Addressing the challenges of detecting time-overlapping compact binary coalescences

TL;DR

This paper investigates how current gravitational wave detection methods perform when multiple signals overlap in time, revealing their capabilities and limitations in identifying and distinguishing such signals.

Contribution

It provides an analysis of the effectiveness of existing search techniques on overlapping signals and suggests modifications for improved detection accuracy.

Findings

Both searches can identify signals separated by at least 1 second.

Clustering routines limit detection of overlapping signals within 1 second.

Modified matched filter searches can estimate parameters of multiple signals.

Abstract

Standard detection and analysis techniques for transient gravitational waves make the assumption that detector data contains, at most, one signal at any time. As detectors improve in sensitivity, this assumption will no longer be valid. In this paper we examine how current search techniques for transient gravitational waves will behave under the presence of more than one signal. We perform searches on data sets containing time-overlapping compact binary coalescences. This includes a modelled, matched filter search (PyCBC), and an unmodelled coherent search, coherent WaveBurst (cWB). Both of these searches are used by the LIGO-Virgo-KAGRA collaboration. We find that both searches are capable of identifying both signals correctly when the signals are dissimilar in merger time, $|\Delta t_c| \geq 1$ second, with PyCBC losing only $1\%$ of signals for overlapping binary black hole mergers. Both pipelines can find signal pairings within the region $|\Delta t_c| < 1$ second. However, clustering routines in the pipelines will cause only one of the two signals to be recovered, as such the efficiency is reduced. Within this region, we find that cWB can identify both signals. We also find that matched filter searches can be modified to provide estimates of the correct parameters for each signal.