|
|
A key element of any effort to redesign the Internet is a strategy
for fostering the research cycle: drastically lowering the barriers
that promising new directions developed by the research community face
before transition to industrial development and deployment within the
commercial Internet. This requires that we move well beyond the
methodologies and facilities used today. An experimental facility that
enables the research community to address architectural questions must
provide a seamless, end-to-end research process for taking ideas from
conception, through validation, to deployment, similar to the
idealized process shown below:
Unfortunately, it is well known within the networking research
community that we lack effective methodologies and tools for
rigorously evaluating, testing, and deploying new ideas. As depicted
below:
Today researchers are able to simulate and experiment with
small-scale prototypes, but are unable to perform experiments at a
meaningful scale. This is not surprising considering the
barrier-to-entry for experimenting with a new network architecture or
service. A research group would have to arrange for dedicated hardware
spread over a wide geographic area, acquire its own networking
bandwidth, provide its own management and operational support, be
responsible for its own security and fault isolation, implement its
own measurement instrumentation, and so on. The consequence of this
chasm is that standards bodies and early commercial adopters look with
a skeptical eye towards any new networking idea backed solely by
simulation results or small-scale experimentation. GENI's main goal is
to radically improve the process by which research goes from the idea
stage through validation to deployment. GENI is essential to the
process of discovering the Future Internet.
|
