IIR Adaptive Filters for Detection of Gravitational Waves from Coalescing Binaries

TL;DR

This paper introduces a hierarchical detection strategy for gravitational waves from coalescing binaries using IIR Adaptive Line Enhancer filters as triggers, demonstrating promising results with low computational cost and robustness.

Contribution

The paper proposes a novel hierarchical detection method employing IIR ALE filters as triggers, improving efficiency and robustness in gravitational wave detection.

Findings

ALE filters are effective as triggers with low computational requirements

A dynamic bank of ALE filters improves detection robustness

The method is suitable for real-time implementation in current interferometers

Abstract

In this paper we propose a new strategy for gravitational waves detection from coalescing binaries, using IIR Adaptive Line Enhancer (ALE) filters. This strategy is a classical hierarchical strategy in which the ALE filters have the role of triggers, used to select data chunks which may contain gravitational events, to be further analyzed with more refined optimal techniques, like the the classical Matched Filter Technique. After a direct comparison of the performances of ALE filters with the Wiener-Komolgoroff optimum filters (matched filters), necessary to discuss their performance and to evaluate the statistical limitation in their use as triggers, we performed a series of tests, demonstrating that these filters are quite promising both for the relatively small computational power needed and for the robustness of the algorithms used. The performed tests have shown a weak point of ALE filters, that we fixed by introducing a further strategy, based on a dynamic bank of ALE filters, running simultaneously, but started after fixed delay times. The results of this global trigger strategy seems to be very promising, and can be already used in the present interferometers, since it has the great advantage of requiring a quite small computational power and can easily run in real-time, in parallel with other data analysis algorithms.