A fast and scalable algorithm for the Monte Carlo simulation of elastic scattering in perturbed media

The simulation of light propagation in perturbed media is important in a variety of applications like atmospheric science, biomedical optics or lithographic optical systems. Based on the theory of elastic scattering on spheroidal particles (G. Mie, 1908) we have developed a Monte Carlo algorithm for the simulation of multiple scattering. In contrast to deterministic ray tracing our approach fully considers the statistical nature of scattering on a microscopic scale. For optical systems this is computationally expensive and requires the usage of parallel processors to reduce execution time. Using a high precision testbed for scattering measurements we have experimentally verified our approach of simulating light propagation in perturbed media. We have further shown that our parallel programming model is physically justified. The approach is scalable with respect to the number of processors available: without re-writing code more extensive Monte Carlo simulations can be computed. This is of special importance if a high statistical accuracy of the results is required.