# An improved algorithm for narrow-band searches of continuous gravitational waves

## TL;DR

This paper introduces a significantly faster algorithm for narrow-band searches of continuous gravitational waves from neutron stars, enabling more efficient detection despite uncertainties in rotational parameters.

## Contribution

The authors develop an improved, faster pipeline based on the 5-vectors framework for coherent searches over wider frequency bands and multiple spin-down values.

## Key findings

- Algorithm is about three orders of magnitude faster than previous methods.
- It can perform fully coherent searches over ~1 Hz frequency bands.
- The pipeline runs efficiently on standard workstations within a few hours.

## Abstract

Continuous gravitational waves signals, emitted by asymmetric spinning neutron stars, are among the main targets of current detectors like Advanced LIGO and Virgo. In the case of sources, like pulsars, which rotational parameters are measured through electromagnetic observations, typical searches assume that the gravitational wave frequency is at a given known fixed ratio with respect to the star rotational frequency. For instance, for a neutron star rotating around one of its principal axis of inertia the gravitational signal frequency would be exactly two times the rotational frequency of the star. It is possible, however, that this assumption is wrong. This is why search algorithms able to take into account a possible small mismatch between the gravitational waves frequency and the frequency inferred from electromagnetic observations have been developed. In this paper we present an improved pipeline to perform such narrow-band searches for continuous gravitational waves from neutron stars, about three orders of magnitude faster than previous implementations. The algorithm that we have developed is based on the {\it 5-vectors} framework and is able to perform a fully coherent search over a frequency band of width $\mathcal{O}$(Hertz) and for hundreds of spin-down values running a few hours on a standard workstation. This new algorithm opens the possibility of long coherence time searches for objects which rotational parameters are highly uncertain.