General question about data that's processed by SETI@Home


log in

Advanced search

Message boards : SETI@home Science : General question about data that's processed by SETI@Home

Author Message
Pavix
Send message
Joined: 4 May 11
Posts: 2
Credit: 1,886,266
RAC: 223
United States
Message 1104353 - Posted: 8 May 2011, 8:35:43 UTC

So I have my laptop crunching away and my desktop and got to thinking about the information that's being crunched. My question is 2 fold, 1) how many times is the same data processed by different people before it's pulled and no longer processed or does all data go through every SETI@Home member? And 2) Is the data we get looked at before we get it? Is there a genuine chance that someone will be the first to crunch an alien intelligence transmission?

Blake Bonkofsky
Volunteer tester
Avatar
Send message
Joined: 29 Dec 99
Posts: 617
Credit: 46,332,781
RAC: 0
United States
Message 1104354 - Posted: 8 May 2011, 8:39:03 UTC - in response to Message 1104353.

1) Each workunit is initially sent to 2 people. As long as both of those people both return a result, and the results match, that's it. Done. If there is an error, or the 2 results don't match, it is sent out a third time. Up to 10 possible people will see the same WU before it just completely errors out and the server gives up.

2) I know there is some very low level analysis and "cleaning" of the data to look for and remove local RFI interference, but beyond that, I'm 99% sure the data is untouched when it reaches our computers
____________

Pavix
Send message
Joined: 4 May 11
Posts: 2
Credit: 1,886,266
RAC: 223
United States
Message 1104355 - Posted: 8 May 2011, 8:40:31 UTC - in response to Message 1104354.

Awesome! Thank you :)

Profile Lint trapProject donor
Send message
Joined: 30 May 03
Posts: 860
Credit: 26,792,207
RAC: 17,769
United States
Message 1104404 - Posted: 8 May 2011, 15:34:33 UTC - in response to Message 1104354.

1) Each workunit is initially sent to 2 people. As long as both of those people both return a result, and the results match, that's it. Done. If there is an error, or the 2 results don't match, it is sent out a third time. Up to 10 possible people will see the same WU before it just completely errors out and the server gives up.


Blake, I believe the current limit for consensus is 7 tries. I've seen a couple of workunits process ok on my machine and then end in "Can't Validate" status after the 7th attempt failed.

Martin

Josef W. SegurProject donor
Volunteer developer
Volunteer tester
Send message
Joined: 30 Oct 99
Posts: 4252
Credit: 1,050,227
RAC: 244
United States
Message 1104506 - Posted: 8 May 2011, 21:36:36 UTC - in response to Message 1104404.
Last modified: 8 May 2011, 21:41:30 UTC

1) Each workunit is initially sent to 2 people. As long as both of those people both return a result, and the results match, that's it. Done. If there is an error, or the 2 results don't match, it is sent out a third time. Up to 10 possible people will see the same WU before it just completely errors out and the server gives up.


Blake, I believe the current limit for consensus is 7 tries. I've seen a couple of workunits process ok on my machine and then end in "Can't Validate" status after the 7th attempt failed.

Martin

That's the usual case since it is highly unlikely there will be 5 results returned as successful processing without getting a strongly similar pair, but it is certainly possible. There's a limit of 5 with errors and 5 "success" as well as the overall 10 limit.

As for how often the data is processed, it should not be forgotten that the same data is split for both S@H Enhanced and Astropulse processing. IOW, at least 4 tasks are generated from each part of the data. Further, the S@H Enhanced splitting is done with about a 20 second overlap, so data near the end or begining of those tasks is also in another pair of tasks.
Joe

Profile Jim-R.
Volunteer moderator
Volunteer tester
Avatar
Send message
Joined: 7 Feb 06
Posts: 1475
Credit: 46,565
RAC: 0
United States
Message 1106149 - Posted: 14 May 2011, 14:54:54 UTC

You must also consider that we are not looking for signals that jump out at us like a multi megawatt radio station nearby. We are looking for signals that are possibly hidden in the background noise so just as happened when the Enhanced app came out, we crunched data that had already been processed by the former SAH app because the Enhanced app could detect much weaker signals buried deeper into the noise. As Josef pointed out, our present data is being crunched by the Enhanced app and the Astropulse app. When we come out with a new app that is able to dig deeper into the noise we may crunch all of the data again.
I like to use the analogy of a quarter buried in the sand. The classic app was like using a pitchfork to dig through the sand looking for the quarter. The Enhanced app is like digging with a garden rake. The next super app might be compared to using a comb. Even though you don't find anything with the pitchfork it doesn't mean there's nothing there. You might be able to find it with the garden rake. And we might miss it with a garden rake so we can look again with a comb.
In the future we might want to look for a different type signal than either Enhanced or Astropulse are looking for so we might crunch the same data again with a different app.
So there's no set answer as to how many times a specific portion of the data is processed.

As to your second question, no one really knows. So many things would have to come together at once for this to happen.
First is whether an alien intelligence actually does exist.
Second is whether they existed at the particular time at which a signal sent by them is reaching us now at a technological level to be able to send such a signal.
Third is whether they actually send a signal at the particular frequencies that we are looking at.
____________
Jim

Join GQFCrunchers
Problems?
Check the Boinc Wiki
Or NEW Enhanced FAQ

Message boards : SETI@home Science : General question about data that's processed by SETI@Home

Copyright © 2014 University of California