The Search for Extra-Terrestrial Intelligence (SETI) is an arduous task. What to look for and how to look for it are extremely difficult questions themselves, and even assuming that we do know what signs of ETI would look like, the process of detecting those signals involves sifting through immense quantities of observational data. In the past, such work required prohibitively expensive supercomputers. Now, however, an emerging technology known as "distributed computing" may hold the key to a more cost-effective SETI by utilizing the spare CPU cycles of idle machines belonging to Internet volunteers. Even SETI skeptics have to agree that processing blocks of information from SETI projects is as least as good a use of idle CPU time as "doing nothing". Still, the inherent complexity of SETI suggests that even the dramatically improved efficiency and processing power that distributed computing makes possible cannot guarantee positive results.
The Target
The goal of SETI is inherent in the name itself--"Extra-Terrestrial Intelligence" is what SETI researchers are interested in. But what are the signs of ETI? This question is very difficult, and has prompted a veritable cottage industry of speculation. Defining the signs of ETI--the "target" for any search--requires an understanding of undiscovered (and possibly nonexistent) civilizations.
First, many SETI researchers believe that the civilizations we are most likely to detect are those similar to or slightly beyond our own in technological development. Following N. S. Kardashev's nomenclature, SETI researchers classify civilizations as Type I, Type II or Type III. Type I civilizations understand basic physics and are equipped for rudimentary interstellar communication. A Type II civilization, on the other hand, colonizes its entire solar system and utilizes all the energy of their sun. Type III civilizations capture energy on a galactic scale, and may even be able to rearrange galaxies.1 As a Type I civilization, mankind has only recently developed the ability to detect a wide range of potential interstellar communications media. Therefore, it is likely that advanced Type II and Type III civilizations, rather than emerging Type I civilizations, could be detected most easily from earth. However, highly advanced Type III civilizations might use communication techniques beyond our ability to detect. A particularly apt metaphor likens our situation to "the inhabitants of the valleys of New Guinea who may communicate by runner or drum, but who are ignorant of the vast international radio and cable traffic passing over, around and through them."2 According to this arguments, either Type II civilizations or those near our stage in Type I are the most likely candidates for detection.
Second, SETI researchers have come to a general consensus that there is a range of radio frequencies in which such civilizations would generally communicate. They assume that the relative energy cost of producing radiation is the same for all civilizations, with less-energetic radio waves being less "energy expensive" than visible light. Assuming fiscal prudence on the part of extraterrestrial civilizations, relatively inexpensive radio transmission should be widely popular, especially for long distance messaging. Also, radio waves are less susceptible to interstellar absorption than other frequencies, so communicating civilizations might utilize radio waves in hopes of getting strong messages through interstellar space undampened.3 "Leakage" from the extraterrestrial equivalents of radio or television broadcasts might also reach Earth.
Finally, SETI researchers hope that interstellar communication might happen in an extremely narrow range of radio frequencies--the so-called "water hole" between the hydrogen line of 1420 megahertz and the first hydroxyl line of 1662 megahertz.4 There are no other known spectral lines between these frequencies, suggesting that civilizations would use this range for communication because it is a very quiet channel. Although this hardly ensures that the "water hole" is the all-purpose intercivilizational frequency, it doesn't make it a bad place to look, either.
The "quietness" of this range of frequencies indirectly implies what it is that SETI researchers are looking for: noise. Beyond the omnipresent cosmic background radiation remaining from the Big Bang, there is almost no natural emission in this range, and even the background radiation is essentially at a constant level. Even low-power transmitters exceed this natural background.5 Still, finding such signals in an interstellar sea of silence is a daunting task for two reasons. First, the search must scan millions of frequency channels, and it must do this repeatedly (every few seconds). Second, those signals above background levels must be examined to eliminate interference from terrestrial radio sources. This second problem complicates the search a great deal, and modern projects muddy the issue even further by attempting to account for Doppler shifts in their algorithms.
In 1959, the first radio SETI project began under Frank Drake as "Project Ozma" at the National Radio Astronomy Observatory in Green Bank, West Virginia. Under the directorship of Otto Struve, Drake turned the 85-foot dish towards two nearby sunlike stars: Tau Ceti and Epsilon Eridani. Except for a "false alarm" generated by some secret military tests, Ozma did not produce positive results. Of course, with only two subjects and a limited period of observation, positive results were unlikely. If anything, Ozma was a practice run.6
After Ozma, new SETI projects began, including the Search for Extraterrestrial Radio Emissions from Nearby Developed Intelligent Populations (SERENDIP) at UC Berkeley. Instead of using dedicated telescope time, SERENDIP's advantage was its "piggybacking" approach--it analyzed data ostensibly collected for some other purpose. The first SERENDIP project, starting in 1979, used Berkeley's Hat Creek Observatory. SERENDIP II used the same National Radio Astronomy observatory telescope that Ozma had, and SERENDIP III began in 1992 using the world's largest radio telescope, the 1000-foot diameter dish at the Arecibo Observatory in Puerto Rico. SERENDIP III examined a 12 megahertz-wide band around 429 megahertz. While SERENDIP III was a much more extensive and powerful search than Ozma, it too did not find any incredibly exceptional signals.
The increased accuracy of SERENDIP III meant increased computational burden. Simply collecting radio telescope data is difficult: to even be usable it must be processed via Fast Fourier Transformation (FFT). This can be anything but fast, especially when it must be repeated billions of times. And while the very broadest search for the most striking signals over many channels has so far failed to produce any convincing ETI candidates, it has nonetheless attracted the patronage of computer companies such as Toshiba America, Sun Microsystems, Xilinx and Intel.7
Even with these sponsors and their generous hardware donations, SERENDIP III's final analysis remains incomplete. The embarrassing six-year backlog shows how massive the computational burden is. With SERENDIP III data-parsing still unfinished, the even more ambitious SERENDIP IV began in June of 1997 in conjunction with systems upgrades at the Arecibo Observatory. Where SERENDIP III observed a 12 megahertz band, SERENDIP IV will examine a 100 megahertz band, increasing the amount of data to be processed still more. Will the analysis of the even greater amount of data collected in SERENDIP IV take decades? With computer demands rising and no particularly striking results, just how long can projects like SERENDIP go on?8
Dealing with Computational Intensity
One new technology for performing computationally Herculean tasks such as the thorough analysis of SERENDIP IV data is to use distributed computing. This technique breaks the problem into smaller constituent parts, sends them out to personal computers around the world via a network such as the Internet and then reassembles the answer after it has been processed block by block by the cooperative computers during their idle time. This is really ideal, because using the idle time doesn't take computing power away from anyone--it only uses time which would normally be spent displaying a screen saver or simply doing nothing while sitting in a wait loop.
Collective computing, even utilizing only a tiny fraction of the PCs available, has already triumphed over traditional supercomputing in the hunt for prime numbers--the most recent largest primes were not discovered with Crays but by an Internet consortium, the Great Internet Mersenne Prime Search (GIMPS). Factoring into primes, finding Golomb rulers and even code-deciphering are also coming to be increasingly dominated by distributed processing techniques.
Distributed computing is not a panacea, however--it is suited only for certain types of applications. If the problem at hand is one in which individual blocks are processed separately, distributed computing is ideal. For algorithms in which each block is processed dependant on other blocks, distributed computing faces bandwidth restrictions. The modern Internet is both too slow and too volatile in terms of throughput for distributed computing to effectively tackle this latter type problem. Still, distributed computing is immensely valuable for the tasks for which it is suited. With 30 million or so computers connected to the Internet and more coming on line each day, the potential power of distributed computing promises to revolutionize problem-solving in many cases.9
SETI@home
Considering the still unprocessed backlogs from SERENDIP III, it is clear that SETI projects like SERENDIP need more computing power. Could distributed computing be put to use in SERENDIP IV? A University of Washington-based organization called Big Science believes so. Under founder David Gedye, Big Science has initiated a project called "SETI@home" to make this a reality.
Distributed computing is well suited to the search for extraterrestrial civilizations for several reasons. First, the problem itself consists of small blocks of data which each require a large amount of processing. Since CPU time, not bandwidth, is the major requirement of the SERENDIP data analysis, distributed computing via the Internet will be very feasible. Second, recruiting volunteers from the Internet public--SETI "hobbyists" willing to devote their computers idle time to SETI@home--has proven to be easy. Although the software to run SETI@home is not yet complete, over 80,000 people have already asked to be involved. And despite being in only an infant state, SETI@home has received a lot of media attention--appearing in the New York Times, on the Discovery Channel, in PC World and on National Public Radio.
It may seem puzzling that public fascination with SETI@home is so high, since SETI research traditionally has had money problems. The congressional decision to end SETI funding in 1993, for instance, might seem to reflect a general distrust of SETI. However, public interest in SETI@home makes sense considering that the PC enthusiasts excited about SETI@home are a very different demographic than grant-controlling bureaucrats and politicians. Furthermore, the appeal of using a resource currently wasted in an epic scientific quest is much greater than grant-controllers interest in shoveling limited funds into a program which has yet to produce much beyond an enormous data backlog.
SETI@home hopes to eliminate this backlog--Big Science estimates that it would only take 50,000 PCs to equal the computational power available to all current SETI projects combined. SETI@home would parcel out the stored tape data to client computers on the Internet, sending a quarter-megabyte to each client every five days. The client software would run as a "screensaver" which would display the status of the SETI@home project around the world.
While this "screensaver" was running, the client would be processing the quarter-megabyte data block, which would contain 50 seconds within a 20-kilohertz range. The algorithm examines this data for strong signals or "chirps" while taking Doppler shifting into account. False alarms of the Ozma variety would be prevented by tests for terrestrial interference.
Once a block was processed, it would be returned to a centralized SETI@homeError computer where the results would be stored and organized. This process, when replicated tens or hundreds of thousands of times, has the capacity to analyze the SERENDIP IV data much more closely than before, perhaps noticing subtle patterns that real-time signal processing missed. The overall results of the search would appear on the SETI@home web site, making the findings immediately available to the public and to the participants.10
On the most basic level, SETI@home provides the capacity to analyze SERENDIP data more finely than real-time processing, in a much more cost-effective manner. This however does not ensure that SETI@home will discover extraterrestrial civilizations soon, or ever, for that matter. Beyond the task at hand, SETI@home may lay the groundwork for a community of computer volunteers and future projects in distributed computing-from the analysis of seismic data for earthquake research to work on the human genome project. According to Gedye, exciting the world about a major volunteer science project is ultimately as important as finding ETI.11
That is not to say that SETI@home is without problems. For all the media attention and public interest, funding has not been forthcoming. Developing new software to run the distributed system and to perform the analysis on the client side is a difficult and expensive process. The SETI@home project has been delayed repeatedly due to lack of corporate sponsorship. "People time," rather than computer power, has proven to be hard to come by, and in the end it seems that expense--the very thing that SETI@home and distributed computing are meant to escape--may be a force as inexorable as gravity.
Whether SETI@home will overcome financial obstacles and become a success remains to be seen. Regardless, distributed computing holds great promise as a future tool for SETI. And beyond discovering signs of ETI, distributed computing may serve to bring the excitement of the scientific process to the consumer level. This is a success whether or not an extraterrestrial civilization is discovered. And besides, there is no reason to consider the project a failure if it does detect extraterrestrial life. If researchers put the entire array of Arecibo data through the most rigorous analysis we can devise and still find no sign of life, well, that too is new information that SETI researchers did not have before. When dealing with marsenne primes or the latest encryption problem, there is always an answer--the only question is how long it will take to find. With SETI, the very existence of an answer is itself a question that needs answering. And although distributed computing holds the key to searching for signs of ETI more efficiently, only people can choose what signs to look for.
Endnotes
1 Goldsmith, Donald and Owen. The Search for Extraterrestrial Intelligence. Second Edition. Addison-Wesley, 1993. (pages 426-427)
2 Sagan, Carl. "On the Detectivity of Advanced Galactic Civilizations." Appendix C from Sagan, ed. Communication with Extraterrestrial Intelligence. Cambridge: MIT Press, 1973. (pages 365-370)
3 Goldsmith, Donald and Owen. The Search for Extraterrestrial Intelligence. Second Edition. Addison-Wesley, 1993. (pages 456-457)
4 Drake, F. et al. "Techniques of Contact." in Sagan, ed. Communication with Extraterrestrial Intelligence. Cambridge: MIT Press, 1973. (pages 284-285)
5 "SETI in the Electromagnetic Spectrum." Online. Internet. 16 Apr. 1998.
6 Sullivan, Walter. We Are Not Alone: The Search for Intelligent Life on Other Worlds. New York: McGraw-Hill, 1964.
7 "SERENDIP Newsletter: Fall 1997." Online. Internet 16 April 1998.
8 "Desription of the SERENDIP Project." Online. Internet. 7 Mar. 1998.
9 Hayes, Brian. "Collective Wisdom." American Scientist, Volume 86, No. 2. Mar. to Apr. 1998 (pages 118-122)
10 "SETI@home" http://setiathome.ssl.berkeley.edu. Online. Internet. 28 Mar. 1998.
11 Dunn, Ashley. "Breaking Down the Search for Extraterrestrials With Distributed Computing." The New York Times. 4 Jun. 1997.
© 1998 Garrett Moritz. All rights reserved.
Launchpad for further explorations (added 3/2001):
More about SETI SETI Music More about Distributed Computing
In print:
Seti Pioneers : Scientists Talk About Their Search for Extraterrestrial Intelligence by David W. Swift (hardcover)
Seti Pioneers : Scientists Talk About Their Search for Extraterrestrial Intelligence by David W. Swift (softcover)
Beyond Contact: A Guide to SETI and Communicating with Alien Civilizations by Brian McConnell (Paperback)
Life Here? There? Elsewhere? : The Search for Life on Venus and Mars/Book, Cards, 1 Videotape, Poster (Life in the Universe Series)
by Seti Institute (Paperback )
Project Haystack : The Search for Life in the Galaxy (Life in the Universe Series) by Seti Institute (Paperback)