Dwaraki A.,University of Massachusetts Amherst |
Seetharaman S.,Infinera |
Natarajan S.,Deutsche Telekom AG |
Wolf T.,University of Massachusetts Amherst
ANCS 2015 - 11th 2015 ACM/IEEE Symposium on Architectures for Networking and Communications Systems | Year: 2015
SDN-enabled networks offer a great degree of flexibility, programmability and support for multiple applications. Applications running on top of a SDN controller could further benefit if network state information were made available to them as part of the SDN framework. Our work investigates the requirements for abstracting network state on the control plane. We intend to show that, by meeting these pre-requisites, network state can be made available with the guarantee of state safety and integrity. We envision that our state management abstraction will provide safety in the data plane and aid better programmability in the control plane. © 2015 IEEE.
Chatelain B.,McGill University |
Laperle C.,Ciena |
Krause D.,Infinera |
Roberts K.,Ciena |
And 6 more authors.
IEEE Photonics Technology Letters | Year: 2010
An intersymbol-interference-free pulse shape is specifically designed for self-phase-modulation (SPM) mitigation. It is experimentally shown to increase the SPM tolerance of a 40-Gb/s single-channel postcompensated dual-polarization quadrature phase-shift keying system by 1.5 dB. Numerical analyses at 40 and 100 Gb/s indicate that the specialized pulse shape can increase the maximum reach for G.652 fiber and G.655 fiber up to 1190 and 820 km, respectively. © 2010 IEEE.
Filion B.,Laval University |
Jiachuan L.,Laval University |
Jiachuan L.,Post University |
Nguyen A.T.,Infinera |
And 3 more authors.
Journal of Lightwave Technology | Year: 2016
We experimentally demonstrate semiconductor optical amplifier (SOA)-based wavelength conversion of 3 × 25 GBd Nyquist-16QAM signal for a flex-grid network. The conversion efficiency (CE) and power penalty of each of three channels during single pumped SOA wavelength conversion are studied with respect to three different channel spacings (or frequency grids). The BER performance of all converted channels fall below the FEC threshold of 3.8 × 10-3, even with a 50 GHz grid. The results show the tradeoff between channel spacing, CE, and BER power penalty. Closely packed channels, which clearly increase spectral efficiency, are also shown to decrease conversion power penalty, potentially counter balancing increased crosstalk levels. These results can be used to optimize routing and spectrum allocation strategy when SOA wavelength converter(s) are present in the optical link. © 2016 IEEE.
Qouneh A.,University of Florida |
Li Z.,University of Florida |
Joshi M.,Infinera |
Zhang W.,NetApp |
And 2 more authors.
Proceedings - IEEE International Conference on Computer Design: VLSI in Computers and Processors | Year: 2012
With silicon optical technology moving towards maturity, the use of photonic network-on-chip (NoCs) for global chip communication is emerging as a promising solution to communication requirements of future many core processors. It is expected that photonic NoCs will play an important role in alleviating current power, latency, and bandwidth constraints. However, photonic NoCs are sensitive to ambient temperature variations because their basic constituents, ring resonators, are themselves sensitive to those variations. Since ring resonators are basic building blocks for photonic modulators, switches, multiplexers, and demultiplexers, variations of on-chip temperature pose serious challenges to the proper operation of photonic NoCs. Proposed methods that mitigate the effects of temperature at device level are either difficult to use in CMOS processes or not suitable for large scale implementation. In this paper, we propose Aurora, a thermally resilient photonic NoC architecture design that supports reliable and low bit error rate (BER) on-chip communications in the presence of large temperature variations. Our proposed architecture leverages solutions at both device and architecture layers that synergistically provide significant improvements. To compensate for small temperature variations, our design varies the bias current through ring resonators. For larger temperature variations, we propose architecture-level techniques to re-route messages away from hot regions, and through cooler regions, to their destinations, thereby lowering BER. Our simulation results show that Aurora provides a robust architectural solution to handle temperature variation effects on future photonic NoCs. For instance, average BER and message error rate (MER) are reduced by 78% and 30% respectively when the combined device and architectural technique (SPF) is applied. From the perspective of power efficiency, Aurora is also superior to conventional photonic NoC architectures by as much as 33%. © 2012 IEEE.
The SCinet network, SC’s Supercomputing Internet, is now live. On November 14, 2015, the Austin Convention Center became home to the fastest and most innovative computer network in the world, delivering more than 1.6 terabits per second of network bandwidth to the SC conference (SC15). SCinet gives the SC conference attendees a unique chance to showcase and discover the latest research in HPC. By building the fastest, most innovative operational network possible every year, SCinet enables data-intensive research and live-use of high performing hardware to run multi-gigabit demonstrations, requiring a fast and robust infrastructure. “This network is unrivaled with regards to its capabilities and the broad-reaching influence, both nationally and internationally, to support demonstrations and experiments that could not be done easily in any other place. It’s a one-of-a-kind environment where research meets production,” says Davey Wheeler, SCinet Chair and Senior Network Engineer from the National Center for Supercomputing Applications (NCSA) at the University of Illinois at Urbana-Champaign (UIUC). “As the SCinet Chair leading the development of this network, this year is a culmination of 17 years of experience working on SCinet from year to year. It is humbling and honoring to be able to work aside these colleagues and see the tremendous talent, dedication and creativity of the volunteers.” SCinet is built by a team of expert volunteers from around the world, taking one year to design the network, three weeks to set it up, four days to operate it, and 24 hours to tear it down. Over 100 engineers from industry, academia and government institutions came together to build this network, using over $22 million in loaned equipment and over 89 miles of newly installed fiber optic cables. “Having been the SCinet Chair for SC07 in Reno, I am intimately familiar with the incredible amount of planning and work that goes into creating what will be the most powerful network. Over 130 SCinet volunteers from more than 15 countries have worked energetically for the past year to provide wired and wireless access to our conference attendees, and the platform for our exhibitors to showcase bandwidth-driven HPC and cloud computing applications. SCinet continues to be a crucial part of SC, and I am extremely grateful for their hard work,” says Jackie Kern, Director of IT Shared Services at UIUC and SC15 Conference Chair. For SC15, SCinet has connected multiple 100 gigabit circuits, bringing an unprecedented 1.62 terabits per second of bandwidth to the Austin Convention Center. Lonestar Education and Research Network (LEARN) leads this effort in collaboration with leading national and international research networks and commodity providers. LEARN and SCinet supports the HPC community by providing multiple 100 gigabit waves and complementary capabilities throughout the SC15 conference events. In addition to the massive external capacity SCinet brings to the convention center, the network is also supporting research initiatives through a half-day workshop, Innovating the Network for Data-Intensive Science (INDIS); and the Network Research Exhibition (NRE). SCinet organizes the INDIS workshop to discuss technical papers and show floor demonstrations dedicated to high performance networking technologies, innovations, protocols, hardware and much more. Further, SCinet is providing the wireless connectivity for more than 11,000 expected conference attendees throughout the conference areas. The SCinet team built the SC15 wireless network using 339 wireless access points to support more than 4,000 simultaneous users on the conference wifi. The wireless network will include support for eduroam (education roaming) service, which allows users (researchers, teachers, students, and staff) from participating institutions to securely access the protected wireless network using their home organization’s login credentials. SCinet is the result of the hard work and significant contributions of many government, research, education and corporate collaborators who have volunteered time, equipment and expertise to ensure SC15’s success. This year, SCinet continued the Contributors Program and would like to give a special thank you to all SCinet contributors and volunteers. Volunteers from the following organizations supporting the development and deployment of SCinet: Alcatel-Lucent, Army Research Laboratory, CABLExpress Corporation, CENIC, CenturyLink, Ciena, Cisco, Clemson University, DFN-Verein, DataDirect Networks, Dell Research, Energy Sciences Network (ESnet), Florida LambdaRail (FLR), Freelance, Georgia Institute of Technology, Idaho State University, InMon Corporation, Indiana University, Indiana University GlobalNOC, Indiana University of Pennsylvania, Infinera, Internet Consulting of Vermont, Internet2, JDSU, Lawrence Livermore National Laboratory, Lonestar Education and Research Network (LEARN), Louisiana Optical Network Initiative (LONI), Louisiana State University, the National Center for Supercomputing Applications, National Oceanic and Atmospheric Administration (NOAA), Pennsylvania State University, Purdue University, REANNZ, Radware, Reservoir Labs, SURFnet, Sandia National Laboratories, The University of Texas at Austin, University Corporation for Atmospheric Research, University of Amsterdam, University of California, San Diego, University of Colorado Boulder, University of Heidelberg, University of Illinois at Urbana-Champaign, University of Michigan, University of Oklahoma, University of Pittsburgh, University of Tennessee, Knoxville, University of Texas, University of Wisconsin-Madison, Utah Education Network, and Virginia Polytechnic Institute and State University.