A 3-D Polynomial-Chaos FDTD Technique for Complex Inhomogeneous Media with Arbitrary Stochastically-Varying Index Gradients

摘要

An enhanced finite-difference time-domain algorithm featuring the polynomial chaos representation is introduced in this paper for problems with stochastic uncertainties. Focusing on the solution of the governing partial differential equations, the new 3-D method uses the Karhunen-Loeve expansion to effectively decorrelate random input parameters denoted by stochastic processes. So, the space dimension is seriously reduced and high accuracy levels are attained, even for media with abrupt and fully unknown statistical variations. These profits are verified via a detailed numerical study.