SonATA Overview

From setiquest wiki

Revision as of 19:15, 21 June 2012 by SigBlips (Talk | contribs)
(diff) ← Older revision | Latest revision (diff) | Newer revision → (diff)
Jump to: navigation, search

Overview of SonATA

The SonATA system is the first software-only incarnation of the SETI Institute's search system looking for narrowband radio signals of extraterrestrial origin. SonATA runs on general purpose computers, and replaces previous generations of search systems that relied on custom hardware in order to achieve the near-real time performance that is required. By implementing the SETI search system in general purpose, commodity servers, it should be possible for SonATA to benefit from future Moore’s Law growth in processing power. SonATA will be able to upgrade its searching capacity by future upgrades to its host computer cluster.

SonATA conducts two types of SETI observations: targeted searches of selected “Target Stars” from a star catalog, or surveys of an area of the sky using a grid of selected positions. Both search strategies attempt to detect narrowband signals of extraterrestrial origin, while recognizing and discriminating against any detected signals of terrestrial origin. The narrowband radio signals may take the form of continuous or pulsed signals that may drift in frequency over time. These classes of signals have been selected because it does not appear that they can be produced by astrophysical emission processes, but are easily generated by technology. The data are processed in near real-time in order to facilitate discrimination of terrestrial signals by means of immediate, automated follow-up observations.


SonataFlow2.png


SonATA currently runs as a bandlimited version on the ATA (Allen Telescope Array) and uses the multi-beam capability of the ATA to target specific stars or survey directions. SonATA is designed to observe with 2 or more beams looking at a different target with each beam. This multi-target observing mode not only speeds up search, but it helps with radio frequency interference (RFI) identification and discrimination. In general a candidate signal that is seen in multiple beams, it is considered RFI that has been detected within the outlying sidelobes of the phased-up beam patterns.

The ATA currently produces 4 dual-polarization IF channels with output digitized time series samples of voltage for a 104.8576 MHz bandwidth each. SonATA uses polyphase filters to break this input data stream into narrower channels, arranges for a 50% overlap in time, and performs FFTs to generate spectral data with a final resolution of 1 Hz, 2 Hz, or 4 Hz. The spectral data are normalized and statistical detection thresholds are set to achieve an acceptable probability of false alarm from noise alone (and in practice, empirically modified to accommodate the local interference levels). Efficient CW detection and pulse detection algorithms process these spectral data, arriving at 1.5 second frame intervals, identifying statistically significant signals, which are then compared to signals detected in other beams and with a dynamic database of known RFI signals. Data collection is followed by signal detection and classification proceeding in a two-stage pipeline that is constrained to complete on fixed observational boundaries. This constraint on completion forces SonATA to ignore subchannels with too many detected candidates. These subchannels are reported as badbands and are, in effect, not actually observed.

Candidates that remain interesting after this first round of discrimination filtering cause the pipeline to be interrupted and trigger follow-up observations to determine whether the candidate is of extraterrestrial origin. The follow-up observations are a sequence of automated off-source and on-source re-observations intended to remove candidates that were due to noise alone, and identify those that are interesting enough to warrant further attention from a human observer.

A software executive program automatically selects the target stars (or directions) and frequencies for each observation, synchronizes the start of new data collection, and the processing of old data collected in the previous observation, reviews the signal candidate classifications, and decides whether to continue the pipeline process by selecting another frequency or target or break the pipeline to begin reobservation and automated follow-up. The SETI Institute team has had more than a decade of experience running the precursors of SonATA on custom built hardware in configurations that invoked two widely separated astronomical observatories functioning as a pseudo-interferometer for effective RFI discrimination based on the predictable differential Doppler signature of signals detected at the two sites. We have only recently begun operating the partially-custom Prelude detection system on the 42-element Allen Telescope Array, and are still learning how best to use the multiple antennas of a true interferometer to help with RFI discrimination.

See also

Index Seeker (SSE) →
Personal tools
Namespaces
Variants
Actions
Navigation
Toolbox