Open Menu Close Menu

Networking

The Deepest Internet Connection on the Planet

The University of Hawaii's ALOHA Cabled Observatory supports real-time observations in the deep ocean.

ALOHA Cabled Observatory

The ALOHA Cabled Observatory transmits data from three miles deep on the ocean floor.

Do you think you have networking issues extending WiFi to remote parts of your campus? Try extending electric power and Internet service to the ocean floor three miles below the surface — the deepest Internet connection on the planet. That is the challenge the University of Hawaii had to overcome to support scientists doing sustained real-time observations in the deep ocean.

Located 60 nautical miles north of the island of Oahu, the ALOHA Cabled Observatory (ACO) is the site of the long-term Hawaii Ocean Time-series open ocean measurement program, visited by research vessels 10 to 12 times each year since October 1988. Researchers use it to study ocean currents, salinity, plant and animal life, and geology.

Campus Technology recently spoke with Brian Chee, director of the University of Hawaii's Advanced Network Computing Laboratory, about the observatory's mission and the difficulties of getting it set up and keeping it connected.

Years ago, UH faculty members got the idea of going to the same spot on the ocean floor to do measurements. On a monthly basis, they would lower tools to the sea floor to take samples and run lab tests to assess chemicals, dissolved gasses, bacteria, detritus, viruses, etc. "It is almost like taking core samples in the arctic ice," Chee said. "This is the longest-running oceanographic time series on earth." But they were only going out to sea once a month, partly because that was all they could afford. The ship time is about $30,000 per day, he said. "Our scientists wanted to have something out there that could allow them to do samples 24 hours a day, so that when things change, they can catch the change rather than just a spot on the curve. So we thought there has to be a better way."

Thus began a decade-long journey to find a way to take measurements on the sea floor on an ongoing basis. The National Science Foundation approved the ACO proposal in 2002, but the telecom company that first planned to donate cable went through bankruptcy, scuttling that option. Eventually, AT&T donated a retired fiber-optic cable to the project. In 2007, the AT&T cable, running from Hawaii to California, was scooped off the bottom of the sea, where it had rested for almost 20 years, and brought to Station ALOHA.

"AT&T also provides an ongoing donation in that we are allowed to use space in its cable landing station in Makaha on Oahu," Chee said. "That is unusual. Normally space in a cable landing station is mega-expensive and it is considered by Homeland Security to be a strategic resource. So all personnel have to be cleared."

Yet progress was slow, and challenges still remain. The engineering team had to figure out how to place and connect switches, cameras and monitoring devices on the bottom of the ocean using a remotely operated vehicle. "Any gear we cannot run at pressure has to be in pressurized vessels," Chee explained. "These vessels are about 10 feet long, 24 inches in diameter and made of titanium. They cost $135,000 each." In 2008, a remotely operated vehicle retrieved an observatory deployed as a proof-of-concept. It turned out to have leaky connectors, allowing water to reach fragile electronic and fiber-optic communication cables.

It took three years to solve these issues and get more funding for the ACO, but on June 6, 2011, the ACO was successfully deployed on the ocean bottom. Five modules are connected together on the sea floor, with a junction box connected to the cable and to the observatory module. Together, they supply 1,200 watts of power and 100 megabits per second of Ethernet communications to sensor systems on these two modules, and to the other three modules. Sensors provide live video of the ocean bottom around the ACO and sound from local and distant sources, as well as data on currents, pressure, temperature and salinity. (The ACO Web site has an interactive diagram explaining the different equipment components.)

Scientists are able to study the movement of deep ocean currents. "And the deep ocean currents are one of the great movers and shakers of the food cycle of the deep ocean," Chee said. "If we understand these better, then we can better predict what is going to happen in El Niño years or in giant plankton blooms." He stressed that the ACO is a platform, not a specific project. "It is gathering data so we can correlate more things. The School of Ocean and Earth Science and Technology at UH combines ocean, earth, air and space and has cross-pollination between them."

The undersea cable is key to those projects' connectivity needs, noted Chee. "A really expensive satellite connection might give us 128 kbps, or the really modern new satellites capable maybe 4 or 5 megabits per second," he said, "whereas the undersea cable is 100 megs, full duplex. It gives us a lot more bandwidth to play with. It gives me the chance to do things with VLANs (virtual local area networks) and section things off, so that even if a device starts behaving badly, I can still get in and shut down that port on the switch, just like a normal co-location."

The cable follows the submerged Kaena Ridge from the research site to the Makaha landing station on Oahu. From there, it feeds into monitoring and recording equipment in the recording station. The signals then travel through several firewalls and connections to UH-West Oahu, from where it is routed back to UH-Manoa. "The bulk of data recording and analysis is done at Manoa," Chee said, "and it is typically from the Manoa campus that we provide SSL/VPN connections to other researchers so they can access the raw data sets. There are hundreds of people using our data on a regular basis, and those that cite our data in their research number in the thousands."

The ACO also works with an array of vendors who provide specialized equipment to work in extreme conditions. For instance, a company called Opengear uses "out-of-band management" to provide a secure alternative connection for accessing network devices at the ocean floor, so that the networking infrastructure can still be accessed even during system or network outages when primary connections go down.

Even today, though, the ACO still does not have 100 percent of its equipment working. For instance, currently the biologists are upset because the lights are out. "They are hammering on us to get the lights back on," Chee said. "The cameras are working fine, but the lights keep going out. The unfortunate thing is if a connector is not made perfectly, you get a short, and it tends to grow and grow," he explained. "So with some of our equipment, the connectors weren't quite perfect, and they shorted out over time. That is why we want to go out and recover these and find out what shorted out, how it leaked, try to fix it and learn from that."

The good news is that after every single dive, more pieces work and Chee and his team have a better understanding of what is going on.

Chee said another reason that progress has been slow is that the NSF has not seen fit to fully fund the project. "We have had to get very creative about how to build this platform," he added. For instance, the AT&T cable uses a proprietary signaling system, so the team had to hire a retired ATT engineer to custom-design a way to convert from 100Base-FX to the proprietary system the cable uses. "That way we can still control the repeaters, power system and switchers, so that if, God forbid, someone should drag an anchor and rip up our cable a bit, and we lose a primary fiber, we can switch over to a secondary fiber," Chee explained.

The team has had to get creative about funding, too — right down to considering doing a deal with Nickelodeon to sell sea floor space and put a SpongeBob Pineapple in view of the webcam. "The scientists said two things: "First, that is cool, it would draw eyes, but are those the eyes we really want on it?" Chee recalled. "And two, do we really want to be the butt of jokes at all the conferences?"

comments powered by Disqus