|
The raw computational power of Xilinx FPGAs
in ganged arrays drives the international search
for extraterrestrial intelligence at 1015 ops per second.
You don’t have to leave Earth to find intelligent
life on other worlds. All you have to do
is tune in ... at the right time ... on the right
frequency ... in the right direction ... with
the right spectrometer ... using the most
powerful supercomputer on this planet.
With the support of the Xilinx University
Program (XUP), the University of California
at Berkeley has emerged as the world leader in
the search for “ET” [pronounced EE-tee, as in
the popular fantasy motion picture E.T.]. UC
Berkeley operates about a half-dozen different
SETI (Search for Extraterrestrial Intelligence)
projects under the umbrella of the
SERENDIP (Search for Extraterrestrial Radio
Emissions from Nearby Developed Intelligent
Populations) Program.
“When we started 25 years ago, we built a
machine that could listen to a hundred channels
at once. We thought that was amazing,”
says Dan Werthimer, Ph.D., director of the
SERENDIP SETI Program at UC Berkeley.
“That was called SERENDIP I. Then we
went to 65,000 different channels with
SERENDIP II – and then Xilinx technology
allowed us to go to four million channels with
SERENDIP III.” And in 1997, “We went to
168 million channels with SERENDIP IV.”
Later this year, SERENDIP V will go
online at the Arecibo Observatory in Puerto
Rico with the capability of simultaneously
processing data from five billion channels
(Figure 1), using several hundred Virtex™-II XC2V6000 and XC2V1000 platform
FPGAs populating dozens of racks of spectrum
analyzer boards.
With such an awesome capability to collect
massive amounts of data, the
SERENDIP scientists need far more computing
power than they could possibly have
with the high-end Sun Microsystems™
workstations at the Berkeley Space Sciences
Laboratory. That’s where you and I come
in. SETI@home volunteers comprise the
largest supercomputer on the planet.
Meanwhile, in the next few years, UC
Berkeley’s Radio Astronomy Laboratory
and the SETI Institute of Mountain
View, Calif., will build the Allen
Telescope Array, a bold innovation in
radio telescope design – and a powerful
new tool for SETI research. As with
SERENDIP, Xilinx will provide the core
technology to enable real-time digital signal
processing (DSP) at the unprecedented
speed of 1015 ops per second.
And what happens if we do find ET?
Virtually all of the world’s scientific community
of SETI researchers have endorsed
a United Nations treaty that requires the
free and open disclosure of everything discovered
and deciphered. How the rest of
humanity will react is anybody’s guess. A
lot of it depends, says Werthimer, on
whether the ET signal is accidental or
intentional.
Tuning In to ET
Radio waves (including television, radar,
cell phones, and other microwave telecommunications)
are considered the optimum
band of the electromagnetic spectrum for
interstellar communication. Radio wavelengths
are relatively free of the absorption
and noise that afflict
other areas of the spectrum.
Additionally, stars
are generally quiet in the
radio wavelengths. This
makes radio frequencies
a natural candidate for
intentional interstellar
communications – or
“leakage” of local transmissions.
Just as the “local transmissions”
of American
television shows, such as
“I Love Lucy” and “The
Honeymooners,” leaked
out into space 50 years
ago (and now have passed
thousands of star systems),
it is conceivable
that we could someday
intercept an extraterrestrial
situation comedy show.
Anatomy of Arecibo
Operated by Cornell University and the
National Science Foundation, the National
Astronomy and Ionosphere Center Arecibo
Observatory is the largest radio telescope
on this planet (Figure 2).
The spherical reflector dish measures
1,000 feet (305 meters) across and covers
20 acres. In what is considered a valid scientific
calculation, Werthimer says the dish
could theoretically hold 10 billion bowls of
cornflakes. Milk, however, would quickly
drain out the almost 40,000 perforated aluminum
panels that make up the dish.
Suspended 450 feet (137 meters) above
the dish is a 900-ton (816 metric tons)
platform that can be placed with millimeter
precision anywhere up to 20 degrees
from the vertical. A “Gregorian dome” on
the platform contains two subreflectors
(secondary and tertiary) to further focus
deep space radio emissions.
The platform also houses ultra-sensitive
radio receivers cooled with liquid helium
(to reduce electron noise) so the infinitesimally
weak signals from outer space can be
picked up amidst all the interstellar static
and radio interference generated on Earth,
orbiting satellites, and probes launched
from Earth.
Piggyback SETI
While most radio astronomers are lucky to
get a day or two a year to use the Arecibo
Observatory, “We figured out how to use the
telescope 24 hours a day all year round by
having our own feed antenna,” Werthimer
says with a certain amount of glee.
“The problem with that is that we
don’t get to point the telescope, but
that’s okay, because we don’t know
where to look anyway,” he grins.
“We call it piggyback SETI.”
Spectacular Spectrometers
Although radio telescope antennas
are visually impressive and quite
essential, they are useless without
the instruments that receive and
process the signals. The real guts of
radio telescopes are spectrometers,
or spectrum analyzers, such as
SERENDIP IV.
SERENDIP IV
The SERENDIP IV spectrometer at
Arecibo consists of 120 Xilinx FPGAs on 40
spectrum analyzer boards working in parallel
to scan 168 million narrow-band (0.6 Hz)
channels every 1.7 seconds.
Each SERENDIP IV board computes a
four million point Fast Fourier Transform
(FFT). This four million point FFT is broken
down into three smaller FFTs (128,
128, and 256 points each). Xilinx chips
comb the resulting power spectra for strong
narrow-band signals and report their findings
to the back-end computers at Berkeley
for subsequent analysis, Werthimer says.
SERENDIP V
“We don’t know what frequency ET will be
transmitting at, so the name of the game in
SETI is to search through as many frequencies
as possible,” explains Aaron Parsons, a
design engineer at the Berkeley Space
Sciences Laboratory. Parsons is designing
SERENDIP V, the next-generation spectrometer
that will be able to process five
billion channels simultaneously.
As of press time, SERENDIP V was still
on the drawing board, but it is on schedule
to be installed at Arecibo later this year.
The whole spectrometer will consist of
40 spectrum analyzer boards, each performing
a pair of 64 million point FFTs to
handle a real-time signal bandwidth of 100
MHz, Parsons says. Because Xilinx platform
FPGAs have the capability of interfacing
with double data rate DRAM
memory chips, SETI engineers will be able
to fit this 64 million point FFT onto a single
Virtex-II XC2V6000 FPGA.
Spectrum Splitting
“We did this by first cutting the spectrum
into coarse frequency bins using the characteristic
frequency response of a 4,096-channel polyphase filter bank,” Parsons
explains (Figure 3). “The output data were
then re-ordered using 256 MB of DRAM
and broken into 16,384 smaller bins using
a dual flow-through FFT we developed. It
uses one-fourth of the space of a traditional
FFT, he says with pride.
Finally, information about the best signals
will be passed to a CPU over a compact
PCI backplane using a Virtex-II
XC2V1000 running a Xilinx PCI core,
according to Parsons.
Xilinx Chief DSP Architect Chris
Dick, Ph.D., consulted on the design of
SERENDIP V. “The signal processing
requirements in the SETI program present
significant computational and I/O
challenges that can only be met using
Xilinx FPGAs,” Dick asserts. “Traditional
processor-based approaches just cannot
deliver the performance required for this
challenging application.
“The highly parallel compute fabric
and I/O capability of the FPGA, however,
are well suited to support the computational
requirements of the filter banks
and FFTs used in the polyphase transform
channelizer,” Dick says.
Key aspects of SERENDIP V were
realized using a recent generation design
flow from Xilinx called System Generator
for DSP. This visual programming development
environment is based on The
MathWorks Simulink® interactive tool
for modeling, simulating, and analyzing
dynamic, multidomain systems. It provides
a natural framework for rapidly
specifying and verifying complex signal
processing systems, according to Dick.
SETI@home Wants You
As fast and as efficient as “ganged” (parallel)
FPGAs are, the real-time data processed
by SERENDIP IV still needs much further
analysis. After the DSP algorithms in
SERENDIP IV break down the incoming
signal into 168 million channels, some of
the outgoing data is recorded onto high-density
digital linear tape – about one 35
GB tape per day.
The tapes are shipped to the Berkeley
Space Sciences Laboratory. Even with highend
Sun workstations, the amount of data
far exceeds the lab’s ability to crunch the
data. Thus, the SETI@home project was
born. The SERENDIP scientists decided
to farm out the massive computing task to
idle computers all over the Internet (distributed
grid computing), so they created
SETI@home “screen saver” software.
Calling the SETI@home data analyzer
software a screen saver is a misnomer. The
only resemblance the data analyzer has to
real screen savers is that it only works when
your computer is idle – and the graphical
display of the data analysis in action is
semi-hypnotic (Figure 4).
If you want to join the hunt for ET,
you can download the free
SETI@home data analysis software at
http://setiathome.berkeley.edu. There are
versions for Windows, Macintosh,
Unix, Linux, BeOS, OS/2, OpenVMS,
and other operating systems.
Once you’ve got the software
loaded, the SETI@home server at
Berkeley sends you a 0.34 MB “work
unit.” This is a very small chunk of data
from the Arecibo tapes. The SETI@home
data analyzer performs
anywhere from between
2.4 trillion and 3.8 trillion
floating point calculations
– including FFTs, dechirping,
and baseline
smoothing, among others.
Once the data is
processed, the SETI@home
program notifies you that it
wants to report its results
back to Berkeley and acquire
another work unit. The only
time the SETI@home data
analyzer needs to be online is when the data
is being transferred.
“When we get the data back from the
participants, we comb through the strong
signals looking for ET,” Werthimer says.
Super Computing
The SETI@home project is an awesome
display of the power of distributed grid
computing. Starting with a base of 1,500
volunteers in 1998, SETI@home has
grown to more than 4.7 million participants
in 226 countries, with 2,000 new
volunteers signing up daily, Werthimer
reports.
“The SETI@home volunteers have
formed the planet’s largest supercomputer,
averaging 60 teraflops [60 trillion floating
point operations per second] and donating
about 1,200 years of CPU time daily,” he
says. “The search for ET is truly an international
effort.”
Allen Telescope Array
While the SETI research will continue at
Arecibo, the SETI Institute and the Radio
Astronomy Laboratory at UC Berkeley
have begun to build the Allen Telescope
Array at the Hat Creek Observatory near
Mt. Lassen in Northern California.
Underwritten by a large donation from
Paul Allen, co-founder of Microsoft, the
Allen Telescope Array will eventually grow
to be a huge, expandable antenna farm
(Figure 5) of as many as 350 20-foot (6.1-meter) offset Gregorian dishes with 8-foot
(2.4-meter) secondary antennas. This
innovative telescope design will cost much
less than an equivalent single dish telescope,
according to Werthimer.
Employing the same piggyback strategy
used at Arecibo, SETI researchers will
scan for ET wherever other radio astronomy
research projects aim the array.
Although these dishes can be “stamped
out like hot tubs” very cheaply, Werthimer
says, no one has ever attempted to build a
giant telescope from so many small dishes
before. The costs of the electronics and
signal processing technology required to
manage such an array were prohibitively
high. The signal processing computation
grows as the square of the number of
antennas, he explains. But, “Thanks to
Virtex-II Pro™ and Spartan™-3 chips,
the peta-op per second signal processing
costs will comprise only a tiny fraction of
the total system costs.”
Each telescope (Figure 6) will have an
extremely wide-band dual polarization
feed, covering 0.5 GHz to 11.2 GHz. This
feed will drive a pair of wide-band, low-noise
amplifiers that output their signals
to a pair of analog optical fiber laser modulators.
All telescope fibers will be routed
to the central electronic lab where the
SETI signal processors and image processors
will be located.
To make maps of the radio sky, the
FPGA-based “imager” for the antenna
array must process real-time data at the
rate of one terabit per second (1012
bits/sec) and compute one peta-op per
second (1015 ops/sec) – that’s 20 times
faster than the SETI@home supercomputer,
Werthimer declares.
First, each telescope signal will be digitized
and broken up into 1,024 spectral
components by means of a polyphase filter
bank. The resulting data from each telescope
will then be sent to a “corner
turner” that will re-order this data by frequency
channel.
Next, each telescope “pair” will be
cross-correlated and integrated (there are
N*(N-1)/2 “pairs” of telescope signals to
correlate). Then the data will be “2D
Fourier transformed” to produce an
image. All signal processing and data
routing will be implemented using several
thousand Virtex II-Pro platform
FPGAs in Rack 2 and Spartan-3 FPGAs
in Rack 4, Werthimer says.
Turning Science Fiction into Fact
SETI scientists operate on the assumption
that there are other intelligent civilizations
“out there.” Otherwise, why look? So, the
question is not “if,” but “how many?” And
where? And when will we find them?
Xilinx University Program
Xilinx donations to SETI research date
back to 1988. Patrick Lysaght, senior director
of Xilinx Research Labs, which includes
the Xilinx University Program (XUP),
reports that within the past five years,
Xilinx has donated more than $1.5 million
in hardware, software, and technical support
to UC Berkeley’s SETI research.
“The search for extraterrestrial intelligence
is a tremendously exciting adventure at
the frontiers of science,” Lysaght says. “Xilinx
FPGAs are uniquely suited to the huge computing
demands of projects such as SETI.”
Werthimer is unrestrained in his gratitude
to XUP: “The Xilinx University
Program has really made this whole thing
possible. Not only did you guys develop
the core technology that we needed to
process 168 million channels simultaneously,
but you have this generous program
that gives us the software and the chips to
build SERENDIP V.”
Werthimer’s gratitude goes beyond
words. In a memo to XUP last year, he
wrote: “We’d like to reciprocate as best we
can for all the wonderful stuff you’ve
donated.” For example: “We have a flow-through
dual FFT and polyphase filter
design that Xilinx and its customers are
welcome to use. It uses one-fourth of the
memory that the Xilinx FFT IP core uses,
and it calculates two complex FFTs
simultaneously at 240 million samples
per second.”
Contact
Why are companies like Xilinx, Sun
Microsystems, Intel™, Toshiba™,
Hewlett Packard™, Quantum™,
Network Appliance™, Fujifilm™ – as
well as non-profit organizations like the
SETI Institute, the SETI League, and The
Planetary Society – contributing millions
of dollars worth of technology and expertise
to support the search for ET? Xilinx
CEO Wim Roelandts puts it this way:
“You can say, well, maybe it is science fiction,
but to prove it isn’t, you need state-of-the-art technology. This is probably the
ultimate science problem we must solve.
The intellectual challenge is enormous.
And that is what is exciting.”
Within the SETI community, the word
“excited” has become almost a code word
for the discovery of ET. Pretty much everybody
uses the word “excited” when they
describe how they’ll feel when a signal
from ET is scientifically confirmed. That
means replicated results from other, independent
observatories, such as those in
Australia, Italy, France, and Argentina.
“We actually want to look at the sky
many, many times,” Werthimer explains.
“One of our most robust algorithms is:
Did you see the signal again, in same
place, when the telescope comes back to
the same place in the sky? Do we see it in
the same place at the same frequency? –
that’s what gets us really excited.”
Unfortunately, in the last 44 years of
serious, scientific research, nobody’s found
anything to get all that excited about.
“I’m optimistic,” Werthimer says. “I
think we might find ET in our lifetimes,
but I think, right now, we’d be very lucky
to find ET. So, I’m sort of counting on
Moore’s Law. If Moore’s Law keeps going,
and if Xilinx keeps on making faster and
better chips, the better the chances we have
of finding an extraterrestrial radio signal.”
Two Scenarios
“There are sort of two scenarios for contact
with ET,” Werthimer reasons. “One is
that we find a signal, and we aren’t really
able to decode it. It could just be a navigational
beacon – there’s no information. All
we would know is that they’re out there.
“The more scary scenario to me,”
Werthimer continues, “is that we might
receive a direct broadcast with a huge
amount of information content in the ET
signal – and that could be used in good ways
or bad ways.”
To prevent the potential abuse of extraterrestrial
intelligence, virtually all SETI
research organizations – including the SETI
SERENDIP Program, the SETI Institute,
the SETI League, and The Planetary Society,
to name just a few – have endorsed Article
XI of the United Nations Treaty on
Principles Governing the Activities of States
in the Exploration and Use of Outer Space.
In part, Article XI decrees that the discoverers
of ET must “... inform the Secretary
General of the United Nations, as well as the
public and the international scientific community,
to the greatest extent feasible and
practicable, of the nature, conduct, locations,
and results ...” of the discovery.
Furthermore, the treaty calls for a transnational
decision on whether to reply to ET –
and if so, what to say.
Conclusion
We may discover ET any time from now
to never. And it’s impossible to predict exactly
how humanity will react to scientifically
validated proof that life as we know it – isn’t.
For many years, the late astronomerphilosopher
Carl Sagan was the standardbearer
for SETI research. He gave as much
thought to the implications of finding ET
as he did to the technology of the search
for ET. In a 1978 essay, he wrote: “The
search for extraterrestrial intelligence is the
search for a generally acceptable cosmic
context for the human species. … It is difficult
to think of another enterprise within
our capability, and at relatively modest
cost, which holds as much promise for the
future of humanity.”
Hyperlinks to SETI Research
|
| Allen Telescope Array: | www.seti.org/science/ata.html |
| Arecibo Observatory:
| www.naic.edu/bigtable.htm |
Origins: Astrobiology:
The Search for Life:
| www.exploratorium.edu/origins/arecibo/ |
| SETI@home:
| setiathome.ssl.berkeley.edu |
| SETI at the University of
California, Berkeley:
| seti.berkeley.edu |
| SETI Institute:
| www.seti.org |
| SETI League:
| www.setileague.org |
| The Planetary Society:
| seti.planetary.org |
| “The Quest for Extraterrestrial Intelligence” by Carl Sagan:
| www.bigear.org/vol1no2/sagan.htm |
| Xilinx University Program:
| www.xilinx.com/univ/ |
Protocols for Contact with ET
|
| Treaty on Principles Governing the
Activities of States in the Exploration
and Use of Outer Space: |
www.oosa.unvienna.org/SpaceLaw/outersptxt.htm |
| SETI Institute:
| www.seti.org/science/principles.html |
| SETI League:
| www.setileague.org/general/protocol.htm |
| The Planetary Society:
| seti.planetary.org/Contact/AfterTheDetection.html |
Printable PDF version of this article with graphics.  (1/25/04) 595 KB |
