This project will develop a plan to leverage the Nation Transparent Optical Network (NTON), jointly funded by DARPA and the private sector, to provide a 20 Gb/s optical research network platform spanning the West Coast of the US, connecting premier university, government and private sector research labs in Washington, Oregon and California. The output of this effort will form the basis for a proposal to NSF to fund the construction and operation of the NTON Research Network Platform. Availability of multi-gigabit/second channels running over commercially available fiber for thousands of km opens new research challenges and opportunities in the construction and operation of complex networks. The precursor work at the component and subsystem level is completed and pioneer applications have been demonstrated. Additionally, the NTON network was shown at SC97 to be accessible from the national research networks. A large scale testbed network is now needed to advance these technologies for the competitive benefit of our educational, industrial and commercial sectors in the global economy. The test bed will enable the evaluation of new network architectures and advanced control and management strategies and to support the development of new infrastructures to enable distributed applications to operate in ways not practical with existing networks. Funding constraints demand maximum leveraging of all available public and private resources. Utilizing the unique, NTON provides this opportunity. The present NTON is the first multi- wavelength, switched, optical network to use the installed commercial fiber base. Initially deployed around San Francisco Bay, the NTON is being extended north to Seattle and south to San Diego. Four 2.5Gb/s channels for SONET traffic and 10 Gb/s of unformatted transport will be available for research purposes. The SF Bay ring will be upgraded to 8 format free wavelengths - each capable of 2.5 Gb/s transport. This NSF planning grant will provide funding to determine the appropriate family of NSF network research sites to be provided access to the test bed, to articulate the types of functional, performance and measurement needs of NSF researchers and to specify the NTON extensions needed to meet these requirements. The planning will entail three specific activities: Activity 1) Identify and validate the potential NSF user communities and their research network needs in terms of functional and performance requirements to provide a definition of what users can expect from the network; Activity 2) Develop preliminary specifications for the additional network architecture, including costing and connection details; Activity 3) Articulate network resource allocation mechanisms to enable a distributed staff of professionals, faculty and students to implement NSF-approved resource allocation policies. Potential User Communities. The potential user communities will be drawn from researchers at NSF-funded colleges and universities. This community of users falls into three broad groupings: Network researchers who are interested in developing new network technologies, algorithms, protocols and management strategies; Applications and infrastructure developers who share the vision, understanding and patience necessary to define, prototype and implement new network characteristics; Applications oriented scientists who are simply interested in faster, more economical and more reliable networks. Examples include applications scientists, chemists and atmospheric scientists. A major component of this planning effort will be to ensure that these groups communicate openly and clearly throughout the requirements and design process. Two workshops are contemplated to encourage this communication. Workshops Planned. Application Requirements. In Activity One, the goal will be to work with the above-cited groups to develop a balanced view of the services and characteristics of the NTON Research Network Platform. Key NSF staff and members of the NSF research communities will collaborate with other network researchers to develop 'strawman' designs for the Research Platform. These will be reviewed, modified and extended at a planned workshop to solicit opinions, identify needs and establish priorities. The result of this activity will be a prioritized set of requirements that can be used to guide further development of the NTON Research Network Platform. Network Architecture. In Activity Two, the goal will be to begin with the existing National Transparent Optical Network (NTON) architectural model and extend it to include the requirements developed in Activity 1. This will be a high level approximation to what will be the final detailed architectural specification of the NTON Network Research Platform. This specification will be reviewed at the second workshop. Platform Management. Concurrent to Activity 2, Activity 3 will be directed toward establishing network resource allocation policies to ensure that the network can be efficiently and effectively implemented by the network management staff in collaboration with the user communities. A set of recommendations will result, detailing the roles and responsibilities of all the participants and establishing the rules of governance for the NTON Research Network Platform.
