NSF-CRI'07 Breakout Session
June 4, 2007
The New Internet -- Goals,
Testing, and Infrastructural Needs
Abstract
A new Internet must offer
substantially larger user value than existing technology, as there is
otherwise no economic rationale for justifying the multi-billion dollar
transition costs. These improvements are more likely to come from
reduced (end-user) operational costs, not improved packet formats,
packet switches, congestion control or routing algorithms. It is unclear
how even large-scale testbeds can address these questions.
Given finite financial
resources, we need to choose where large new testbeds are most useful
and whether testbeds should take the form of physical infrastructure or
"kits" that can be physically replicated. Other models, such as resource
cooperatives, may also offer a more bottom-up model for resource sharing
where department-scale testbeds are not appropriate.
This breakout session will consider the above
questions and issues, and will discuss the possible role that the CRI
program at NSF may play in meeting the needs of the research community.
Session Leader and Moderator:
Henning Schulzrinne, Columbia University |