Mock data challenge for the Einstein Gravitational-Wave Telescope

The Einstein Telescope (ET) is conceived to be a third generation gravitational-wave (GW) observatory. Its amplitude sensitivity would be a factor 10 better than advanced LIGO and Virgo and it could also extend the low-frequency sensitivity down to 1--3 Hz, compared to the 10--20 Hz of advanced detectors. Such an observatory will have the potential to observe a variety of different GW sources, including compact binary systems at cosmological distances. ET's expected reach for binary neutron star (BNS) coalescences is out to redshift $z\ensuremath{\simeq}2$ and the rate of detectable BNS coalescences could be as high as one every few tens or hundreds of seconds, each lasting up to several days. With such a signal-rich environment, a key question in data analysis is whether overlapping signals can be discriminated. In this paper we simulate the GW signals from a cosmological population of BNS and ask the following questions: Does this population create a confusion background that limits ET's ability to detect foreground sources? How efficient are current algorithms in discriminating overlapping BNS signals? Is it possible to discern the presence of a population of signals in the data by cross correlating data from different detectors in the ET observatory? We find that algorithms currently used to analyze LIGO and Virgo data are already powerful enough to detect the sources expected in ET, but new algorithms are required to fully exploit ET data.