Accelerating the discrete dipole approximation by initializing with a scalar solution and using a circulant preconditioning. (April 2023)

• Record Type: Journal Article
• Authors: Chaumet, Patrick C.; Maire, Guillaume; Sentenac, Anne
• Is Part Of: Journal of quantitative spectroscopy & radiative transfer. Volume 298(2023)
• Publication Date: 2023-04
• Subjects: Scattering -- Discrete dipole approximation -- Iterative method -- Preconditioner; Spectrum analysis -- Periodicals; Radiation -- Periodicals; Analyse spectrale -- Périodiques; Rayonnement -- Périodiques; Radiation; Spectrum analysis; Periodicals; 543.0858
• DOI: 10.1016/j.jqsrt.2023.108505
• Languages: English
• ISSNs: 0022-4073

Abstract

Abstract: The discrete dipole approximation (DDA) is a method of choice for simulating the electromagnetic scattering by objects of arbitrary shape and permittivity. To recover the field inside the object, it requires the iterative solving of a dense linear system which can be time consuming. To ease this task, we propose to start the inversion with the solution of the recently introduced scalar approximation [Chaumet et al. J. Opt. Soc. Am. A, 39, 1462 (2022)]. This initial guess allows a reduction of the time required for the solving of the linear system up to 50%. In addition, we study the interest of preconditioning the system to accelerate convergence. We show that the gain can be up to a factor of 5, especially for homogeneous objects on a plane substrate.