Stochastic FDTD/VFETD schemes with MOR for random nanomaterials and graphene nanostructures

A stochastic spectrally-controllable methodology for the precise design of 3-D nanomedia and graphene-based components with statistical parameters is introduced in this paper. The new technique develops a set of reduced-order models for a hybrid finite-difference time-domain/vector finite-element time-domain (FDTD/VFETD) formulation, tailored to comply with the specifications of nano-EMC environments. Moreover, state-space approximations are delivered via Krylov-based Laguerre polynomials to get single-run evaluations of the mean value and standard deviation, unlike the unduly needs of multiple-realization Monte-Carlo FDTD approaches. The hybrid algorithm, accelerated via GPU programming, is verified by realistic nanoscale setups with complex geometries, large electrical sizes, and curved interfaces.