Matched filtering of gravitational waves from inspiraling compact binaries: Computational cost and template placement

Physical Review D - Particles, Fields, Gravitation and Cosmology

Volumn 60, Issue 2, 1999, Pages

  Sathyaprakash, B S ^a  

^a CALIFORNIA INSTITUTE OF TECHNOLOGY   (United States)

Author keywords

[No Author keywords available]

Indexed keywords

EID: 17044396821     PISSN: 15507998     EISSN: 15502368     Source Type: Journal    
DOI: 10.1103/PhysRevD.60.022002     Document Type: Article

References (41)

1. A. Abramovici, Science 256, 325 (1992).
2. B. Caron, Class. Quantum Grav. 14, 1461 (1997).
3. H. Lück, Class. Quantum Grav. 14, 1471 (1997).
4. K. Tsubono, Gravitational Wave Experiments (World Scientific, Singapore, 1995), p. 112.
5. R. Narayan, T. Piran, and A. Shemi, Astrophys. J. 379, L17 (1991)
6. Astrophys. J.E. S. Phinney, 380, L17 (1991). In the intervening years astronomers have estimated event rates both lower (based on statistics of the few known binary pulsars) and higher (based on theoretical evolutions of progenitor populations), but their estimates have begun to converge again on the 1991 numbers. See 6 and 7 for recent examples of the high-rate and low-rate camps, respectively.
7. S. F. P. Zwart and H. N. Spreeuw, Astron. Astrophys. 312, 670 (1996).
8. I. H. Stairs, Astrophys. J. 505, 352 (1998).
9. B. Barish, LIGO Advanced Research and Development Proposal, Caltech/MIT, 1996 (unpublished).
10. L. Blanchet, T. Damour, B. R. Iyer, C. M. Will, and A. G. Wiseman, Phys. Rev. Lett. 74, 3515 (1995).
11. H. Tagoshi, M. Shibata, T. Tanaka, and M. Sasaki, Phys. Rev. D 54, 1439 (1996).
12. Some physicists are in the habit of calling this optimal filtering method simply “Wiener filtering.” However, in the larger signal processing community, “Wiener filtering” refers to an altogether different technique. In order to avoid confusion we call this method “matched filtering.”
13. C. W. Helstrom, Statistical Theory of Signal Detection, 2nd edition (Pergamon Press, London, 1968).
14. One might try to avoid using a large family of templates by finding a set of basis functions on which to expand the signal. However, this technique does not work very well for inspiral chirps (and many other types of signals) because the number of basis functions required to achieve nearly optimal signal-to-noise ratio is comparable to the number of templates.
15. T. A. Apostolatos, Phys. Rev. D 52, 605 (1995).
16. S. Droz and E. Poisson, Phys. Rev. D 56, 4449 (1997).
17. T. Damour, B. R. Iyer, and B. S. Sathyaprakash, Phys. Rev. D 57, 885 (1998).
18. R. Balasubramanian and S. V. Dhurandhar, Phys. Rev. D 50, 6080 (1994).
19. A. Królak, K. D. Kokkotas, and G. Schäfer, Phys. Rev. D 52, 2089 (1995).
20. T. A. Apostolatos, Phys. Rev. D 54, 2421 (1996).
21. B. S. Sathyaprakash, Phys. Rev. D 50, R7111 (1994).
22. B. S. Sathyaprakash and S. V. Dhurandhar, Phys. Rev. D 44, 3819 (1991).
23. B. S. Sathyaprakash and S. V. Dhurandhar, J. Comput. Phys. 109, 215 (1993).
24. S. V. Dhurandhar and B. S. Sathyaprakash, Phys. Rev. D 49, 1707 (1994).
25. B. J. Owen, Phys. Rev. D 53, 6749 (1996).
26. B. F. Schutz, in The Detection of Gravitational Radiation, edited by D. Blair (Cambridge University Press, Cambridge, England, 1991), p. 406.
27. S. D. Mohanty and S. V. Dhurandhar, Phys. Rev. D 54, 7108 (1996).
28. S. D. Mohanty, Phys. Rev. D 57, 630 (1998).
29. C. Cutler and É. É. Flanagan, Phys. Rev. D 49, 2658 (1994).
30. R. Balasubramanian, B. S. Sathyaprakash, and S. V. Dhurandhar, Phys. Rev. D 53, 3033 (1996)
31. Phys. Rev. DR. BalasubramanianB. S. SathyaprakashS. V. Dhurandhar54, 1869(E) (1996).
32. The idea of using the Fisher information matrix as a Riemannian metric dates back to the early days of matched filtering and parameter estimation: see N. Wiener, The Extrapolation, Interpolation and Smoothing of Stationary Time Series with Engineering Applications (Wiley, New York, 1949)
33. and C. R. Rao, Bull. Calcutta Math. Soc. 37, 81 (1945).
34. E. Poisson and C. M. Will, Phys. Rev. D 52, 848 (1995).
35. Strictly speaking, the upper limit of integration in the noise moments should be the lesser of the sampling frequency and the “coalescence frequency” (Formula presented) which represents (roughly) the point where adiabatic inspiral of the bodies ends. Both of these frequencies are much greater than (Formula presented) and thus our approximation is quite good, except on the high-mass end of template parameter space where (Formula presented) becomes comparable to (Formula presented). Since most templates populate the low-mass end of parameter space, the effect on the total number of templates is small. However, detailed algorithms for placing high-mass templates will have to take (Formula presented) into account—which may be problematic since (Formula presented) is poorly known when the binary components have comparable masses. Using an infinite upper frequency is safe in the sense that the errors cause us to space templates slightly more closely than needed, i.e., we slightly overestimate the loss of signal-to-noise ratio.
36. P. C. Peters, Phys. Rev. 136, B1224 (1964).
37. S. Droz, D. J. Knapp, E. Poisson, and B. J. Owen, Phys. Rev. D 59, 124016 (1999).
38. B. Allen et al., GRASP software package, available from http://www.lsc-group.phys.uwm.edu/.
39. B. J. Owen, Phys. Rev. D 58, 084020 (1998).
40. T. Nakamura, M. Sasaki, T. Tanaka, and K. S. Thorne, Astrophys. J. 487, L139 (1997).
41. J. H. Taylor, Rev. Mod. Phys. 66, 711 (1994).