Hearing Protection for High-Noise Environments. Attachment 4: Parallelization of the Acoustic Integral-Equation Solver and Example Applications
Abstract
We have recently developed a fast iterative volumetric integral-equation solver for acoustic scattering and propagation problems involving inhomogeneous media. The solver incorporates a Fast Fourier Transform (FFT)-based matrix compression technique and the corresponding fast matrix-vector multiplication algorithm. The method consists, essentially, of the method-of-moments (MoM) computation of the near-field interactions involving physical sources and fields, and evaluation of the far-field by means of the FFTs operating on equivalent source and field distributions defined on nodes of a regular Cartesian grid. The FFT-based matrix compression is particularly well suited to volumetric discretization (with a fairly uniform discretization) and to sub-wavelength problems, i.e., spatial resolution scale much smaller than the wavelength. Both of these features are characteristic of realistic anatomical models and propagation problems in the biological applications we are considering.
Document Details
- Document Type
- Technical Report
- Publication Date
- Nov 30, 2009
- Accession Number
- ADA517903