SonATA (or Sno-ATA?)





As a new guest blogger here, I recently had the great opportunity of being able to ask SETI Institute Astronomer Dr. Peter Backus  a few questions about the software known as SonATA (an acronym for "SETI on the Allen Telescope Array") which processes data as it comes into the control room/lab at Hat Creek, California, site of the Allen Telescope Array (ATA). 
 
 The above is a shot from a remote camera at the Hat Creek site of the ATA, taken earlier today.  You can see a few dishes in the background, and you can also see a current image at the site here http://www.seti.org/page.aspx?pid=1311 .  According to Dr. Backus the snow can adversely affect data acquisition via the dishes, as it can diminish signal and affect the drive mechanisms.  To mitigate weather effects, they often point the telescopes so as to minimize snow accumulation and point them toward the Sun to enhance melting once some snow has built up.
 
     I asked Dr. Backus about Open SonATA and it's relationship to setiQuest.  According to Dr. Backus “The Open Source development of SonATA code is a major part of setiquest.  It controls all aspects of the observing from selecting antennas, calibrating the beamformers, selecting targets, pointing the antennas, processing data, mitigating RFI, and following up on candidate ETI signals.  Eventually the Open SonATA software will collect data for distribution to the Cit(izen) Sci(ence) App.”  He also says that “…The Open SonATA v1.0 to be released soon will be essentially identical to the code running on the observing system at Hat Creek.  There may be a few files specific to using the ATA that may not be included in the release.”
 
     Per Peter the ATA will be acquiring data day and night (except during maintenance periods, currently at around 8 hours per week, and during times of inclement weather, like the recent snows they've had.)   There may be other acquisition times that data will not be available for setiQuest or the Citizen Science app., especially when  radio astronomers are using the scope who might have proprietary information to preserve (like pointing positions) that researchers do not want to make public until they are ready to publish their work..  So, real-time data for setiQuest and the General Release Citizen Science application will not be available all of the time, but when possible it is hoped that people around the world will be able to participate in the search in near-real-time, when the data are available.
 

Only yesterday the Allen Tlescope array looked like this, when the snow started falling.  This panorama was made from 18 separate webcam images stitched together for from the ATA cam page.  I wanted to post an image from today, but the images were not available, perhaps because of the snow and ice?

     So what is SonATA?  Aside from appearing to be a homonym of a classical musical form, or automobile brand, SonATA is a software-based approach to SETI signal acquisition and processing (as opposed to the customized hardware intensive approach used in past projects.)  SonATA is designed with future in mind, to be open, scalable, and with consideration of the potential increases in computational speed anticipated by "Moore's Law".

      By utilizing off-the-shelf, constantly increasing computer processing speeds expected in the coming years, and the continued improvements in memory, storage capacity and networking, SonATA permits the possibility of improved performance at a fraction of the cost of creating dedicated signal processing hardware.  SonATA incorporates similar processing algorithm approaches that were used in previous SETI projects, including NASA's original SETI program (20 years ago, known as the High Resolution Microwave Survey or HRMS) .
 
      According to Peter “The first release of Open SonATA code will have the CW and pulse algorithms in use at the ATA.  They trace their roots back to the NASA Targeted Search System but have a few more features and tweaks.”
 but much of it is totally new, it being intended to be one of the first software clients used by the SETI Inst. that can run on a readily available platform.  SETI scientists realized awhile back that not only were off-the-shelf machines processing speeds increasing at a pace that exceeded their ability to match by building dedicated hardware, but indeed, now those same scientists are betting on Moore's Law to help them out of a tough situation, handling the immense amounts of data they anticipate the ATA will be generating.  In the past they used to ship to radio-telescopes around the world a container/shelter/building known as the Mobile Research Facility (MRF) containing 10 racks of gear that included signal processor hardware, but back in 2002 they made the move from digital signal processors being hardware-based to software-based, and eventually created a predecessor program to SonATA called Prelude (somebody likes classical music over there!) which has been used on the already constructed 1st 42 dishes of the Allen Array.

     Prelude is a direct descendant of the previous Phoenix program that was used in the MRF's.  And Phoenix's software can be traced back to NASA's HRMS, so as Dr. Backus says, there's software DNA (so-to-speak) in SonATA that can be traced all the way back to NASA's HRMS!  Dr. Backus decribed Prelude as “…a transition between full custom hardware and full software.  It had some commercial DSP boards (Motorola "Blue Wave") and two small custom boards in a commercial PC.”  The Prelude system also had 20 PCs and other components in four equipment racks.  Peter says “The current SonATA system uses four Dell PowerEdge servers and processes three times as much data.”
 

The above is a diagram Peter forwarded to me that he happened to have on hand for the system.

The ATA's receivers work at frequencies from around 500 mhz to 11 Ghz, and their received RF electrical impulses are then converted to analog optical fiber signals and sent to the Control Room/site lab.  In there are 4 intermediate systems, 2 correlators for radio astronomy and 2 for data sent to either SonATA, or captured and stored locally on disks, including data for setiQuest (initially setiQuest data will then be trickled out within a day or days via their network server.)  For the Beta software (which is a 'closed beta' test not open to the public, which is slated to be released in March 2011) the data used will be recorded (non-real-time) data, but it is envisioned that for the coming Citizen Science software, near-real-time data is hoped to be distributed, when available.

Key scientists, engineers and programmers for SonATA include: Peter Backus, Jane Jordan, Ken Smolek, and Jon Richards.
 
Lastly, Dr. Backus also mentioned that he anticipates the publishing of a paper on ~180 stars that are known to have planets orbiting them, which were surveyed by the ATA, many of which having known exoplanets that were observed in a 2009 campaign.  SonATA's predecessor system 'Prelude' surveyed the 180 stars for signals from E.T.'s. (no luck yet though) and they will be able to set limits on the chances of signals coming from those locations.  Another data point (reed of hay) in the immense haystack (the Universe.)  Maybe the needle will be found in the next pitch of the hay by setiQuest, and you?
 
Keep searching!  Jason W. Higley