GW4 - Space GW Detection (LISA) |
Speaker_ |
Pai, Archana |
Co-autors |
Eric Chassande-Mottin |
Talk _ |
Best Chirplet-chain : detecting "unmodelled" GWave chirps |
Abstract _ |
Violent astrophysical events like binary mergers or core collapses are believed to be sources of strong gravitational waves (GW). Such events involve either poorly known or complex physics. It is thus difficult to obtain an accurate model for the emitted GW. This motivates exploratory searches (based on a partial information) as opposed to targeted searches (relying on a precise signal model). In this work, we address the detection problem for a special class of unmodelled GW, namely "unmodelled GW chirps". We propose a `quasi-physical' chirp model -- a chirplet chain (CC). This model is general enough to closely approximate any physically realistic GW chirp. We produce a finite grid of template waveforms in the form of CCs which sample the resulting set of possible chirps and propose a search algorithm -- best CC -- in the time-frequency domain. The 'best CC' algorithm establishes a clear link between the method and an optimality criterion. We also assess the performance, robustness and computational cost of the proposed method with several benchmarks using simulated data. |