APPLICATIONS

Team will use new Breakable Experimental Network to evaluate concepts and capabilities for GENI: A research team involving the Renaissance Computing Institute (RENCI), Duke University and Infinera is one of 29 teams across the nation that will participate in the first phase of the National Science Foundation’s Global Environments for Network Innovation (GENI) project. GENI is a multi-year effort to construct infrastructure for a large-scale experimental network that will enable the worldwide research community to test ideas and clean-slate designs in a range of technology areas including network design, distributed systems, and cyber-security. GENI will be built using a “spiral development” approach. Spiral 1 awards were announced Sept. 29 by BBN Technologies, the advanced technology solutions firm that is managing the project for the NSF. The awards, totaling $12 million, fund research to build, integrate, and begin to operate the first prototypes of the GENI suite of network research infrastructure. The RENCI-Duke-Infinera project will receive $640,000 over three years. GENI Spiral 1 focuses on ways to discover, schedule and control resources for large-scale research experiments and to measure capabilities. Multiple competing approaches are being funded to provide design insights for the evolving suite of experimental infrastructure. The RENCI-Duke-Infinera winning proposal leverages the strengths of each organization. RENCI and Duke will use ORCA (Open Resource Control Architecture)—a software framework developed at Duke—to implement a model for the GENI control plane and deploy it on the newly implemented Breakable Experimental Network (BEN) in order to create a ‘GENI island’ – a miniature version of the future GENI testbed. BEN, a RENCI-managed, regional experimental network testbed designed to push the limits of networking research, links RENCI’s home office in Chapel Hill to sites at UNC Chapel Hill, Duke and NC State University. Infinera will provide engineering support, integrating the Infinera DTN platform deployed on BEN to enable flexible provisioning of bandwidth. Based on large-scale photonic integration, the Infinera DTN delivers 100 Gigabits/second of optical capacity on a pair of photonic integrated circuits, enabling optical networks with unprecedented capacity, flexibility, speed, and intelligence. Infinera’s Bandwidth Virtualization feature enables a high level of flexibility and programmability in an optical platform for the wide-ranging research agenda planned for the GENI project and BEN. “Infinera is delighted to support RENCI in exploring new architectures and experiments towards the development of the next-generation Internet,” said Infinera Chief Technology Officer Drew Perkins. “We are very excited that Infinera’s Bandwidth Virtualization capability is a near-perfect match with RENCI’s vision of a highly programmable optical layer.” “BEN will serve as a miniature GENI prototype allowing us to experiment with different arrangements of network resources with special attention being paid to the provisioning of the optical substrate.” said Ilia Baldine, a senior network researcher at RENCI and principal investigator for the project. “Using BEN and with ORCA as a foundation, our plan is to deploy, demonstrate and evaluate the fundamental concepts and capabilities of GENI, such as end-to-end slicing in an optical network that connects heterogeneous computing and storage resources.” Successive spirals of GENI funding will refine and extend the GENI suite in response to the research community’s evolving needs and interests. “The GENI engagement is a key part of a regional initiative to investigate new structures for ‘virtualized’ server networks involving other RENCI partners,” said Jeff Chase, a professor of computer science at Duke and co-principal investigator on the project. “The next generation of researchers and educators will be able to tap into a programmable ‘cloud’ of servers, networks, and storage for research in networking, distributed computing, and other areas.” For more on GENI, visit www.geni.net . For more on BEN, visit ben.renci.org .