|
NGI Implementation Plan
Section 3.1: High Performance Connectivity |
 
|
|
|
|
|||
 |
|
||||||||||||||||||||||||||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
|
|
||||||||||||||||||||||||||||||||
|
|
|||||||||||||||||||||||||||||||||
3. Goal 2
|
Goal 2: Next Generation Network Testbed (continued) |
||||||||||||||||||||||||||||||||
|
|
|||||||||||||||||||||||||||||||||
|
3.2 Goal 2.2
|
Goal 2.2: Next Generation Network Technologies and Ultrahigh Performance Connectivity |
||||||||||||||||||||||||||||||||
|
|
|||||||||||||||||||||||||||||||||
|
3.2.1 Introduction
|
Introduction This Goal 2.2 addresses the development of ultrahigh speed switching and transmission technologies, and the demonstration of end-to-end network connectivity at 1+ Gbps. Because of its high risk and pioneering nature, networks involving about 10 NGI sites and applications will be implemented. Attaining this goal, together with the technologies developed in Goal 1, will be the pathway to terabit-per-second (Tbps) networks, operated with the appropriate network management and control and guaranteed end-to-end QoS. Working in partnership with industry is the key to a shared infrastructure that can be profitably used to support high end scientific users and large numbers of commercial users. This is a joint agency effort with DARPA as the lead, with participation by NASA, NSF, DoE (beginning in FY 1999) and other Federal agencies. The following sections describe the NGI Goal 2.2 implementation plan for the agencies included in the Congressional FY 1998 NGI appropriations. DoE is not a formal participant in the NGI in FY 1998. The Administration plans to propose adding DoE as a formal participant beginning in FY 1999. DoE's participation in the NGI beginning in FY 1999 is described in Appendix E. |
||||||||||||||||||||||||||||||||
|
|
|||||||||||||||||||||||||||||||||
|
3.2.2 Strategy
|
Strategy The technology of choice to achieve a Tbps network is wavelength division multiplexing (WDM), which is a technique of mixing many wavelengths onto the same optical fiber. This is equivalent to opening up the narrow communication links into multiple-lane communication highways. DARPA's Broadband Information Technology (BIT) program has pioneered much of today's WDM effort. Whereas WDM is currently implemented at the physical layer, the aim here is to integrate WDM and its management with the upper layers of ATM and IP. To this end, a new network architecture that specifically addresses the access loop will be designed. These technologies and architectures hold the promise of eventually satisfying the goal of an infrastructure that is shared by both high end users, typical users, and network researchers. Some of the network nodes will be chosen to coincide with some Goal 2.1 nodes. The architecture will be designed such that portions of the Goal 2.1 networks can interconnect to and gracefully evolve into Goal 2.2 networks and further demonstrate the continuing evolution of network performance. Partnership with long distance and local exchange carriers is a key to ensuring the early adoption of this technology and to ensuring its affordability. With respect to the generation-after-next network technologies, Goal 2.2 will explore optical, electronic, and hybrid switching techniques. The goal is to pave the way to Tbps packet switching systems. On the optical side, hybrid dense WDM and optical time division-multiplexing (TDM) systems will be explored. On the electronic side, a distributed electronic switching design, as opposed to a single monolithic Tbps module, will be pursued. Resulting devices and systems will be initially field tested in the Goal 2.2 network. These research activities will be a combination of government, academia, and equipment vendor efforts and collaboration. |
||||||||||||||||||||||||||||||||
|
|
|||||||||||||||||||||||||||||||||
|
3.2.3 Metrics
|
Metrics Goal 2.2 will focus first on the deployment of at least one metropolitan network (for example, five-node network), with the appropriate management and control software. This network will be operational at least 80 percent of the time and will be capable of delivering 20 Gbps to each node. As the tools of Goal 1 and the broadband local access technologies become available, they will be incorporated into this network to experiment with providing ultra-high speed end-to-end QoS, management of lead user infrastructure, data integration, and network security. A true test of the success of a network is the range of new applications it will enable. To this end, at least 10 new applications will be tested on this ultrahigh speed network. In Phase 2 of Goal 2.2 the network will be expanded to a wide area network with about 10 nodes performing similar functions, as in Phase I. In this phase, agency applications will be linked to demonstrate a distributed, heterogeneous, multidomain, and multivendor environment. Since the number of nodes that can be built is limited by the available resources, the scalability and network management of hundreds of ultrahigh speed nodes will be examined by simulation and modeling. The following is a description of the implementation plans and milestones by participating agencies. |
||||||||||||||||||||||||||||||||
|
|
|||||||||||||||||||||||||||||||||
|
3.2.4 Agency Specifics
|
Agency Specifics |
||||||||||||||||||||||||||||||||
|
|
|||||||||||||||||||||||||||||||||
|
3.2.4.1 DARPA
|
Defense Advanced Research Projects Agency Wide Area Broadband Core DARPA's Broadband Information Technology (BIT) program has developed basic WDM transmission capabilities and will soon demonstrate a metropolitan network of five nodes, with link transmission capacities of 20 Gbps. DARPA will extend these technologies and deploy them in more complex, mesh-like topologies that involve long distance links. The metropolitan testbed will be expanded into a wide area network with about 10 nodes using WDM technology. This wide area backbone will have sufficient aggregate transmission and switching resources to support hundreds of users at Gbps rates. This network will share the fiber facilities with the general public. Tbps Multiplexing and Switching DARPA will develop the generation-after-next multiplexing, switching, and routing technologies that will bridge the gap among packet-based Gbps tributaries and the WDM-based optical core. This task will also lay the groundwork for the direct optical support of packet-based communication. A major component of this task will be to investigate statistically sound techniques for performing "space-division"-like spreading of the resultant TDM traffic across a set of wavelengths. A second component will be the design and demonstration of a highly parallel and distributed switching fabric. Taken together, these efforts will enable the development of a highly distributed approach to Tbps switching, based on a combination of optical and electronic technologies, with many-to-many multicast capability. Broadband Local Trunking The need to provide select sites with "orders-of-magnitude-above-average" access to the network core has been a recurring source of delay in commissioning advanced research facilities. This task will explore novel and cost effective approaches to delivering broadband access to select sites within a geographically restricted area. DARPA will examine the terrestrial extension of SuperNet rate facilities to the building and explore the effectiveness of high capacity (>150 Mbps) radio frequency (RF)-based trunking. In addition, wireless broadband local access will be addressed as one of the DARPA tasks in Goal 2.2. Technology Demonstration and Field Trials Most of the technologies to be developed by the previous tasks are associated with the physical, link, and networking layers. This task will seek opportunities to demonstrate the newly developed capabilities through collaboration with some of DARPA's application oriented activities, such as the Human Computer Interaction, Information Management, and Intelligent Collaboration and Visualization programs. Milestones
|
||||||||||||||||||||||||||||||||
|
|
|||||||||||||||||||||||||||||||||
|
3.2.4.2 NSF
|
National Science Foundation NSF will participate actively in NGI Goal 2.2 through select connections to the ultrahigh speed networks, as well as the direct funding of competitive research proposals by campus-based investigators. NSF will participate with DARPA and other agencies in ultrahigh speed networking links and technologies through NSF's two major supercomputer partnerships (Partnerships for Advanced Computational Infrastructure, PACI) centered at San Diego Supercomputing center (SDSC) and the National center for Supercomputing Applications (NCSA). The focus will be on protocols and technologies for advanced, distributed computing. NSF strategies will center first on peer evaluation of Goal 2.2 research and new end-to-end network technologies, and later on the deployment of Goal 1 technologies to Goal 2.2 networks. Among other activities, NSF will:
|
||||||||||||||||||||||||||||||||
|
|
|||||||||||||||||||||||||||||||||
|
3.2.4.3 NASA |
National Aeronautics and Space Administration Strategy NASA will partner with DARPA to have at least two NASA sites be active participants in ultrahigh speed testbeds. NASA will investigate the feasibility and performance of engineering application demonstrations across these testbeds. The goal is to achieve an end-to-end high speed hybrid network capable of supporting both wireless and bounded media applications. Milestones The goal of NASA's program in collaboration with NGI is to accelerate R&D in select core technologies (transmission, fast switching, wavelength division multiplexing, network security). The following milestones illustrate NASA's participation:
|
||||||||||||||||||||||||||||||||
|
|
|
||||||||||||||||||||||||||||||||
|
|
|||||||||||||||||||||||||||||||||