I recently received some very interesting questions regarding algorthms. I'd like to share these remarks with you, and please feel free to answer to them specifically:
Put yourself in the shoes of a civilization who decides to so send a message to another star. Wouldn't you begin with a beacon and code the message in the beacon? Your object is to get detected - not hide in the noise. In other words - wouldn't you go out of your way to insure that the signal was detected? So, has anyone really thought about how to create a beacon with a high likelyhood of detection with minimal analysis? I've looked thru 2020 and searched a little on line. I see discussions of coding messages - but I don't see any practical ideas about how you would insure that the signal was detected. Thanks.
These are great questions. People have worried (and continue to worry) about the problem of how the SETI signal will be encoded. We are starting a major project right now that will lead to new ideas and algorithms (know variously as "DATA/API" or "setiCloud"). setiCloud will have its first release to the open source communityh on July 22 at OSCon (http://www.oscon.com/oscon2010). This will be followed up with a specialized SETI conference in Mountain View, CA (http://seticon.com/).
Rather than answer your questions directly, allow me to point you to a paper we recently wrote that has a good introduction to why we normally search for narrow-band tone-like signals. The paper compares spectral-line SETI with a newly proposed signal type and an autocorrelation algorithm to discover this signal type: