Method for all-sky searches of continuous gravitational wave signals using the frequency-Hough transform

TL;DR

This paper introduces a hierarchical pipeline using the frequency-Hough transform for all-sky searches of continuous gravitational waves, improving sensitivity and candidate rejection over existing methods.

Contribution

It presents a novel hierarchical data analysis pipeline with an efficient frequency-Hough transform for all-sky gravitational wave searches, enhancing sensitivity and false candidate rejection.

Findings

The pipeline achieves competitive sensitivity curves.

The frequency-Hough transform improves computational efficiency.

Enhanced candidate selection and data cleaning methods.

Abstract

In this paper we present a hierarchical data analysis pipeline for all-sky searches of continuous gravitational wave signals, like those emitted by spinning neutron stars asymmetric with respect to the rotation axis, with unknown position, rotational frequency and spin-down. The core of the pipeline is an incoherent step based on a particularly efficient implementation of the Hough transform, that we call frequency-Hough, that maps the data time-frequency plane to the source frequency/spin-down plane for each fixed direction in the sky. Theoretical ROCs and sensitivity curves are computed and the dependency on various thresholds is discussed. A comparison of the sensitivity loss with respect to an "optimal" method is also presented. Several other novelties, with respect to other wide-parameter analysis pipelines, are also outlined. They concern, in particular, the construction of the grid in the parameter space, with over-resolution in frequency and parameter refinement, candidate selection and various data cleaning steps which are introduced to improve search sensitivity and rejection of false candidates.