The SPIIR online coherent pipeline to search for gravitational waves from compact binary coalescences

This paper presents the SPIIR pipeline used for public alerts during the third advanced LIGO and Virgo run (O3 run). The SPIIR pipeline uses infinite impulse response (IIR) to perform zero-latency template filtering and this process is accelerated with graphics processing units (GPUs). It is the first online pipeline to use the coherent statistic, the network likelihood ratio statistic, to select candidates from individual detectors. Here we express this coherent method using singular-value-decomposition (SVD) technique and show that single-detector signal-to-noise ratios can be directly used to construct this network statistic. Coherent searches are not widely used due to known problem of searching over high-dimension parameter space. To address this challenge, we pre-select single-detector candidates and accelerate the process using GPUs. We demonstrate that the O3 latency of the pipeline is around nine seconds, and present an achievable roadmap to reduce the latency to one to five seconds. The detection performance is reported using a segment of public O2 data. During the O3 online run, SPIIR registered triggers associated with 37 of the 56 non-retracted public alerts. The extreme low-latency nature makes it a competitive choice for joint time-domain observations, and offers the tantalizing possibility of making public alerts prior to the completion of the merger in neutron-star systems.