IU research team receives grant to develop GENI experimental network tools
Beth Plale
Researchers are working on new models of computer networks that may someday replace the Internet, and researchers at Indiana University have been awarded a grant to develop tools to ensure that detailed network conditions can be measured for research.
As computer network experiments increase in complexity and size, it has become increasingly difficult to fully understand the circumstances under which a network experiment was run, particularly when it comes time to reproduce the results. A collaborative team from Indiana University will lead the effort to provide essential tools related to the history and authenticity of an experiment's data set (called "provenance") for the Global Environment for Network Innovations (GENI) computer network.
The GENI provenance effort, which is supported by a $484,485 grant from the National Science Foundation (NSF), will be led by principal investigator Beth Plale, director of the Data to Insight Center in the Pervasive Technology Institute at Indiana University. Plale, a professor at the IU Bloomington School of Informatics and Computing, partnered with co-PI Chris Small, a network engineer and research scientist at the IU Global Research Network Operations Center (Global NOC). The collaboration leverages Plale's expertise in data provenance with Small's deep knowledge of advanced computer networking.
Sponsored by the NSF, GENI supports the development of a national-scale suite of infrastructure for experimental research in groundbreaking network science and engineering. Provenance information is valuable to GENI scientists, helping them to accurately understand, repeat and learn from experiments conducted on the network over time.
"Provenance gives an orientation on data collection that is considerably different from auditing or performance monitoring," said Plale. "Provenance can provide a better understanding of how experiments were run because it can capture relationships between events in the layers in the GENI infrastructure. We feel that provenance collection will add considerable value to GENI for new users and for those wanting to repeat experiments they or other computer scientists carried out earlier in time."
The provenance collection will be based on the Karma tool, developed with funding from the NSF Strategic Data for Cyberinfrastructure (SDCI) program, and applied to collection in an informatics application for pharmaceutical discovery. It will also be applied to a weather modeling and analysis framework. The GENI Provenance Registry, called NetKarma, will capture the activities of user experiments conducted on a slice of the GENI network by monitoring all layers of the complex network of computers.
"This research grant in network science is a testament to IU's national leadership in the area of advanced networking and was made possible by IU's unique collaboration involving the IU Global NOC, PTI and the School of Informatics and Computing," said Brad Wheeler, vice president for information technology and CIO at Indiana University.
