Detecting the Cosmological Stochastic Background of Gravitational Waves with FastICA

TL;DR

This paper explores how the early universe's gravitational wave background depends on gravity theories and proposes a neural network method to enhance detection sensitivity with interferometers like VIRGO.

Contribution

It introduces a neural network approach leveraging non-Gaussian features to improve detection of the cosmological stochastic gravitational wave background.

Findings

Detection sensitivity can be significantly improved using the proposed neural network method.

The stochastic background depends on the form of the f(R) gravity theory.

Neural network algorithms can identify non-Gaussian signals in gravitational wave data.

Abstract

We show that the stochastic background of gravitational waves, produced in the early cosmological epochs, strictly depends on the assumed theory of gravity. In particular, the specific form of the function f(R), where R is the Ricci scalar, is related to the evolution and the production mechanism of gravitational waves. Using a neural network algorithm which only requires non-Gaussian nature and independence of the input signals we conclude that, in order to detect a CSB signal, the interferometric sensitivity of detector like VIRGO will be improved.