Learn more about how Ohio researchers are successfully applying high-performance computing to their research.
With the Internet becoming almost ubiquitous, many end-user applications that include office applications (e.g. desktop videoconferencing, data backup for disaster recovery) and day-to-day consumer applications (e.g. online gaming, video-on-demand) have become network-dependent. These end-applications use network protocols that are complex and resource intensive.
This requires the best-effort structure of today’s Internet to support massive flows of voice, video and data traffic while still maintaining consistent end-application performance. End-application performance over the Internet is directly impacted by the end-to-end performance bottlenecks present at the end-hosts and at the intermediate network paths. The bottleneck factors include cross-traffic congestion, cyber-attacks and optical link failures. Understanding the limitations and demands of advanced end-applications and developing suitable adaptation technologies is vital for supporting existing and emerging end-applications on the Internet.
The Networking Research group at OSC/OARnet is engaged in evaluating and developing novel and innovative, network-based end-applications. Also, the group develops techniques and open-source tools for network-awareness required for end-users to identify and isolate bottleneck scenarios. The end-applications targeted are: (a) Voice and Video over IP (VVoIP) for multi-point videoconferencing and video streaming (IPTV), (b) Remote Instrumentation for remote access of expensive scientific instrument resources, (c) Thin Client Cloud Computing for virtualization of user desktops along with their applications and data, and (d) Large-scale Data Transfers for grid computing and disaster data recovery. In addition, there are projects underway in the network security arena to enable Secure Videoconferencing, Detection of Active Worm Propagation, and Network Forensics Training. The group has setup several Network Monitoring Test Beds to collect human perceptual measures, throughputs for large-scale data transfers, and both active/passive network measurements.
Research findings from the above research projects are being implemented as human-aware and network-aware end-applications such as “Remote Instrumentation Collaboration Environment (RICE)” and as open source network measurement tools such as “H.323 Beacon” and “ActiveMon.” These open source software packages are being widely used at institutions across Ohio, as well as in other academic and commercial communities worldwide.
- OnTimeMeasure: Measurement Service for NSF GENI
- Multi-Domain Network Performance Sampling
- Thin Client Cloud Computing
- Remote Instrumentation and Collaboration Environment (RICE)
- Secure Videoconferencing
- Voice and Video over IP (VVoIP) Traffic Studies
- Network Monitoring Test Beds
- H.323 Beacon Project
- ActiveMon Project
- Network Forensics Training
- Large-Scale Data Transfers over the Internet
- OSCnet Measurement Project
- Detection of Active Worm Propagation
- Remote Backup of Critical Administrative Data for Disaster Recovery
- NSF Network Connections Project
Internet2 American Distance Education Consortium The Ohio State University University of Cincinnati Cincinnati State Community College Central State University Wright State University University of Toledo Southern State Community College NC-ITEC Apparent Networks, Canada Polycom VMware IBM VoIP Laboratory, UFRJ, Brazil HEAnet, Ireland Samara Institute of Technology, Russia King Fahd University of Petroleum and Minerals, Saudi Arabia Huawei Technologies, China
Contact the Networking Research group.
David E. Hudak, Ph.D.
Director of Supercomputer Services