provide early production native IPv6 service. Because IPv6 uses 128-bit addresses, it offers a theoretical maximum of about 256 trillion, trillion, trillion addresses. This should allow sufficient addressing scalability to keep up with the current and future growth of the Internet, allowing universal accessibility. Other features designed into IPv6 include built-in security, dynamic automatic configuration, multicast, mobility, QoS, and an ability to allow routing systems to operate more efficiently. Full implementation of IPv6 is needed to effectively deploy large numbers of future wireless devices.

EMERGE

In FY 2000, DOE funded EMERGE, the ESnet/Midwest Research and Engineering Network (MREN) Regional Grid Experimental Testbed. EMERGE's goals are to place routers, Grid middleware, and applications test suites at DOE labs to make them differentiated service (DiffServ)- or QoS-enabled, explore interoperability of ESnet (an IP-over-ATM network) and Abilene (IP-over-SONET), and extend these QoS experiments to Europe and Asia via the NSF-funded STAR TAP.

EMERGE will facilitate advanced data flows not adequately addressed by today's "best efforts only" network. The testbed is driven by DOE computational science applications--particularly climate, combustion, and high-energy physics--that may require guaranteed bandwidth and low-latency networking. Tool development focuses on remote instrument control, data mining and extremely large data sets, collaborative environments, tele-immersion, ultra-high resolution rendered imagery, and unicast/multicast digital video.

Anticipated FY 2000 accomplishments include:

  • 100 Mbps DiffServ networking available to science labs through EMERGE routers
  • A Grid services package (GSP), a common suite of middleware services, deployed for use with DOE laboratory and university applications nationwide
  • Deployment of a common test suite of tele-immersion applications
  • Demonstration of DOE applications over the EMERGE testbed and ESnet
  • Published quantitative measurements that document improvements when DiffServ is enabled
  • Plans for testing alternative QoS strategies

Led by the University of Illinois at Chicago's EVL, in cooperation with ANL and ESnet, EMERGE is establishing a testbed with the following other sites:

Application sites:

  • University of Wisconsin-Madison (Engine Research center)
  • University of Illinois at Urbana-Champaign (center for Simulation of Advanced Rockets)
  • University of Chicago (center for Astrophysical Thermonuclear Flashes)

Tool development sites:

  • University of Wisconsin-Madison (computer science: high-energy physics data grid)
  • University of Illinois at Urbana-Champaign (NCSA: GSP)
  • University of Illinois at Chicago (EVL: application-level tele-immersion network performance test suites)
  • Northwestern University (International center for Advanced Internet Research: video serving and Internet2 DiffServ interoperability over MREN)

Follow-on efforts with NIST will involve testing multiprotocol label switching (MPLS) to manage DiffServ and other flows, extending the GSP and incorporating visualization into monitoring tools, adding haptic and rendering flows to the tele-immersion network performance tests, creating a test suite for multiresolution compressed digital video, continuing interoperability testing and tuning with ESnet and Abilene, and increasing international cooperation. In the latter area, CERN, Holland, Russia, and Singapore have expressed interest in participating in DiffServ and MPLS tests.

Digital Collaboration
Services (DCS)
ESnet is providing the DOE research community with the next generation of videoconferencing capabilities. Digital Collaboration Services (DCS) makes the audio, video, and data interactivity of a dedicated videoconference facility available on desktop computers, taking another step toward the virtual laboratory where scientists at distant DOE sites can collaborate in real time. DCS 2.0 is based on communications and applications standards adopted by the International Telecommunications Union (ITU) that allow interoperability with other networks and users.

DOE networking tools In coordination with DARPA, DOE has developed Netlogger, a tool that provides real-time analysis for end-to-end performance monitoring. An easy-to-read graphical user interface (GUI) interface indicates bottlenecks in the network. DOE has also developed PingER, a network monitoring tool that provides ping end-to-end reporting on Internet link performance. PingER measures response, jitter, loss, and reachability; provides national and international coverage of 536 remote nodes at 381 sites in 55 countries on six continents; and correlates with network performance monitoring tools. PingER primarily supports the high-energy physics community.



Global Observation
Information Network
(GOIN)


The GOIN demonstration project in Hawaii in March 1999 linked U.S. global observation researchers with partners in Japan. Using links supplied by NASA, DoD, and NSF, NOAA's PMEL demonstrated the first NOAA applications over the NGI, including OceanShare, a collaborative environment for oceanographic research, and 3-D tools using VRML to demonstrate the evolution of El Niño, fisheries larval drift, and fur seal feeding trips. Working with these innovative applications over the NGI provides greater opportunities for collaborating and accelerating scientific inquiry. Additional GOIN demonstrations used distributed real-time data to produce visualizations of natural Earth phenomena--such as sea ice movement and ocean thermal gradients--and used satellite thermal data to identify agricultural productivity.



Load balancing


NOAA has experimented with geographically distributed mirroring and load balancing to alleviate peak loads without wasteful over-provisioning. This is especially useful at the agency's National Hurricane center, which experiences peak NOAA traffic during a hurricane landfall along the populous eastern seaboard. These experiments demonstrated their usefulness during the past hurricane season, when the load on NOAA's Web services exceeded that of any other governmental body as measured by unique accesses per month tracked by Media Metrix. The hurricane information was made available at three geographically distributed access points while critical operational weather data flowed uninterrupted.



LSN applications
R&D


LSN applications R&D focuses on applying beta software from commercial providers and the research community to mission problems such as NOAA's weather modeling and NIH's medical collaboratories. Near-term development and deployment include Internet tools, products, and services such as high performance parallel interface (HiPPi) standards, application-specific multicast and QoS, and Internet videoconferencing development.



Next Generation Internet


The Federal NGI initiative, in coordination with other Federal agency networking research programs, is creating the technical and infrastructure foundation for a more powerful, flexible, secure, and intelligent Internet in the 21st century. Tightly coupled with LSN network R&D and infrastructure support, NGI is helping build partnerships among academia, industry, and government to keep the U.S. at the cutting edge of information and communications technologies and stimulate the introduction of new multimedia applications in the Nation's schools, businesses, and homes.


The NGI initiative has two thrusts:

Thrust 1 . NGI will advance R&D and experimentation in next generation networking technologies to add functionality and improve performance in:

  • Reliability
  • Security
  • Robustness
  • Differentiated services including multicast and audio/video-also known as quality of service (QoS) and class of service (CoS)
  • Network management, including allocation and sharing of bandwidth

Other areas of concern for Thrust 1 include:

  • Middleware
  • Developing and deploying a software platform using high-speed Linux at a wide range of end-user sites
  • Protocol stack tuning to improve performance

    Thrust 1 is supported by DARPA's SuperNet, an NGI testbed providing a 1,000-fold increase in end-to-end performance over 1997 capabilities, or approximately 1 Gbps for research end users.

Thrust 2 . NGI will develop and demonstrate revolutionary applications in enabling technologies such as:

  • Collaboration technologies
  • Digital libraries
  • Distributed computing
  • Privacy and security
  • Remote operation and simulation

    And in disciplinary applications such as:

  • Basic science
  • Crisis management
  • Education
  • The environment
  • Federal information services
  • Health care
  • Manufacturing
    Thrust 2 is supported by an NGI testbed providing a 100-fold increase in end-to-end performance over 1997's Internet, or approximately 100 Mbps for research end users.



Thrust 1
accomplishments
and plans


NGI Thrust 1 programs focus on R&D and testbed deployment and demonstration of technologies to provide network growth engineering and to enable the reliable and secure provisioning, management, and end-to-end delivery of differentiated classes of service.


NSF


NSF has implemented a wide range of networking research programs addressing network growth engineering, QoS, and security. Network engineering research projects focus on scaling up the Internet in size and speed, developing performance measurement and middleware, and addressing flow-based or aggregate-based QoS and congestion control. Specific projects address:

  • Congestion management
  • Terabit routing with QoS
  • Advance reservations, routing, and scheduling
  • Scheduling algorithms for high-speed switches with QoS guarantees
  • QoS-enabled middleware for global distributed computing

NSF programs address network security and assurance, end-to-end security, and policy functions such as pricing and cost recovery. Specific projects include secure and robust agent technology and secure multicast.

 

High performance network
service providers (HPNSPs)
NSF has designated a category of commercial HPNSPs that offer advanced network services over broadband networks to university and Federal agency sites to provide the high performance services needed by the NGI. The Abilene network-the first network to receive this designation-coordinates closely with the JET to provide connectivity and meet other requirements of the NGI community. Additional network providers have expressed interest in providing HPNSP services.



DARPA's SuperNet


The goal of DARPA's SuperNet program is guaranteed ultra-high bandwidth on demand over the shared national networks infrastructure. SuperNet will demonstrate multi-Gbps transmission end-to-end by:

  • Developing streamlined networking protocol stacks
  • Stressing end-to-end architecture and performance
  • Developing technologies for regional, metropolitan, and local area networks
  • Implementing a dynamically controlled optical layer
At the SC99 conference, NGI researchers and industry partners set 2.4 Gbps world network speed and performance records, transmitting studio-quality high-definition television (HDTV) streams long distance over NGI testbeds to SC99.


The goal of DARPA's SuperNet program is guaranteed ultra-high bandwidth on demand over the shared national networks infrastructure.


NIST

 

Advanced encryption
standard (AES)
NIST is developing AES, a royalty-free encryption standard to be used by government and industry to protect information over the next 30 to 50 years. It is testing candidate algorithms using multiple platforms and computing systems, comparing efficiency testing results and determining the validity of cryptanalysis, and validating conformance of COTS products.

Public key
insfrastructure (PKI)
NIST's PKI program will ensure the development of commercially available PKI products and services that are interoperable and sufficiently secure to meet the needs of Government agencies and the general public. NIST is helping users and suppliers establish PKI standards and specifications, interoperability, correctness, and quality, and is publishing its analysis of PKI component security requirements and guidance and developing pilots for automated key recovery systems and Web-based electronic certification.

Internet quality of
service (QoS)
NIST is expediting commercial deployment of standardized Internet QoS technologies, devising tools to aid developers of adaptive Internet applications, developing techniques and tools to test distributed multiparty QoS routing and signaling protocols, and evaluating proposed algorithms and protocols for scalable QoS routing and signaling. For example, NIST has deployed NISTnet, a tool that emulates controllable Internet performance dynamics. NISTnet was demonstrated at SC99 (page 102).

Hybrid-fiber coax
(HFC) access
Hybrid-fiber coax (HFC) is a means for connecting home computing systems to commercial communications networks. NIST is evaluating protocols for HFC media access control (MAC) and is studying end-to-end performance to improve ATM and TCP/IP traffic over HFC networks. NIST has incorporated HFC network protocols into its ATM network simulator and has published results on contention resolution algorithms, bandwidth allocation, and priority schemes for end-to-end ATM and TCP/IP traffic over HFC networks.

Dense wave division
multiplexing (DWDM)
NIST is accelerating the development of dynamically reconfigurable DWDM networks and evaluating proposed algorithms for wavelength assignment and routing in WDM networks. WDM dramatically increases bandwidth on existing fiber.

Agile networking
infrastructures
NIST is developing agile networking technologies to enable programmable and reconfigurable communication infrastructures, middleware for adaptive, reconfigurable distributed systems, and measurement techniques to enable resource control in active networks. NIST is also evaluating measurement and standardization requirements for networked pervasive computing, which allows mobile wireless network access to the Internet.



Thrust 2
accomplishments
and plans


NGI Thrust 2 applications span a wide range of societal and technical areas, and the 100x testbed that supports applications development provides high performance, reliable service for a rigorous applications testing environment. The 100x testbed network:

  • Develops and implements advanced technologies including QoS, multicast, security, network management, and protocol performance management
  • Emphasizes end-to-end performance rather than point-to-point service
  • Emphasizes transparency
  • Includes campus environments (LANs) and user end systems
  • Provides coordinated management of the network resources
  • Provides reliable service through a robust architecture (NGI network backbone, regional and GigaPop structure, and Next Generation Internet Exchanges [NGIXs])

The network providers implement these capabilities, which are in turn coordinated by the JET. In FY 1999, the JET focused on interconnection points among the JETnets (vBNS, ESnet, NREN, DREN, NISN, and Abilene) at the NGIXs. NGIXs were implemented at NASA Ames Research center (NGIX-West), Chicago (NGIX-Midwest), and University of Maryland (NGIX-East), and peering arrangements are currently being developed at these exchanges. Further cooperation with the university community was provided by joint meetings among the Federal agencies, GigaPop operators, university campus representatives, and Internet2. This coordination has focused the university and Federal committees on NGI network performance measurements and upgrading campus networking capabilities and services.

In FY 2000 the JETnets implemented multicast in the NGI backbone. University campus networks need to implement and tune multicast tools in their networks to enable end users to effectively use multicast services, which are particularly effective in broadcasting lectures and meetings, providing interaction among remote environments in real time.

IP QoS is being deployed in a collaborative testbed. ESnet, NREN, and vBNS participate in the Qbone project that is implementing inter-domain DiffServ. vBNS has offered reserved bandwidth service using RSVP/ATM. All JETnets are part of the 6Bone project that is developing IPv6, which provides protocols for a significantly increased address space and expedites network services such as QoS. vBNS is testing a native IPv6 service over separate routers.



Performance
measurement and
improvement


Measuring the performance of the Internet is critical in identifying problems such as bottlenecks and resolving them to provide improved network performance. Active measurement capabilities of the JETnets include the Advanced Measurement Platform (AMP) and Survey to provide throughput, round trip time (RTT), topology, one-way delay, delay variation, and packet loss data. Passive network performance measurements use Coral/OCXmon, which is non-invasive and provides detailed characteristics of individual flows. Additional measurement and visualization tools are under development. The JET is identifying large data flows and high-bandwidth users on the NGI networks to offer them assistance in improving the performance of their applications. The Team will provide fine-tuned laptops to measure end-to-end performance from specific subnets to assist local network operators in identifying bottlenecks and other network problems. The JET is fostering close cooperation among commercial vendors and the NGI agencies to develop Linux software on a platform to tune the protocol stack to provide improved, flexible Internet performance for a range of high performance applications.

 



NSF applications

Hundreds of applications are being developed by NSF-funded university researchers. While many require the high-bandwidth capabilities of the NGI, others require the advanced services being developed by the NGI initiative, such as QoS, security, multicast, collaboratory tools, and visualization software. Examples of the revolutionary applications funded by NSF include:
Ecology
Species analyst

The species analyst is a set of tools for simultaneous access to multiple biological collection databases that can be used to predict species distribution as a function of global climate change, species invasions, and species disturbances by humans. Led by the University of Kansas, this project is also being conducted at the University of California-Berkeley, Harvard University, the University of Michigan, the University of Nebraska, SDSC, and the University of Mexico.

 

Education and teaching
Collaborative development
of 3-D life science
educational resource

Yale University's School of Medicine and NLM are creating 3-D educational materials using NLM's life science image databases. To assess levels of acceptable video compression, Yale and NLM experimented with digital video encoding of the video records of patients suffering from neuralogic motion disorders such as Parkinson's disease.

Two data mining and
visualization projects

The University of Illinois at Chicago's CAVERN project is conducting research in tele-immersion and intelligent data mining for use in computational science, collaborative distance-independent education, and the Every Citizen Interface to the National Information Infrastructure.

The Web TerraFly project at Florida International University allows users to fly over and manipulate data retrieved from a high performance semantic multimedia spatial database that includes satellite and aerial photography data.

Megaconference With Ohio State University as the lead institution, this pilot project has established an ongoing videoconference collaboration linking 65 network engineers and researchers on three continents to discuss advanced networking technology.

Widea area
interactive teaching
With Oregon State University in the lead, this application has provided the first regular scheduled Internet2 video class in graduate-level plant pathology among Oregon State, Kansas State University, and the University of Nebraska. The application uses enough bandwidth to enable high-quality video at these institutions.

Widea area interactive teaching With Oregon State University in the lead, this application has provided the first regular scheduled Internet2 video class in graduate-level plant pathology among Oregon State, Kansas State University, and the University of Nebraska. The application uses enough bandwidth to enable high-quality video at these institutions.

Humanities, arts, and
archaeology

center for Electronic Reconstruction of
Historical and
Archaeological
S ites (CERHAS)


CERHAS--a collaborative research project of the University of Cincinnati, Cleveland State University, and Wright State University-is building software for students and professors in the arts and humanities. CERHAS researchers are exploring real-time, networked motion capture, avatars, artificial personalities, and secure personal operating systems to enable collaborative virtual reality. Target applications will be distributed to users at participating institutions via Web-DVD.

Variations digital music
library
Variations provides access to more than 5,000 titles of near CD-quality digital audio to computer users in the Cook Music Library of the Simon Music Library and Recital center at the University of Indiana-Bloomington.

Manufacturing
Scaling Internet
connections to
support research
applications

This NSF Science and Technology center project focuses on automated machining direct from CAD sketches. Brown University, CalTech, Cornell University, the University of North Carolina, and the University of Utah are the collaborating institutions. Using the vBNS, Brown has been able to maintain its software on Utah systems, speeding research experiments and facilitating the display of 3-D graphics at remote locations to the point where applications appear to be running locally.

Multimedia
Large-scale video
network prototype

Supported by the Southeastern Universities Research Association (SURA) and BBN/GTE Internetworking, this advanced networking project, which involves researchers at the University of North Carolina-Chapel Hill, the University of Tennessee-Knoxville, the Georgia Institute of Technology, and NYSERNet, is investigating and deploying model architectures for video services to support distributed video in higher education.

Remote science
and networking

Remote observing


A new remote observing application will provide authorized scientists anywhere on an NGI network with access to the more than $1 billion in research instruments at the Mauna Kea Observatories in Hawaii, eliminating the need to travel and acclimate to the remote 15,000-foot-high facility. Institutions involved in this research include the University of Hawaii, the University of California, CalTech, the University of Washington, and others.


The Mauna Kea Observatories sit at an altitude of 4,200 meters, near the summit of Mauna Kea on the Island of Hawaii. Currently, nine telescopes plus the Hawaii Antenna of the Very Long Baseline Array are in full operation. The cloud-free, dry, and stable atmosphere on Mauna Kea permits more detailed astronomical studies than are possible elsewhere. NGI technologies will enable scientists at remote locations to access the instruments at this unique site.
Telemedicine
Distance-independent
telemedical diagnosis


Eastern Carolina University along with the Ohio Supercomputer center, the Ohio State University, the Northeast Parallel Architecture center, and Syracuse University are funded by NSF and NLM to study biomedical applications that require NGI capabilities such as:

  • Novel rendering techniques for extremely large volumetric data sets and methods to guarantee network performance
  • Evaluating IP video over the network to determine the technologies needed to meet diagnostic quality needs and standards
  • An investigation into the effect of real-time encryption of IP video to ensure patient confidentiality
  • Qualitative analysis of how the NGI can be used to transmit cine-angiograms rapidly without degrading the images

Psychological services
for the hearing-impaired
This University of Rochester research project uses teleconferencing technology to bring deaf consumers with mental health care needs into contact with sign language-fluent practitioners. Directly communicating with sign-fluent psychologists and social workers avoids problems caused by differences between American Sign Language and English as well as the awkwardness of using a sign language interpreter in a sensitive mental health setting.

Veterinary science
Virtual rounds

Virtual rounds is an application being developed by the colleges of veterinary medicine at the University of Tennessee-Knoxville, Auburn University, the University of Georgia-Athens, and North Carolina State University to share animal clinical cases via live videoconferencing. Sessions will be captured and stored on a video server to be available on demand, creating a potential for digital video libraries to support veterinary medical instruction.

Weather, atmospheric
research

Advanced Regional
Prediction System
(ARPS)


ARPS is a fully automated, functionally complete numerical forecast environment that can predict weather on scales ranging from continents to cities. It was the first model of its kind in the world designed for massively parallel computers and is uniquely suited to accommodate observations from Doppler radars. NSF, NOAA, the Federal Aviation Administration (FAA), AlliedSignal (now part of Honeywell), and American Airlines support this University of Oklahoma research.

Space Physics and
Aeronomy Research
Collaboratory
(SPARC)

Space physics researchers at some 380 sites around the world study phenomena such as magnetic storms on the sun that can interfere with radio and television reception, disrupt electrical power transmission, and threaten orbiting spacecraft. They do so by controlling and gathering data from more than a dozen instruments located around--and above--the globe, directly accessing advanced supercomputer models of upper atmospheric phenomena and state-of-the-art communication tools that include "chat rooms" and a shared whiteboard. All research is recorded for replay, annotation, or asynchronous collaboration with colleagues.



DARPA applications


DARPA's NGI applications include:

  • CSU-CHILL radar for remote sensing and meteorological analysis. Doppler radar data are collected at distributed sites, processed, and made available on the network at speeds of 240 MBps to 2.88 GBps. The data are used to determine meteorological characteristics such as rain, hail, ice crystals, and turbulence.
  • Matisse computer microvision workstations optically monitor microelectromechanical systems (MEMS) device response over varying focal planes. A scientific microscope provides ultra-high resolution motor control and stroboscopic light-emitting diode illumination, and a closed coupling device camera system provides a megapixel camera and frame grabber. A typical data set is 10 GB.
  • Digital Earth. Part of an interagency initiative, this application is an open, distributed, scalable, multiresolution 3-D representation of the Earth into which massive quantities of geo-referenced information can be embedded. Digital Earth uses VRML and a standard browser and plug-in to navigate the 3-D model.



NIH applications

NLM awards

NLM initiated a three-phase NGI research program to develop innovative medical projects to demonstrate application and use of NGI capabilities, including:

  • Quality of service
  • Medical data privacy and security
  • Nomadic computing
  • Network management
  • Infrastructure technology for scientific collaboration

NLM has awarded 15 Phase 2 awards to develop local telemedicine testbed applications. Those that prove effective will be implemented for end users under Phase 3 awards. Phase 2 awards under way or to be made in FY 2000 include:

Applications layer
security solution for
stationary/nomadic
environments

This project will evaluate security techniques within an open security architecture. The proposed solution is based on security shared among collaborating parties, nomadic computing, and the privacy of medical information.

Biomedical tele-immersion

By combining teleconferencing, telepresence, and virtual reality, tele-immersion enables teachers and students to interact with 3-D models, especially in surgical education. NGI data privacy and security guarantees allow tele-immersive environments to be derived from models of patient data.

Networked 3-D virtual
human anatomy

The goal is to build an online virtual human cadaver, based on the Visible Human data set, that students can explore with a variety of tools.

Patient-centric tools
for regional collaborative
cancer care

 

This project will investigate the application of collaborative tools in the Seattle area Cancer Care Alliance (CCA) to:

  • Enhance the CCA partners' existing clinical care programs to be highly collaborative patient-centered interdisciplinary efforts
  • Allow a fully integrated team approach to cancer, including state-of-the-art diagnosis, treatment, and management through collaboration by distributed clinicians and researchers
  • Accelerate the dissemination and application of new knowledge for cancer diagnosis and treatment

Radiation oncology
treatment planning/care
delivery application

 

This project will implement and evaluate NGI capabilities for radiation oncology treatment planning and care delivery, providing diagnostic support, treatment planning, and remote verification of proper equipment operation from the Comprehensive Cancer center to a Johns Hopkins University treatment facility.

Rural health science
education

 

This project is developing a plan to evaluate computer and interactive compressed video technologies to support rural health science education. Such an application will enable delivery of interactive educational programming, such as continuing medical education, clinical information systems, library services, and consultation. Beneficiaries include students and health care professionals.



DOE applications

Collaboratories



Funded in FY 1999, DOE's work in NGI applications was largely completed in FY 2000. DOE funded five collaboratories with multiple sites including universities and national laboratories. DOE supports two testbeds to demonstrate advanced services to university sites and improve access and capabilities for university applications development researchers.

Distributed X-ray
crystallography
Using tools being developed by the ANL Globus group and the DOE 2000 Common Component Architecture Forum, this project is building NGI network-based instrumentation including high-speed data collection, reduction, storage, and visualization, and real-time instrument control for the acquisition of macro molecular x-ray crystallographic data from the LBNL Advanced Light Source.

Integrated grid
architecture and the
Earth system grid
DOE is creating an integrated grid architecture built on a fabric of networks, archives, computers, display devices, and other technologies and associated local services. Grid services include protocols, authentication policy, resource management, instrumentation, and discovery. A remote data toolkit, remote computational toolkit, visualization toolkit, asynchronous collaboration, and remote sensors will help address the needs of end-user applications. One such application is the Earth system grid for climate modeling, projections, and impact assessments, in which more than 100 universities, laboratories, and centers participate.

Combustion Corridor

Combustion of fossil fuels accounts for approximately 85 percent of the energy expended in the U.S., and modeling this process is critical to increasing efficiency and reducing pollution. Combustion modeling codes can easily generate terabytes of data that must be analyzed by researchers dispersed across the U.S. The DOE Combustion Corridor application uses grid storage API, global naming services, the Globus resource reservation system, networks, disk caches, and PC clusters to interactively and collaboratively visualize these data.

Such visualizations typically represent complex, 3-D scientific problems varying over time, such as how two gases mix in a turbulent environment. To visualize these models, researchers previously required access to very powerful computing systems, and moving their large files onto local workstations was either impossible or impractical. In image-based-rendering-assisted volume rendering (IBRAVR), developed at LBNL, a large data set is now partially pre-rendered on a large computational engine close to the data and the final image rendering is performed on a workstation.

Corridor One DOE's Corridor One--visualization at a distance-uses the grid to integrate data servers, analysis and manipulation engines, visualization backend servers, visualization clients, display device interfaces, and advanced networking services.



NASA applications


Biomedical image
collaboratory



This experiment relies on a Visible Human Viewer developed on OpenStep, an Apple API. This application can show sections of a human body so a researcher can recognize anatomical objects. Anatomical terms can be attached using NLM's Unified Medical Language System (UMLS).

Collaborative electron
microscopy of
quarantined samples
This NGI experiment will remotely control and observe microscope imaging of quarantined samples returned from Mars at NASA's Ames Research center (ARC) from remote sites at ARC, ORNL, and Oregon State University.

Digital Earth/Mars/Sky Digital Earth (DE) is an interagency initiative to define and prototype a framework to make all U.S. geo-referenced data available via Web-style point-and-click user interfaces over high performance networks. NASA's DE research is a collaborative effort with the Federal Geographic Data Committee and the Open Geographic Information System Consortium, an organization of business, academic, and government officials concerned with interoperability of geographic data. This work supports NASA's goal of human telepresence throughout the Earth and solar systems, which requires high performance remote access and visualization of large Earth and space data sets.

Distributed video
observation of shuttle
processing and
launch activities
This project will demonstrate the delivery of broadcast quality video streams over a high performance IP network. Multiple video sources at the Kennedy Space center will be transmitted to desktops at a variety of locations, and viewers will be able to choose among available sources. The primary goals are to use COTS technologies to provide greater video surveillance flexibility at reduced cost.

Virtual collaborative
clinic (VCC)
NASA, in coordination with NLM, the Navajo Nation, Abilene, vBNS, and the CALREN-2 GigaPoP, is developing the virtual collaborative clinic (VCC) to demonstrate a high performance testbed that allows medical colleagues to simultaneously review medical images remotely in real time. This requires 30 to 50 Mbps and multicast technology.



NIST applications


NIST has deployed a pilot collaboratory for robotic arc welding and published preliminary results on usage. NIST researchers are deploying and assessing a manufacturing collaboratory testbed, evaluating collaboration processes and an industrial pilot collaboratory, and developing quantitative evaluation methods for collaboration technologies.

NIST has developed interfaces for virtual manufacturing applications, a testbed to demonstrate real-time, multiuser, interactive simulation of manufacturing equipment control, and extensions to VRML for device behaviors. NIST is building a remote interface to a weld cell with real-time display of a data-driven VRML weld controller model and synchronized real-time video with a VRML model.

In FY 2001 NIST plans to:

  • Link the full immersion CAVE environment at the University of Illinois to a welding robot at NIST via the NGI
  • Establish remote control of a welding robot in the virtual world



New Starts


SII Program
FY 2000-2001

The Scalable Information Infrastructure (SII) component of LSN R&D responds to the PITAC recommendation that the Federal government establish a high-priority focus on research seeking fundamental advances in the science and engineering of networking technologies that will enable the Internet to grow or "scale up" to meet requirements for ever-increasing numbers of users and types of devices. Begun in FY 2000, the SII R&D program is still being developed to address demands for 21st century technologies such as deeply networked systems, anytime/anywhere connectivity, and network modeling and simulation.

  • Deeply networked systems. Researchers will develop software and network architectures enabling large numbers of diversified devices such as low-cost wireless sensors to provide real-time information from distributed sensors. Such devices could provide, for example, real-time information on air and water pollution and improve our ability to monitor the environment and respond to man-made disasters; "guardian angels" that monitor the health and safety of fire fighters, law enforcement officials, soldiers, and home health care patients; and crisis management centers that use sensors carried by response teams and airplanes to improve responses to forest fires, floods, and hurricanes. Research will focus on naming, addressing, and network configuration, and on developing less expensive network interfaces.
  • "Anytime, anywhere" connectivity. Research will focus on wireless technology to provide all U.S. citizens with high-speed connections to the Internet, whenever and wherever they are needed. Wireless networks can extend services such as distance learning and telemedicine to remote rural areas in the U.S. and to markets in developing countries. This research will emphasize developing standards, improving bandwidths, and addressing wireless architectures to provide the wide geographic and mobile coverage needed for wireless Internet connectivity.
  • Network modeling and simulation. The goals are to develop tools to model network behavior to test new network technologies before implementing them on testbeds and the full Internet, and to enable faster than real-time simulation to allow network operators to forecast network behavior, allowing them to intercede to prevent network congestion or collapse.

Federal agency SII programs

Federal agency SII activities will explore fundamental aspects of networking, including the ability to depict and troubleshoot very large systems through network modeling, simulation, and emulation; underlying network technologies such as optical and wireless; and middleware enabling large-scale systems, information management, and information and service survivability. DARPA and NSF-supported researchers are developing the network architecture, middleware, and human interfaces for pervasive networking environments-systems capable of effectively and efficiently networking large numbers of very small distributed and embedded devices-including highly scalable and adaptable network protocols for evolving application requirements; QoS, such as for network control and scheduling; and network security and privacy tools. Research addressing development of on-demand wideband access in scalable mobile networking will aim for multimegabit wireless access with seamless global roaming.

Agile networking infrastructures

Future networking communities will consist of highly fluid groups and locations accessing the Internet through a wide variety of information services, embedded devices, and facilities. In this environment, network infrastructures must be "agile," meaning able to respond quickly to variations in connectivity and bandwidth requirements and differences among access devices, and to dynamically reorganize end-to-end network services in response to changes in the composition of collaborating teams. The agile networking infrastructures effort will support the specification, standardization, measurement, and testing of protocols and software to support these capabilities. Program goals include the development of virtual overlay networks (VONs) that can be established and readjusted on demand in minutes by distributed mobile collaborative users.

Network research center

NSF and NIST plan to establish a high performance distributed network research center enabling researchers to study new protocols they are developing to improve Internet performance. The center will focus on discovering and managing the limiting features of networks, providing advanced performance features not now available on large-scale networks; distributed applications and their interaction with network protocols; middleware enabling efficient computer-aided software development; and tools and expertise to enhance the Internet knowledge base, particularly for minority-serving organizations.

Network security

NSF- and NIST-funded researchers will explore techniques, protocols, and procedures with the potential to enhance the overall security of packet-switched telecommunications networks, including the ability to support secure tele-collaboration activities.

Prototype access testbeds

Researchers will develop prototype testbeds to demonstrate very high performance campus networks that can serve as models for upgrading campus networks nationwide to take advantage of new backbone capabilities. The research will test potential delivery systems ranging from wireless to all optical and capabilities for interaction among wireless, optical, and wireline systems.

left right