GPO visits RENCI Print
Written by Ilya Baldin   
Monday, 15 June 2009 20:31

Several representatives from the GENI Project Office will be visiting RENCI on July 7, 2009. As part of the visit Harry Mussman, GENI Senior Systems Engineer will give a talk to the Triangle networking research community titled "GENI: Overview & Plans".


Is there a science for understanding the complexity of our networks so that we can engineer them to have
predictable (or adaptable) behavior?  In response to this fundamental research challenge, the National
Science Foundation (NSF) Directorate for Computer and Information Science and Engineering (CISE),
and the research community, are developing a comprehensive strategy to advance research and
education in Network Science and Engineering.
In parallel, community planning for the suite of infrastructure that will support NetSE experiments has
been underway since 2005. This suite is termed the Global Environment for Network Innovations (GENI).
Although its specific requirements will evolve in response to the evolving NetSE research agenda, the
GENI infrastructure suite is being planned to allow researchers from diverse disciplines across computer
and information science and engineering, as well as from economics and the social sciences, to escape
today's Internet-circumscribed research environment.
The suiteís conceptual design is now clear, and includes the following core concepts:
*    Slice-based Experimentation ñ GENI experiments will run on a ìsliceî, an interconnected set of
reserved resources, or ìsliversî, on heterogeneous substrate platforms and networks. 
Researchers will remotely discover, reserve, configure, and program, debug, operate, manage,
and teardown resources on these platforms and networks to setup, utilize and then teardown the
slivers necessary to complete an experiment.
*    Virtualization ñ Whenever feasible, the substrate platforms and networks will be ìvirtualizedî to
allow multiple researchers to simultaneously share the infrastructure, and operate without
disturbing another experiment, or being disturbed.
*    Programmability ñ When a platform includes a host or other node, researchers will typically be
able to download software into the host or node to control the behavior of the resultant sliver.  In
the case of a network node, this will specialize the function of a node, i.e., provide a custom
routing function or even a software-defined radio transceiver.
*    Federation ñ Different parts of the GENI suite are owned and/or operated by different
organizations, and the NSF portion of the GENI suite forms only a part of the overall

As envisioned in these community plans, GENI will support a wide range of experimental protocols, and
data dissemination techniques running over facilities such as fiber optics with next-generation optical
switches, novel high-speed routers, city-wide experimental urban radio networks, high-end computational
clusters, and sensor grids.  GENI is envisioned to include extensive instrumentation that makes it easy to
collect, analyze, and share measurements. 

The GENI Project Office (GPO) was formed at BBN Technologies in 2007 and has been charged with
identifying the infrastructure suite that could best support NetSE experiments.  The GPO has begun a
series of community-based planning and prototyping activities, notably Design and Prototyping
Solicitations #1 and #2.  The first set of selected prototypes will complete ìSpiral 1î at the end of one year, and
will attempt a first integration of a majority of necessary ingredients for a ìreal GENIî. 

Note that there is no pre-ordained outcome for these activities:  the resultant GENI infrastructure suite could be the existing
Internet, existing testbeds, federations of testbeds, something brand new (from small to large),
federation of all of the above, and perhaps a federation with related international efforts.
In this talk, we will present an overview of the GENI development effort, an introduction to the GENI
architecture, and a discussion of how interested researchers can get involved in shaping the facility.

Harry Mussman is a Senior Systems Engineer, one in a group of GENI Project Office systems engineers at BBN Technologies, responsible for supporting the GENI community through documentation and coordination.  Harry brought a strong background in network architecture and systems engineering to the GPO in early 2008.  Before joining BBN, he was Chief Voice-over-IP Architect at a startup, BridgePort Networks.   During the period from 1999 through 2002, he was responsible for Voice-over-IP network architecture at GTE Internetworking/Genuity.  Harry received his BSEE degree from the University of Michigan, his MSEE from Northwestern University and a PhD from Stanford University.

Last Updated ( Wednesday, 17 June 2009 15:28 )