On Higher Order Imperative in Computational Electromagnetics

Methods of computational electromagnetics (CEM) such as Finite-Difference-Time-Domain (FDTD), Finite-Element-Method (FEM), Method of Moments (MoM), and others are widely used for design of today's wireless systems, analysis of biological effects of electromagnetic (EM) fields, prospecting for natural resources, and in various other important areas. While in most cases each of the above methods can be used for solution of a particular problem, the choice of a specific discretization approach is generally stipulated by a variety of factors. These factors include the level of required accuracy in approximation of EM fields, electrical size of the model, its material properties, and type of computational resources available for solution of the problem, among others. It can be shown that solution of the same problem can be achieved exponentially faster when the best method is used as opposed to its alternatives [1].