Online Pipelines
================

During the third LIGO observation run a number of search pipelines will be
running in a low latency, online mode. These can be divided into two groups,
modeled and unmodeled. The modeled (CBC) searches specifically look for signals
from compact binary mergers of neutron stars and black holes (BNS, NS-BH, and
BBH systems). The unmodeled (Burst) searches on the other hand, are capable of
detecting signals from a wide variety of astrophysical sources in addition to
compact binary mergers: core-collapse of massive stars, magnetar star-quakes,
and more speculative sources such as intersecting cosmic strings or as-yet
unknown GW sources.

Modeled Search
--------------

**GstLAL**, **MBTAOnline**, **PyCBC Live** and **SPIIR** are matched-filtering
based analysis pipelines that rapidly identify compact binary merger events,
with sub-minute to ∼1 minute latencies. They use discrete banks of waveform
templates to cover the target parameter space of compact binaries, with all
pipelines covering the mass ranges corresponding to BNS, NS-BH, and stellar
mass BBH systems. However, GstLAL and PyCBC Live and SPIIR also include
intermediate-mass BBH systems and the O2 banks differ in detail from pipeline
to pipeline.

A coincident analysis is performed by GSTLAL, PyCBC Live, and MBTAOnline, where
candidate events are extracted separately at each detector via
matched-filtering and later combined across detectors. SPIIR extract candidates
of each detector via matched-filtering and look for coherent responses in other
detectors that a localization of the source can be constructed. Of the four
pipelines, GstLAL and MBTAOnline use several matched filters to cover the
detector bandwidth i.e., the matched filter is split across multiple frequency
bands. All pipelines also implement different kinds of signal-based vetoes to
reject instrumental transients which cause large SNR values but can otherwise
be easily distinguished from compact binary coalescence signals

**GSTLAL** [#GSTLAL]_ is a matched-filter pipeline designed to find
gravitational-waves from compact binaries in low-latency. It uses the
likelihood-ratio, which increases monotonically with signal probability, to
rank candidates, and then uses Monte Carlo sampling methods to estimate the
distribution of likelihood-ratios in noise. This distribution can then be used
to compute a false alarm rate and p-value.

**SPIIR** [#SPIIR]_ [#SPIIRThesis]_ applies zero-latency SPIIR filters to
approximate matched-filtering results. It selects high-SNR events from each
detector and find coherent responses from other detectors. It constructs
background by time-shifting detector data one hundred times over a week to form
a background statistic distribution used to evaluate foreground candidate
significance.

**MBTA** [#MBTA]_ constructs its background by making every possible
coincidence from single detector triggers over few hours of recent data. It
then folds in the probability of a pair of triggers passing the time
coincidence test.

**PyCBC Live** [#PyCBC1]_ [#PyCBC2]_ estimates the noise background by
performing time-shifted analyses using triggers from a few hours of recent
data. Single-detector triggers from one of the LIGO detectors are time shifted
by every possible multiple of 100 ms, thus any resulting coincidence must be
unphysical given the ∼10 ms light travel time between detectors. All such
coincidences are recorded and assigned a ranking statistic; the false alarm
rate is then estimated by counting accidental coincidences louder than a given
candidate, i.e. with a higher statistic value.

Unmodeled Search
----------------

**cWB** [#cWB]_ is a power excess algorithm focused to identify
gravitational-like signals with short time duration. It uses a wavelet
transformation to identify time-frequency pixels which can be grouped in a
single cluster if they satisfy neighboring conditions. A tuned version for
compact-binary coalescences chooses the time-frequency pixels if they mainly
follow a frequency increasing pattern. A maximum-likelihood-statistics
calculated over the cluster is used to identify the proper parameter of the
event, in particular the probability of the source direction and the coherent
network signal-to-noise ratio. The last one is used to assign detection
significance to the found events.

**oLIB** [#oLIB]_ uses Q transform to decompose GW strain data into several
time-frequency planes of constant quality factors :math:`Q`, where :math:`Q
\sim \tau f_0`. The pipeline flags data segments containing excess power and
searches for clusters of these segments with identical :math:`f_0` and
:math:`Q` spaced within 100 ms of each other. Coincidences among the detector
network of clusters with a time-of-flight window up to 10 ms are then analyzed
with a coherent (i.e., correlated across the detector network) signal model to
identify possible GW candidate events.

Coincident with External Trigger Search
---------------------------------------

**RAVEN** [#RAVEN]_ In addition, we will operate the Rapid On-Source VOEvent
Coincidence Monitor (RAVEN), a fast search for coincidence between GW online
pipeline events and gamma-ray bursts or galactic supernova notifications. A
similar approach is under development for high energy neutrinos.

.. |apj| replace:: *Astrophys. J.*
.. |cqg| replace:: *Class. Quantum Grav.*
.. |prd| replace:: *Phys. Rev. D*

.. [#GSTLAL]
   Messick, C., Blackburn, K., Brady, P., et al. 2017, |prd|, 95, 042001.
   https://doi.org/10.1103/PhysRevD.95.042001

.. [#SPIIR]
   Hooper, S., Chung, S. K., Luan, J., et al. 2012, |prd|, 86, 024012.
   https://doi.org/10.1103/PhysRevD.86.024012

.. [#SPIIRThesis]
   Chu, Q. 2017, Ph.D. Thesis.
   https://api.research-repository.uwa.edu.au/portalfiles/portal/18509751

.. [#MBTA]
   Adams, T., Buskulic, D., Germain, V., et al. 2016, |cqg|, 33, 175012.
   http://doi.org/10.1088/0264-9381/33/17/175012

.. [#PyCBC1]
   Nitz, A. H., Dent, T., Dal Canton, T., Fairhurst, S., & Brown, D. A. 2017, |apj|, 849, 118.
   https://doi.org/10.3847/1538-4357/aa8f50

.. [#PyCBC2]
   Dal Canton, T., & Harry, I. W. 2017.
   https://arxiv.org/abs/1705.01845

.. [#cWB]
   Klimenko, S., Vedovato, G., Drago, M., et al. 2016, |prd|, 93, 042004.
   https://doi.org/10.1103/PhysRevD.93.042004

.. [#oLIB]
   Lynch, R., Vitale, S., Essick, R., Katsavounidis, E., & Robinet, F. 2017, |prd|, 95, 104046.
   https://doi.org/10.1103/PhysRevD.95.104046

.. [#RAVEN]
   Urban, A. L. 2016, Ph.D. Thesis.
   http://adsabs.harvard.edu/abs/2016PhDT.........8U