Analytical theory and numerical approaches to the simulation and optimal design of multilayer mirrors and optical Bragg waveguides are developed. Optimal refractive index profiles in planar and cylindrical geometry are found. Reduction of the full set of Maxwell's equations to a scalar boundary value problem in the lowcontrast approximation is discussed. An efficient finite-difference scheme is used to simulate realistic Bragg fibers.

The aim of this workis to increase the instability of the marching-on-in-time (MOT) method that is used in the analysis of wide-band electromagnetic pulse scattering from structures made of thin wires. The stability problem has been identified with the advent of the MOT method in 1991, and although several improvements have been suggested to overcome this difficulty no exact solution has been found [1]. In this thesis two methods (the Newmark-Beta formulation and the analytic integration) to suppress the instabilities of the MOT algorithm for thin wire scatterers have been proposed and their effects on the stability have been inspected. The results are compared with the results obtained with time domain method of moments (MOM) [2] and it is observed that the results are both stable and accurate. It is shown how the stability is changed by a determined β parameter. Also, Newmark-Beta formulation results for selected different types of structures such as dipole antenna illuminated by a Gaussian pulse given in [3], three pole structure given in [4], loop antenna given in [3] has been shown.

In the present work the radiation of sound from a bifurcated circular waveguide formed by a semi-infinite rigid duct inserted axially into a larger infinite tube with discontinuous wall impedance is reconsidered through an alternative approach which consists of using the mode matching technique in conjunction with the Wiener-Hopf method. By expressing the total field in the appropriate waveguide region in terms of normal modes and using the Fourier transform technique elsewhere, we end up with a single modified Wiener-Hopf equation whose solution involves an infinite system of algebraic equations. This system is solved numerically and the influence of some parameters on the radiation phenomenon is shown graphically. The equivalence of the direct method described in [1] and the present mixed method are shown numerically.

The diffraction by a finite parallel-plate waveguide with four-layer material loading is rigorously analyzed by means of the Wiener-Hopf technique for the H-polarized plane wave incidence. Taking the Fourier transform for the unknown scattered field as well as the Helmholtz equation and applying boundary conditions in the transform domain, the problem is formulated in terms of the simultaneous Wiener-Hopf equations. The Wiener-Hopf equations are solved via the factorization and decomposition procedure together with the use of rigorous approximation procedures leading to an efficient approximation solution. The scattered field in the real space is evaluated explicitly by taking the inverse Fourier transform. Illustrative numerical examples on the radar cross section (RCS) are presented and the far field scattering characteristics of the waveguide are discussed.

In this paper, a method based on clonal selection algorithm (CLONALG) is proposed for null steering of linear antenna arrays by controlling only the positions of selected elements. The CLONALG is a relatively novel population-based evolutionary algorithm inspired by the clonal selection principle of the human immune system. In order to illustrate the accuracy and flexibility of the proposed algorithm, several numerical examples of Chebyshev pattern with the single and double nulls imposed at the directions of interference are given.

This work reports some recent advances in diffraction theory by canonical shapes like wedges or cones with impedancetype boundary conditions. Our basic aim in the present paper is to demonstrate that functional difference equations of the second order deliver a very natural and efficient tool to study such a kind of problems (For a thorough and up-to-date overview of the scattering and diffraction in general the readers are referred to a special section of the journal ``Radio Science'' edited by Uslenghi [1].). To this end we consider two problems: diffraction of a normally incident plane electromagnetic wave by an impedance wedge whose exterior is divided into two parts by a semi-infinite impedance sheet and diffraction of a plane acoustic wave by a right-circular impedance cone. In both cases the problems can be formulated in a traditional fashion as boundary-value problems of the scattering theory. For the first problem the Sommerfeld-Malyuzhinets technique enables one to reduce it to a problem for a vectorial system of functional Malyuzhinets equations. Then the system is transformed to uncoupled second-order functional difference-equations (SOFDE) for each of the unknown spectra. In the second problem the incomplete separation of variables leads directly to a functional difference-equation of the second order. Hence, it is remarkable that in both cases the key mathematical tool is an SOFDE which is an analog of a second-order differential equation with variable coefficients. The latter is reducible to an integral equation which is known to be the most traditional tool for its solution. It has recently been recognised that reducing SOFDEs to integral equations is also one of the most efficient approaches for their study. The integral equations which are developed for the problems at hand are both of the second kind and obey Fredholm property. In the problem of diffraction by a wedge the generalised Malyuzhinets function is exploited on the preliminary step then ``inversion'' of a simple difference operator with constant coefficients leads to an integral equation of the second kind. The corresponding integral operator is represented as a sum of the identical operator and a compact one [2]. However, in the second problem the situation is slightly different: the integral operator can be represented by a sum of the boundedlyinvertible (Dixon's operator) and compact operators. This situation was earlier considered by Bernard in his study of diffraction by an impedance cone, and important advances have been made (see [3-6]). The Fredholm property is crucial for the elaboration of different numerical schemes. In our cases we exploited direct numerical approaches based on the quadrature formulae and computed the farfield asymptotics for the problems at hand. Various numerical results are demonstrated and discussed.

A method based on adaptive neuro-fuzzy inference system (ANFIS) for computing the effective permittivity and the characteristic impedance of the micro-coplanar strip (MCS) line is presented. The ANFIS is a class of adaptive networks which are functionally equivalent to fuzzy inference systems (FISs). A hybrid learning algorithm, which combines the least square method and the backpropagation algorithm, is used to identify the parameters of ANFIS. The effective permittivity and the characteristic impedance results obtained by using ANFIS are in good agreement with the theoretical and experimental results reported elsewhere.

The societal and technological priorities of the world have been continuously changing because of complex computer and technology-driven developments in everywhere like communications, health, defense, economy, etc. Electromagnetic (EM) systems have become more and more complex however the explosive growth of computer capabilities has revolutionized the design and analysis of such complex systems. This has made interdisciplinary exposure necessary in modern EM engineering, also has brought up discussions of educational challenges and novel teaching approaches that confront wave-oriented EM engineering in the 21st century. This paper reviews EM computer simulation strategies and summarizes novel virtual tools that may be used in connection with classical EM lectures as well as with a newly proposed EM Engineering Program.

Excitation of the electromagnetic fields by a wide-band current surge, which has a beginning in time, is studied in a cavity bounded by a closed perfectly conducting surface. The cavity is filled with Debye or Lorentz dispersive medium. The fields are presented as the modal expansion in terms of the solenoidal and irrotational cavity modes with the time-dependent modal amplitudes, which should be found. Completeness of this form of solution has been proved earlier. The systems of ordinary differential equations with time derivative for the modal amplitudes are derived and solved explicitly under the initial conditions and in compliance with the causality principle. The solutions are obtained in the form of simple convolution (with respect to time variable) integrals. Numerical examples are exhibited as well.

This article presents a new approach based on artificial neural networks (ANNs) to calculate the characteristic parameters of elliptic and circular-shaped microshield lines. Six learning algorithms, bayesian regularization (BR), Levenberg-Marquardt (LM), quasi-Newton (QN), scaled conjugate gradient (SCG), resilient propagation (RP), and conjugate gradient of Fletcher-Reeves (CGF), are used to train the ANNs. The neural results are in very good agreement with the results reported elsewhere. When the performances of neural models are compared with each other, the best and worst results are obtained from the ANNs trained by the BR and CGF algorithms, respectively.

The aim of the present paper is to reveal the effect of motion on the scattering by an edge. To this end one considers a canonical structure formed by a perfectly conducting half-plane illuminated by a time-harmonic and uniformly moving infinitely long line source. The relevant line source is located parallel to the edge and moves with a constant velocity which is also parallel to the half-plane. This is the dual of a previously studied problem in which the halfplane was moving uniformly. The present problem is first reduced into a Wiener-Hopf problem in the sense of distribution and then solved by an ad-hoc method. The edge-diffracted field is discussed in detail.

After a short presentation of the boundary layer method extended to strongly elongated objects by Andronov and Bouche [1], the author develops some techniques for deriving explicit formulas for the asymptotic currents on a strongly elongated object of revolution excited by an electromagnetic plane wave propagating in the paraxial direction. The performance of the different techniques are demonstrated by comparing numerical results obtained for the asymptotic currents on an elongated prolate ellipsoid with those obtained by solving the EFIE.

Ray tracing algorithms rely on two dimensional or three dimensional database. They use ray optical techniques referred to as the uniform theory of diffraction (UTD) using building database given as polygons. Building geometries can also be modelled as having non-planar geometries, and this would be important in modeling of shadowing loss due to curved structures in urban radio propagation. To demonstrate modelling of buildings as non-polygonal geometries, a particular building composition involving 3D cruved geometries is chosen, and shadowing loss for this building composition is studied via UTD ray tracing. Building structure considered in this study involves main canonical shapes of non-planar geometries including cone, cylinder and sphere. Single and multiple interaction of surface diffractions, effect of creeping waves are taken into consideration in the analysis.

It is well known in B-scan ground penetrating radar (GPR) imagery that the underground scatterers generally exhibit defocused, hyperbolic characteristics. This is mainly due to the data collection scheme and the finite beam width of the main lobe of the GPR antenna. To invert this undesirable effect and obtain focused images, various migration or focusing algorithms have been developed. In this paper, we survey the performance of our recent focusing algorithms, namely; hyperbolic summation (HS) and frequency-wavenumber (w-k) based synthetic aperture radar (SAR) focusing. The practical usage of these focusing methods were tested and examined on both simulated and measured GPR data of various buried targets. The simulation data set is obtained by a physical optics shooting and bouncing ray (PO-SBR) technique code. Measurements were taken by a stepped frequency continuous wave (SFCW) radar set-up. Scattered C-band field data were measured from a laboratory sand box and from outdoor soil environment. The proposed focusing methods were then applied to the B-scan GPR images to enhance the resolution quality within these images. The resultant GPR images obtained with the proposed algorithms demonstrate enhanced lateral resolutions.

In this work a new method based on the adaptive neuro-fuzzy inference system (ANFIS) was successfully introduced to determine the characteristic parameters, effective permittivities and characteristic impedances, of conventional coplanar waveguides. The ANFIS has the advantages of expert knowledge of fuzzy inference system and learning capability of neural networks. A hybrid-learning algorithm, which combines least-square method and backpropagation algorithm, is used to identify the parameters of ANFIS. There are very good agreement between the results of ANFIS models, experimental works, conformal mapping technique, spectral domain approach and a commercial electromagnetic simulator, MMICTL.

The scattering of electromagnetic waves from a Frequency Selective Surfaces (FSSs) composed of a new one- and two-turn square spiral shaped periodic structures are investigated by using modal expansion method for a linearly polarized transverse electric (TE) and transverse magnetic (TM) incident waves. The Moment Method (MM) of Galerkin type is employed by expressing the current induced on the metallic surfaces in terms of Piecewise Sinusoidal (PWS) basis functions to determine the FSS structure reflection and transmission coefficients.

Tunable frequency selective surfaces (FSSs) based on split ring resonators (SRRs) are presented. Tuning performance is achieved by means of several on/off switches placed between the rings of each SRR element. The band-stop FSS response is dynamically tuned to different frequency bands at different switching states. In addition, loadings placed at the corners of outer ring elements, forming a fan-like shape, with additional switches are shown to offer rather fine-tuning capability. A dual-layer FSS is also introduced to demonstrate a filter response over a larger frequency band, and also offers tunable dualband operation via switching. By using complementary SRR elements, a tunable band-pass response instead can be obtained using a similar switching configuration. Practical switch modeling is also examined in the paper along with the scanning performance of the SRR-FSS. The numerical analysis of the FSS designs is accomplished using a fast periodic array simulator, and the measurements demonstrate preliminary validation of the proposed switching configuration.

Computational aspects of EM pulse propagation along the nonuniform earth surface are considered. For ultrawide-band pulses without carrier, the exact wave equation in a narrow vicinity of the wave front is reduced to a time-domain version of the Leontovich- Fock parabolic equation. To solve it by finite differences, we introduce a time-domain analog of the impedance BC and a nonlocal BC of transparency. Numerical examples are given to demonstrate the influence of soil conductivity on the received pulse waveform. For a high-frequency modulated EM pulse, we develop an asymptotic approach based on the ray structure of the monochromatic wave field calculated at the carrier frequency. As an example, a problem of target altitude determination from overland radar data is considered.

The plane wave diffraction by a finite parallel-plate waveguide with four-layer material loading is rigorously analyzed for the case of E polarization using the Wiener-Hopf technique. Introducing the Fourier transform for the scattered field and applying boundary conditions in the transform domain, the problem is formulated in terms of the simultaneous Wiener-Hopf equations, which are solved via the factorization and decomposition procedure together with a rigorous asymptotics. The scattered field is evaluated explicitly by taking the inverse Fourier transform and applying the saddle point method. Representative numerical examples of the radar cross section (RCS) are shown for various physical parameters and the far field scattering characteristics are discussed in detail.