Signal Processing - July 2016 - 13

"We have shown that our improved
CAA design can increase the capacity
compared to conventional arrays by a
factor of 20-30% at small antenna
spacings," Bahrami says. "The biggest challenge now is to go even
beyond this."

Approximately two-thirds of the
Earth's surface is covered by water, and
it is inevitable that the number of
underwater sensors will grow significantly over the next few years. Yet data
retrieval from submerged sensors continues to be hindered by the limited
speed of existing underwater communication networks, which remain steadfastly stuck somewhere in the range of
late 20th-century dial-up modems. This
stubborn obstacle hampers a wide
range of activities, including searchand-rescue operations, pollution monitoring, tsunami detection, and other
types of important work conducted in
bodies of waters such as oceans, bays,
rivers, and lakes.
State University of New York at
Buffalo (SUNY-Buffalo) researchers
are now working to help underwater
communications at least partially catch
up with over-the-air data transmission
rates. "The remarkable innovation and
growth we have witnessed in landbased wireless communications has
not yet occurred in underwater sensing
networks, but we are starting to change
that," says Dimitris Pados, a professor
of electrical engineering in the SUNYBuffalo's School of Engineering and
Applied Sciences.
Pados and several coresearchers are
developing new types of hardware and
software, including waterproof modems
and open-architecture protocols,
designed to address underwater transmission speed issues. The teams'
efforts are currently focused on combining a software-defined radio (SDR)
with underwater acoustic modems.
Their prototype underwater communication system is a cognitively self-optimized technology that works by jointly
adapting link signal waveforms and
network session routes to maximize

state University of new york at BUffalo

Underwater wireless

Figure 3. SUNY-Buffalo students test an enhanced underwater communications technology on
Lake Erie.

In May 2015 at Lake LaSalle on
network throughput and/or spectral
the SUNY-Buffalo campus, the reefficiency under a wide range of opersearch team demonstrated its technoloating conditions.
gy for the first time, achieving
"Link waveforms are cognitively
communication at 200,000 bits/s over
optimized over the whole spectrum
an underwater distance of 200 m. The
accessible by the transceiver nodes (allsystem consisted of two Ettus Respectrum channelization) and routes
search USRP N210
are cognitively optiSDN modules using
mized over all accesstate university of New
Teledyne RESON
sible nodes," Pados
York at Buffalo researchers TC4013 transducers
says. He notes that
are now working
with a 1 Hz-170 kHz
the technology is
operational frequenbased on well-underto help underwater
cy range. Further
stood signal processcommunications at
tests were conducted
i n g t h e o r y. "A n
least partially catch up
on nearby Lake Erie
elementary pulse sigwith over-the-air data
(Figure 3).
nal is selected and
transmission rates.
Pados says he
shaped to occupy all
expects to see many
hardware accessible
significant applications for the techfrequency bandwidth-raised cosine
nology emerge over the next few years
pulse, chirps, and others," Pados
as it matures and achieves increasingly
explains. A finite number of different
faster throughput speeds. Real-time
sign/phase shifted repeats of this pulse
underwater sensing (pollution, temperforms the final waveform design,
ature, or sea currents), sea-life moniwhich will then carry the information
toring, port surveillance, wireless
symbols on its back. "The specific
diver-to-diver communication, wireva l u e s o f t h e sign/phase shift
less diver/underwater vehicle commusequence-that we will call waveformnication, untethered sea exploration,
carrier code-are determined near-real
search-and-rescue operations, undertime by principal- component signal
water wireless video feeds, and offanalysis techniques that calculate the
shore drilling monitoring are just some
code with highest interference avoidof the technology's potential applicaance properties at any given time and
sp
tions, he notes.
receive node location," Pados says.

IEEE Signal Processing Magazine

July 2016

Table of Contents for the Digital Edition of Signal Processing - July 2016