J-transform applied to the detection of Gravitational Waves: preliminary results

TL;DR

This paper introduces a novel spectral analysis method using J-transform for detecting gravitational waves, effectively separating signal from noise in interferometric data, and demonstrates its success on simulated low SNR data.

Contribution

The paper presents a new J-transform based spectral analysis technique for gravitational wave detection, capable of lossless undersampling and effective noise separation.

Findings

Successfully applied to simulated data with low signal-to-noise ratio.

Effectively separates signal and noise in the complex plane.

Demonstrates potential for real gravitational wave data analysis.

Abstract

We propose to apply to the detection of Gravitational Waves a new method developed for the spectral analysis of noisy time-series of damped oscillators. From the Pad\'e Approximations of the time-series Z-transform, a Jacobi Matrix (J-Matrix) is constructed. We show that the J-Matrix has bound states with eigenvalues strictly inside the unit circle. Each bound state can be identified with one precise damped oscillator. Beside these bound states, there is an essential spectrum sitting on the unit circle which represents the noise. In this picture, signal and noise are clearly separated and identified in the complex plane. Furthermore, we show that the J-transform enjoys the exceptional feature of lossless undersampling. We take advantage of the above properties of the J-transform to develop a procedure for the search of Gravitational Wave bursts in interferometric data series such as those of LIGO and VIRGO projects. Successful application of our procedure to simulated data having a poor signal to noise ratio, highlights the power of our method.