SETI@home

In 1999, SERENDIP spun off an imaginative project that quickly became the most exciting landmark in the SETI movement's 48-year history.

Sifting through huge volumes of radio data for narrowband signals takes massive computing power. Given limited budgets, this is a critical bottleneck in the search. Even the powerful supercomputer specially designed for SERENDIP must limit itself to looking only for simple signals with certain predetermined characteristics.

In 1994 David Gedye, a Seattle computer scientist, had a brainstorm. He realized that deeper analysis of SETI radio data would be a perfect project for "distributed computing" by tens of thousands of volunteers using home computers. Development of the project stalled in 1998 due to fundraising problems, but The Planetary Society saved the day by putting up a matching grant of $50,000 (from its Carl Sagan memorial Fund for the Future) and then persuading Paramount Pictures to grant $50,000 for the society to match. As a result, the SETI@home project burst onto the world in May 1999 and has been running ever since.

The idea is simple. You download a program that can install itself as your computer's screen saver if you wish. It then fetches files of data ("work units") recorded by the SERENDIP radio receiver. The program unobtrusively mulls over the data whenever your computer has nothing else to do. When the analysis is done (typically after 2 to 10 hours of processing time per work unit, depending on the speed of your CPU and memory), your computer sends back the results and downloads more work units to chew on.

SETI@home became wildly popular the moment it was unveiled. As of April 2010, ten years into the project, more than 8 million people from nearly every country in the world had run SETI@home for at least a while. As of May 2009, 140,000 were running it enough of the time to be classified as "active users."

What It Does

SETI@home analyzes only one narrow, 2.5-MHz segment of SERENDIP's much wider band. The chosen segment is usually centered on the 1,420-MHz hydrogen line. It is wide enough that a transmitter broadcasting exactly at the hydrogen frequency could be moving toward or away from Earth by up to 270 kilometers per second and still be picked up despite the resulting Doppler shift. This velocity range is wide enough to include most of the stars and planets in the Milky Way (at least in the part of the sky accessible to Arecibo) but generally none in any other galaxies.

Within this 2.5-MHz band, SETI@home provides several important boosts to Project SERENDIP. It increases the sensitivity tenfold, thereby enlarging the volume of space that SERENDIP listens to (for a transmitter of a given power) by a factor of 10^3/2, or about 30. That means 30 times more stars and planets get searched. SETI@home also broadens SERENDIP's ability to discriminate signal widths; it looks at channels that range from 0.075 to 1,220 hertz in bandwidth.

The software can also spot signals that drift in frequency, by as much as 50 hertz per second. A very narrow signal from deep space is expected to show at least a slight frequency drift (up to 0.14 hertz per second as seen by Arecibo) for a simple reason: the receiver is on a rotating planet! An alien transmitter might be too. If the transmitter is in space, it might be in some kind of rapid orbit giving the signal a greater frequency drift. For instance, a transmitter in a low orbit around an Earth-mass planet would show changing Doppler drifts of up to 40 hertz per second.

Each work unit that you receive from SETI@home represents 107 seconds of listening time by Arecibo and about 10 kilohertz of bandwidth. So it takes 256 work units, lined up side by side, to cover the experiment's entire bandwidth for those 107 seconds.

During that brief time, the receiver's beam usually sweeps about 0.3° to 0.6° across the sky, depending on what else the telescope is doing. The beam is about 0.1° wide, so your work unit represents a scan of a tiny strip of sky 0.1° wide by 0.3° to 0.6° long — about as much sky as would be covered by a small rice grain held at arm's length.

As with SERENDIP, promising signals that are analyzed out of the data (and that don't match known radio-frequency interference) will be followed up with dedicated observations. In March 2003 the SETI@home team got its first chance to take control of the Arecibo radio telescope and use it for dedicated followups. At 155 promising signal sites (and 61 other interesting SETI target sites), nothing striking turned up. Fuller analysis of all the SETI@home signals (3.4 billion were on file as of March 2009) will take much longer.

Continued; click "Next Page" below.