NSF Awards $2.3 Million Grant for High-Performance Wireless Research and Education Network

The National Science Foundation (NSF) has awarded a $2.3 million, three-year research grant to UC San Diego to create, demonstrate, and evaluate a non-commercial, prototype, high-performance, wide-area, wireless network for research and education. Principals on the proposal are with the National Laboratory for Applied Network Research (NLANR) at SDSC, and the Scripps Institution of Oceanography (SIO).

The project involves a multi-institution collaboration led by Hans-Werner Braun of the NLANR group at SDSC and co-principal investigator Frank Vernon of the Institute for Geophysics and Planetary Physics (IGPP) at SIO, and includes researchers from other institutions such as Paul Etzel, Director of Mount Laguna Observatory and Chair of the San Diego State University (SDSU) Astronomy Department, and researchers from UCSD's Center for Wireless Communications (CWC).

The NSF research grant includes $1.4 million in the first year to get the network up and running, and the remainder over the next two years of the three-year grant for evaluation, upgrades, and adding additional applications from a variety of different scientific disciplines, and support for educational uses.

"This demonstration project promises to fill an important gap in many fields of scientific research that increasingly need high-speed network access from hard-to-reach sites, and it's an important plus that it also can serve education needs in remote, underserved communities," said Aubrey Bush, director of the NSF Advanced Networking Infrastructure and Research (ANIR) division.

The project will create a wireless backbone network in southern California that will initially include backbone nodes on the UC San Diego campus and a number of mountaintops in San Diego County including Mt. Woodson, Mt. Laguna, Mt. Palomar, and Toro Peak (Figure 1). To increase network robustness, redundant network links will be run between backbone nodes. Researchers in various disciplines and educational communities will be able to gain Internet connection through this backbone network.

"The problem we are addressing is that you do not have to go very far from the urban core before you lose access to high-speed Internet connectivity, or sometimes any Internet connection at all," Braun says. And there are researchers in many disciplines who work or live in such areas, as well as many schools now left out of the Information Age because they are in areas where there is no business incentive for commercial service providers. Existing alternatives to providing remote areas with high-speed Internet access, such as fiber optics and satellite links, are prohibitively expensive, so Braun and his collaborators are investigating a cost-effective alternative to providing network access for both scientific research and education.

"This project is interesting to us because not only will we be doing specific performance-related network research in a wide-area wireless networking environment that enables broadband 'last mile access,' but at the same time, we will be providing real connectivity services for the day-to-day activities of scientific researchers in disciplines across astronomy, earthquake monitoring, and ecology, as well as to remote, underserved educational users," Braun said. Getting researchers with real applications involved from the start gives valuable information about the actual needs and problems of building this kind of network, and speeds up the learning process for everyone. For this project, the NLANR Measurement and Analysis Group at SDSC will build upon its extensive experience in both research and engineering aspects of high-performance research networks.

Project researchers expect to make rapid progress in implementing the first phase of the wireless network, which includes building the critical backbone links and nodes, as well as the connections to existing collaborators. "Once we have the backbone running, this will make it possible to connect other collaborators with a wide range of research needs at relatively low cost, similar to how NSFNET worked," Braun said. "And because we are using unlicensed spectrum, this eliminates red tape, and researchers will be able to do a field installation without a lot of professional help, sometimes even on their own."

Two initially quite different scientific applications that the network will support are those of geophysicists of IGPP-SIO, whose earthquake sensors are arrayed along the earthquake faults that cut through the deserts and mountains of southern California, and astronomers at SDSU's observatory on Mt. Laguna. "The network requirements for earthquake monitoring are quite different from those of many other researchers," Vernon says. Earthquake sensors are often deployed in remote, rural areas that usually lack both phone service and electric power. "Our sensors need to be very self-contained and run on solar power, and generate continuous data, but only at relatively low speed per sensor, so we don't need large bandwidth. What is critical for an earthquake network is robustness, especially during an earthquake, and ubiquity—that the network is available everywhere we need to deploy earthquake sensors," Vernon says.

A second primary application is to connect SDSU's observatory at Mt. Laguna. "Although we're only 35 line-of-sight miles from our home campus at SDSU, the observatory is still too remote for us to be served by anything but a single, dedicated phone-line that can only transmit data at 30 kilobits per second," Etzel says. Yet modern CCD-equipped telescopes generate images of more than eight megabytes each, hundreds per night. In addition to needing to transmit these sky images to colleagues or to the home campus for analysis, real-time connections are needed for coordinated observations with other observatories, including satellites, requiring high-performance traffic both to and from the observatory.

Moreover, observing by remote control over the Internet is now routinely raising efficiency and opening significant educational opportunities as astronomy becomes a global enterprise. "By providing high-speed Internet access with this wireless network, we'll not only enable much more efficient operation of our active astronomy program, we'll also provide a real-world testbed for this prototype network," Etzel says.

A third application for this type of wireless network is in environmental observing systems. Such systems—seamless linkages of environmental sensors, computers, and databases—will provide researchers as well as policymakers, resource managers, educators, and the public with valuable information about the present and future state of the land and oceans for purposes from detecting and forecasting climate variability to managing natural resources for sustainable use and mitigating natural hazards such as storms, fires, and earthquakes.

SDSC has created a special "Observing Systems Team" (OST) to collaborate with researchers at SIO and other UC San Diego research units to focus on delivering real-time data for environmental monitoring and response. The development of a prototype environmental observing system, as well as other national observing system initiatives, hinges on the successful deployment of wireless technologies such as the network being developed by NLANR researchers

A quite different part of the wireless project is delivering Internet access to remote, underserved communities, and the project is planning to provide wireless links to hard-to-reach schools. "Not only are these schools without high-speed Internet access, they are often without any form of Internet access at all," Braun says. The schools will be provided access through links to backbone nodes, and project researchers will explore the possibility of extending this service to other schools with the same urgent need.

—PT