1. Introduction to gravitational waves
The discovery of the first gravitational wave (GW) signal due to the merger of two black holes [1] made by the LIGO
and Virgo collaborations in 2015 propelled the field of GW astronomy into a new era. Since then, several tens of binary
coalescence signals have been detected, including the merger of two black holes with a wide and continuous range of
composite stellar size masses [2,3,4,5]. In addition, evidences for the binary merger of a black hole and a neutron star
[6] and two neutron stars [7,8] have been collected as well. The latter observation was followed up by several observations
in the electromagnetic spectrum [9,10], which kick-started the field of gravitational-wave multi-messenger astronomy.
The signals of all these events, which have characteristic transient features, are of relative short duration (O(1-100 s)),
corresponding to the time during which the signal remains within the sensitive frequency band of current earth bound
interferometers, and have relatively large amplitudes that exceed the intrinsic noise levels of the detectors. Many other
astrophysical sources are predicted to yield detectable GW signals that are either continuous in nature and which typically
originate from asymmetrical rotating compact objects such as pulsars, or are burst-like, such as supernovae type objects.
None of these signals have been detected so far, but could well be in reach of operational and planned Earth-based and
satellite borne GW detectors [11]. Finally, there is also bound to be a stochastic gravitational wave background (SGWB)
in our Universe that can contain several components. The first component is of astrophysical nature and consists of
the random superposition of individually unresolved signals from the entire population of astrophysical sources listed
above. In addition, signals from cosmological events or structures could be present as well [12]. The SGWB is persistent
but can have an intermittent nature, has no phase coherence, and is in several experimental conditions, such as for
unresolved binary coalescences and cosmological signals searched for by current earth based interferometers, buried under
the intrinsic noise level of a single detector. Such a signal has a small but non-negligible contribution to the total energy
content of our Universe. Its detection, and in particular a primordial or cosmological component, would be as significant
as the discovery of the cosmic microwave background (CMB) [13,14]. Its spectral structure will yield information on the
dynamical properties of its contributors and on the cosmological evolution of our Universe, and up to its earliest time
scales, way before the decoupling of the CMB.
In the past decades, many review papers on the SGWB have been published [15,16,17,18,19,20,21], where the
philosophy and explanation of the mathematical framework in this paper has the most overlap with the earlier work in
[19]. In this paper we discuss the properties of the possible components of the SGWB, their theoretical and experimental
bounds, the state-of-the-art of detection techniques and an outlook for future observations. We will refrain from giving an
extensive review of the astrophysical, cosmological and particle physics inspired models for the generation of a SGWB, as
these have been presented in other recent reviews, see e.g. [22,23,20,12,24,25,26]. We try to complement previous work
by giving an update on the latest results and upper limits achieved by Earth-based interferometric gravitational-wave
detectors as well as pulsar timing arrays. The main focus of this paper will be on the analysis techniques that are used or
being investigated for data analysis of Earth-based interferometers. Since we are nearing the first detection of a SGWB
with the continuously increasing sensitivity of the detectors, we will discuss several techniques that can be used to prove
the observed signal is due to GWs and not to a terrestrial or instrumental noise source.
4