Routing (electronic design automation)

Quantum mechanics

Algorithm

Bayesian probability

Astrophysics

LIGO

Computer science

Statistical physics

Computer network

Binary black hole

Routing protocol

Physics

Amplitude

Mathematics

Link-state routing protocol

Artificial intelligence

Gravitational wave

Black hole (networking)

Binary number

Bayesian inference

Arithmetic

Physical review. D/Physical review. D.

The space-based gravitational wave detector LISA will observe mergers of massive black hole binary systems (MBHBs) to cosmological distances, as well as inspiralling stellar-origin (or stellar-mass) binaries (SBHBs) years before they enter the LIGO/Virgo band. Much remains to be explored for the parameter recovery of both classes of systems. Previous MBHB analyses relied on inspiral-only signals and/or a simplified Fisher matrix analysis, while SBHBs have not yet been extensively analyzed with Bayesian methods. We accelerate likelihood computations by (i) using a Fourier-domain response of the LISA instrument, (ii) using a reduced-order model for non-spinning waveforms that include a merger-ringdown and higher harmonics, (iii) setting the noise realization to zero and computing overlaps in the amplitude/phase representation. We present the first simulations of Bayesian inference for the parameters of massive black hole systems including consistently the merger and ringdown of the signal, as well as higher harmonics. We clarify the roles of LISA response time and frequency dependencies in breaking degeneracies and illustrate how degeneracy breaking unfolds over time. We also find that restricting the merger-dominated signal to its dominant harmonic can make the extrinsic likelihood very degenerate. Including higher harmonics proves to be crucial to break degeneracies and considerably improves the localization of the source, with a surviving bimodality in the sky position. We also present simulations of Bayesian inference for the extrinsic parameters of SBHBs, and show that although unimodal, their posterior distributions can have non-Gaussian features.

Exploring the Bayesian parameter estimation of binary black holes with LISA