Reobservations Report No. 2: Reobserving, Recording, and Reprocessing
As SETI@home revisits its most promising candidate signals between March 18th and 20th, the Arecibo radio telescope will be receiving vast amounts of radio data from the targeted regions of the sky. All of this information must be preserved for later processing and analysis. Only when this is completed will SETI@home Chief Scientist Dan Werthimer and his team know for sure whether the data contains traces of an intelligent signal from the stars.
To preserve the data, Werthimer and his colleagues will rely on two different recorders located at the Arecibo Observatory. One of these is the SETI@home recorder itself - the same one used to record the data from the "piggy-back" receiver throughout the year. It will record the reobservations data just as it records ordinary SETI@home signals, and with the same 2 bit accuracy.
For the first time, however, during the reobservations, Werthimer and his crew will have use of another recorder. This is Arecibo's "radar" recorder, built for those occasions when the giant dish is used as a radar, bouncing electromagnetic signals off planets, moons, and asteroids. Astronomers employ this method to map the various objects in our Solar System, and they need a recorder of the highest sensitivity to preserve the radar measurements.
Because of these requirements, Arecibo's 8 bit radar recorder is more sensitive than SETI@home's 2 bit recorder. This suits SETI@home Chief Scientist Dan Werthimer just fine: "During ordinary observations," he explained, "we have so much data to record with SETI@home that we record very few bits, but then we sacrifice some sensitivity." "With follow up observations," he added "we can afford to record a lot of data, because it's only for a short time."
Both recorders will operate simultaneously during the reobservations, preserving all the incoming data at their different degrees of sensitivity. Once the observations are completed and the data delivered back to Berkeley, however, the information captured by the two recorders will be treated very differently.
The 2 bit recorder's data will be processed like any other SETI@home data tape: it will be divided into normal work units and sent out for analysis by the project's 4 million users around the world. It is, after all, only appropriate for SETI@home that any potential alien signal will be detected on one of the personal computers that make up its vast distributed computing network.
The analysis of the radar recorder's data will take considerably longer. Since the format of the recordings is different from the usual SETI@home tapes, they cannot be analyzed by the existing SETI@home client program - the one installed on millions of personal computers worldwide. A different type of distributed computing system is needed, and fortunately one is at hand.
For some months now, SETI@home Project Director David Anderson and colleagues have been working on perfecting the "Berkeley Open Infrastructure for Network Computing," more commonly known as BOINC. Inspired by the phenomenal success of SETI@home, BOINC is a consortium of various scientific distributed computing projects, all of which share the basic infrastructure for sending out calculations and collecting processed data. Now Anderson and his team plan to modify the existing SETI@home program to make it compatible with BOINC standards, and to adapt it to the analysis requirements of the data collected by the radar recorder.
All this, however, is some months in the future. In the meantime SETI@home users can look forward to processing those special work units from the Arecibo reobservations on their personal computers, using the tried and true SETI@home client program. This time they will be analyzing not a random section of the sky, but only that select set of the best and most promising candidates signals. Any volunteers?