Conventional computational electromagnetics toward nanoscale, optical, and plasmonic applications

Summary form only given. Over the last six decades, many computational electromagnetics (CEM) methods have been developed for the analysis and design of communication systems, optoelectronic and biomedical devices, MEMs, and most of the other electronic devices. We have also experienced the tremendous impact of CEM methods in remote sensing, EMC/EMI analysis, chip design, and geophysical prospecting. Even though all of these methods have been solving Maxwell\´s equations fully or approximately, they might differ dramatically from each other depending on their applications. In some scenarios, frequency domain solvers (MoM [1, 2, 3], FEM [2, 3, 4], and FMM [2, 3, 5]) are more efficient then time domain methods (FDTD [6, 7] and PSTD [7, 8]) or vice versa. In some cases implementation of periodic Green\´s functions in an integral equation solver is sufficient for 1 per cent of error whereas sometimes one has to use the fast multipole method to solve that large problem of finite periodicity. No matter which approach is followed for the development of an algorithm, the question is, does this algorithm also work in the optical regime? Or for an algorithm working for microwave, is converting "permittivity" from a real variable to a complex variable enough to handle metamaterials or plasmonic structures? In this talk, I will try to answer these questions and explain how we can push the frontiers of conventional computational electromagnetics toward nanoscale, optical, and plasmonic applications.