The signal detection problems facing those who search the sky for evidence of extraterrestrial civilizations or interesting natural phenomena are enormous. Among those problems are the fact that we don't know much about what to expect. In particular, we don't know exactly where to look in the sky, or what frequencies might be used, or the possible forms of the transmissions. Moreover, the many powerful sources of natural and terrestrial sources of interfering signals must not be confused for extraterrestrial ones. There is also the massive computer power required, which has driven the development of specialized hardware and software as well as distributed computation over thousands of Internet-connected personal computers across the world using the SETI@home computational screen-saver. Although many of the computational techniques used in this search are far more sophisticated than those covered in this web site, they begin with the basic concepts covered here.
One of the reoccurring themes of this site has
been that the more you know about your data, the more
likely you are to obtain a reliable measurement. In the
case of possible extraterrestrial signals, we don't know
much, but we do know a few things.
We know that electromagnetic radiation over a wide range
of frequencies is used for long-distance transmission on
earth and between earth and satellites and probes far
from earth. Astronomers already use radio telescopes to
receive natural radiations from vast distances. In order
to look at different frequencies at once, Fourier
transforms of the raw telescope signals
can be computed over multiple time segments. We
previously saw a simulation that showed how
hard it is to see a periodic component in the presence
of on equal amount of random noise and yet how easy it
is to pick it out in the frequency spectrum.
Also, transmissions from
extraterrestrial civilizations might be in the form
of equally-spaced pulses, so their detection and
verification is also part of SETI signal processing.
Interestingly, triplets and other groups of equally
spaced pulses appear in the Fourier transforms of high
frequency carrier waves that are amplitude
or frequency modulated (like AM
or FM radio). Of course, there is no reason to
assume, nor to reject, that extraterrestrial
civilizations might use the same methods of
communication as ourselves.
One thing that we know for
sure is that the earth rotates around its axis once a
day and that it revolves around the sun once a year. So
if we look at a fixed direction out from the earth, the
distant stars will seem to move in a predictable
pattern, whereas terrestrial sources will remain fixed
on earth. The huge Arecibo
Observatory dish in Puerto Rico is fixed in
position and is often used to look in one selected
direction for extended periods of time. The field of
view of this telescope is such that a point source at
a distance takes 12 seconds to pass. As
SETI
says: "Radio signals from a distant transmitter
should get stronger and then weaker as the telescope's
focal point moves across that area of the sky.
Specifically, the power should increase and then
decrease with a bell shaped curve (a
Gaussian curve). Gaussian curve-fitting is an excellent test to
determine if a radio wave was generated 'out there'
rather than a simple source of interference somewhere
here on Earth, since signals originating from Earth
will typically show constant power patterns rather
than curves". Also, any observed 12 second peaks can
be re-examined with another focal point shifted
towards the west to see if it repeats with the
expected time and duration.
We also know
that there will be a Doppler
shift in the frequencies
observed if the source is moving relative to the
receiver; this is observed with sound
waves as well as with electromagnetic
waves
like radio or light. Because the earth is
rotating and revolving at a known and constant speed,
we can accurately predict and compensate for the
Doppler shift caused by earth's motion (this is called
"de-chirping"
the data).
For more on the details of SETI signal processing, see
SETI@home.