A wavelet-convolution neural network distinguishes simulated lensed from unlensed gravitational waves with 92.2% accuracy (AUC 0.965) using wave-optics diffraction patterns.
Probing compact dark matter with gravitational wave fringes detected by the Einstein Telescope
1 Pith paper cite this work. Polarity classification is still indexing.
abstract
Unlike the electromagnetic radiation from astrophysical objects, gravitational waves (GWs) from binary star mergers have much longer wavelengths and are coherent. For ground-based GW detectors, when the lens object between the source and the earth has mass $\sim 1-10^5M_\odot$, the diffraction effect should be considered since the chirping wavelengths are comparable to the scale of the barrier (its Schwarzschild radius). The waveform will thus be distorted as the fringes. In this work, we show that signals from the third-generation GW detectors like the Einstein Telescope (ET) would be a smoking gun for probing the nature of compact dark matter (CDM) or primordial black holes. Detection of the lensing effects becomes harder when the lens mass is smaller. ET is more sensitive than LIGO, the constraint is available for CDM mass $>5M_\odot$ while LIGO can only detect the mass $>100M_\odot$. For a null search of the fringes, one-year observation of ET can constrain the CDM density fraction to $\sim10^{-2}-10^{-5}$ in the mass range $M_{\rm{CDM}}=10M_\odot-100M_\odot$.
citation-role summary
citation-polarity summary
fields
astro-ph.IM 1years
2025 1verdicts
CONDITIONAL 1roles
background 1polarities
unclear 1representative citing papers
citing papers explorer
-
Identifying Microlensing by Compact Dark Matter through Diffraction Patterns in Gravitational Waves with Machine Learning
A wavelet-convolution neural network distinguishes simulated lensed from unlensed gravitational waves with 92.2% accuracy (AUC 0.965) using wave-optics diffraction patterns.