Electromagnetically characterizing disordered systems: Composite material design and foliage penetrating radars

Effective electromagnetic representations of composite systems provide a useful simplification in many analyses. As the number of applications of wireless communication swells so does the need to characterize the individual systems and their coupling into and propagation through a variety of environments. This dissertation explores the utility of effective permittivity and permeability representations to simplify the electromagnetic analysis of both ordered and disordered composite systems. The necessity of an accurate account of a system's structure within an effective representations, and the need for consistency within different approaches for the same structure is shown. First, quasi-static effective representations are described, and their natural suitability is shown to grouping in terms of the connectivity of the structure they describe. Then, a consistent perturbational account of scattering within any quasi-static effective permittivity is derived. In going beyond the quasi-static limit the interplay between electric and; magnetic fields within the structure being described becomes increasingly important. To demonstrate this the inclusion of electric and magnetic resonances in an ordered composite structures are demonstrated for an ordered array of dielectric spheres, and an ordered array of conducting disks. The utility of the non-static effective simplification is then demonstrated by solving for the near and far field scattered from an electric line source antenna radiating over the photonic bang gap crystal made up of arrays of the disks. In all these effective formulations the necessity of accounting for a composite's structure is emphasized, and the general characteristics differentiating similar structures within each theory are explored.; After various types of effective representation are demonstrated an effective model of a forest is created at UHF and lower frequencies. This model uses the quasi-static, perturbationally non-static, and excitation dependent functional effective descriptions to consistently simplify the stochastic forest structure. A realistic forest structure is obtained using Lindenmayer algorithms and measured tree parameters and then used to create a layered effective model. In the framework of this model coupling into and propagation through the forest are investigated, and the effects of each layered component are explored. This dissertation explores the applicability of effective representations of composite systems for simplification, and demonstrates the use of that simplification in describing electromagnetic propagation within foliage in order to facilitate the design of foliage penetrating radars.