Hank Hogan, email@example.com
FALMOUTH, Mass. – Future oceanographic research could be done
more cheaply and efficiently, thanks to a communication innovation powered by the
latest LEDs and detectors. Researchers from Woods Hole Oceanographic Institution
used these in a demonstration of remote optical communication under water, which
they described at the 2010 Ocean Sciences Meeting in Portland, Ore. Unlike acoustic
communication techniques that have been the standard, the new approach is fast enough
“We’ve done that, certainly. It’s compressed
video,” said team leader Norman E. Farr. “We can transmit anywhere from,
say, 500 kilobits to 15 megabits per second. That’s all based on range, rate,
power and water clarity.”
Underwater data communication by
light is shown in this artist’s conception of how the optical modem could
function at a deep-ocean-cabled observatory. Autonomous underwater vehicles (in
yellow) collect sonar images (downward bands of light) and other data at a hydrothermal
vent site and transmit this through an optical modem to receivers stationed on moorings
in the ocean. The moorings connect to a cabled observatory, and the data is sent
back to shore. Scientists, in turn, can send new instructions to the vehicles via
the optical modem. Illustration by E. Paul Oberlander, Woods Hole Oceanographic
A senior engineer at Woods Hole, Farr explained that the limit
of the optical communication link would be 200 m in clear water. In particulate-laden
water, the distance would be less. In either case, the rate would be at the low
end of the speed at the extremes in range. At shorter distances between transmitter
and receiver, the data rates would be higher.
Typically, underwater communication is done using sound, not light.
The acoustic approach offers the advantage of working over great distances. That
plus is balanced by slow data rates and transmission, the latter a result of the
slow speed of sound in water and the former, of sound’s wavelength.
These constraints mean that oceanographic research is done with
tethered remote-operated vehicles. The cabling provides video and data connections,
but the cost is a more expensive remote vehicle and support ship. Better and faster
communication would allow the cable to be cut.
For an optical data link, the Woods Hole team needed a light source
in the visible, from about 400 to 500 nm. That’s the window for transmission
through water. The source had to be low-powered because it would have to run in
battery-powered systems. And because of size constraints, it also had to be compact.
Finally, it had to be able to be modulated rapidly, as that would determine the
maximum data rates possible.
In developing a solution, the researchers used recently available
high-power LEDs. They put these into a hemispherical transmitter, choosing this
approach over the alternative of a tight beam because it offered better total performance.
For a receiver, they used photomultiplier tubes, selecting them
over avalanche photodiodes because of their sensitivity. They placed bandpass filters
before the receiver to suppress everything not in the signal transmission window.
As for the rest of the optics, they built these so that the receiver, as with the
transmitter, would operate over a hemisphere.
For remotely operated vehicles, communication by light and sound
ensures that vessels don’t lose contact, Farr said. “As soon as you’re
outside of the optical hemisphere, you still have acoustic communications, so you
can get the vehicle back into high-bandwidth range.”
Tests have indicated that the system allows video, an important
point, as many of the applications assume a human will be in the loop for control.
For example, a pilot might steer a remote vehicle to a location near a hydrothermal
vent with the intention of collecting samples and making measurements. To do that,
the pilot must be able to see what is going on in real time. In practice, this might
be done by having a surface ship on station above the vent, with a line containing
an optical transmitter/receiver lowered to about 100 m above the point of interest.
This July, a Woods Hole team will conduct the first large-scale
deployment of the system at the Juan de Fuca ridge off the northwestern US. There
it will be used to collect and transmit data from wellheads situated near the undersea
ridge. Expectations are that the system will last for several years, but part of
the reason for the deployment is to test that.
As for the future, Farr noted that longer distances may be possible
with brighter LEDs, more sensitive receivers or both. Data rates could go up as
well, with higher LED modulation rates.
However, there are limits, he said. “You get to the point
where you start to get dispersion because of multipath in the water. But we’re
nowhere near that.”