Number 196 - September 1999
You Can Try To Find ET
by George Cottay
    Please imagine that, many years ago on a planet circling a sun somewhere on a distant arm of our galaxy, an intelligent form of life sent a powerful radio signal out into the void. They carefully chose a frequency for their signal of about 1.419580076 GHz to give it the best chance of arriving relatively intact somewhere out there in the universe.

    Because of the near unimaginable distances involved, the signal did not reach earth until early this year when it was picked up by a special receiver and data processor on Cornell University's giant Arecibo radio telescope in Puerto Rico. That, however, is where it stops. We have heard from intelligent life but don't know it yet.

    The task of filtering all other radio frequency noise and detecting a pattern is so complex that SETI, the organization using the Arecibo facility for listening, cannot even begin separating the ancient signal from all the noise and detecting its pattern. The cost of assembling the necessary computing power would be astronomical.

    That's where you come in. If you have a fairly powerful computer (32 megs of memory and video resolution of 800 by 600 and 256 colors or better), and an Internet connection, you can easily donate some of your machine's processing power to the task of finding ET's signal in the data haystack.

    http://setiathome.ssl.berkeley.edu

    Unlike much of what you hear about the Internet, this one is for real. You can download a smallish program from SETI (about five minutes at 28.8), complete a very brief registration, download your own chunk of data for analysis (again about five minutes), and turn your machine loose on it. In a day or two when the process is complete you upload the results, and get another set of information for your machine to comb for signs of life.

    The process is really simpler to accomplish than to read about. SETI has designed the process well. You do very little work. The work your computer does just uses capacity that would otherwise be wasted while it's sitting there or doing simple tasks. You have your choice whether you want the analysis to happen all the time or just when you aren't doing anything else.

    If one of the data chunks your machine turns out to provide evidence of extraterrestrial intelligence, you, not your machine, will get credit for the discovery. You need not worry about raising false alarms. The analysis and data return processes are automated. All possible signal discoveries will be thoroughly checked.
    At the time this is written, just four days after the project went public, a quarter-million machines were at work processing SETI data. Based only on jobs returned, not counting the many thousands still in progress, over 600 years of processing time have been saved.

   With computers all over the world working on the task, SETI expects available data from Arecibo to be fully analyzed, a task that otherwise might never have been completed. Response so far raises the possibility of finishing early and moving on to record and analyze a wider band of the radio spectrum.

    My considered mature opinion of all this is WAY COOL!

    Even though I suspect we are not alone in possessing what passes for intelligence in this outpost of the Milky Way, I'd rate the prospects of SETI finding a meaningful signal as rather remote. The sky is too large, the radio spectrum too wide, and time too long for odds to be high.

    For me, the combined facts that we are listening and using computers around the world to make sense of what we hear, are worth some excitement even if we don't hear from ET.

    George Cottay directs the Quad Cities Online Training Center and writes Online. Questions, comments and suggestions are welcome at cottay@qconline.com or by snail mail at 1724 4th Ave., Rock Island, IL 61201. Online is also on the web, linked from www.qconline.com/index.html


TOGGLE Editor's Note
    I just started participating in this program and was fascinated with the display, but perhaps I'm easily amused. Based on the observed rate it was going, shown on the screen, I calculated that my machine's analysis of the first chunk of data it received would take 30 to 31 hours. It was mentioned, however, that displaying the analysis slows the calculation because some time, however short, is spent displaying it.

   You can shut off the display by sending it to the background and go on with your computing. The screen on my "green" machine, with a 450MHz Pentium III, goes dark when it is idle. After being idle overnight, but still powered up and analyzing, the estimated time to complete is more like 25 or 26 hours. You don't have to do anything once the process is started. It is all automatic. If the program is interrupted, it saves its place and starts up again where it left off--a very well-designed program.
 
  Number 196 - September 1999