Abstract
With advances in system integration and packaging, the capabilities of hand-held devices and embedded bio-sensors have grown to a phenomenal scale. This in turn has led to a constant interaction between human beings and ambient electromagnetic waves. Hence there is a need for studying the effects of radiation on human physiology and also the performance of systems in such an environment. The system designers seek a full-wave solution of the entire system, taking into account a variety of environments in which it operates. However, attempts to do this substantially increase the complexities involved in computing full-wave solutions, especially when the problems involve multi-scale geometries with very fine features. For such problems, even the well-established numerical methods, such as the time domain technique finite difference time domain and the frequency domain techniques, e.g. the finite element method and the method of moments, are often challenged to the limits of their capabilities. In an attempt to address these challenges, we propose to handle the multiscale problems in three different ways, based on the dimension and the complexity of the fine features involved in the problem. Furthermore, we illustrate the efficacy of the above techniques via several examples, and the results obtained by the proposed techniques are compared with other existing numerical methods for the purpose of validation.