New Monte Carlo Integration Models for Underwater Wireless Optical Communication
Ruqin Xiao, Pierre Combeau, Lilian Aveneau
Abstract
Underwater wireless optical communications are knowing a high interest in recent years, with many military and civil applications. The design of efficient systems relies on powerful simulation tools. The most used simulation algorithms are based on a paper from 1989, written by Prahl and designed for light propagation through human tissue. This method relies on Monte Carlo Simulation, following the path of a photon in the participating media. As such, it is difficult to propose evolution or optimization. In addition, a large amount of photons are needed to obtain good results with low error, <i>e.g</i>. for turbid water. With Monte Carlo Integration method, there exist a lot of optimization techniques to reduce the variance. This paper proposes a mathematical formalization of the propagation of light in water or any other participating medium as an integral problem. Therefore, it opens the way for a large number of future optimizations. A very straightforward and simple variance reduction technique is proposed as an example. Our simulation results show that this new technique has a lower sample variance as expected, and thus better convergence rates.