We generalize to scalar pulses with finite duration a previous work [1] in which a new approach to diffraction at plane apertures is developed for scalar harmonic waves. A particular attention is given to rectangular pulse modulated signals for which an exact solution to the diffraction problem is obtained. As an example, the diffraction of a truncated harmonic pulse is investigated and the numerical problems to be solved are discussed with an important simplification when one is only interested in the diffraction pattern far from the aperture. More works are needed for apertures with no simple geometrical form.

Auseful and flexible method based on the tabu search algorithm for the pattern synthesis of linear antenna arrays with the prescribed nulls is presented. Nulling of the pattern is achieved by controlling the amplitude-only and both the amplitude and phase of each array element. To show the versatility of the present method, some design specifications such as the sidelobe level, the null depth and the dynamic range ratio are considered by introducing a set of weighting factors in the cost function constructed for the tabu search algorithm. Several illustrative examples of Chebyshev pattern with the imposed single, multiple and broad nulls are given.

A complete set of three dimensional multilayered media Green's functions is presentedfor general electric andmagnetic sources. The Green's functions are derived in the mixed potential form, which is identical with the Michalski-Zheng C-formulation. The approach appliedin this paper is basedon the classical Hertz potential representation. A special emphasis is on the formulation of the dyadic Green's functions G^HJ and G^EM. In these functions the derivatives due to the curl operator are taken in the spectral domain. This avoids the needof the numerical differentiation. Furthermore, it is foundthat the Hertzian potentials satisfy several useful duality and reciprocity relations. By these relations the computational efficiency of the Hertz potential approach can be significantly improvedandthe number of requiredSommerfeldin tegrals can be essentially reduced. We show that all spectral domain Green's functions can be obtained from only two spectral domain Hertzian potentials, which correspond to the TE component of a vertical magnetic dipole and the TM component of a vertical electric dipole. The derived formulas are verified by numerical examples.

The variational principle is employed to obtain the parameters dynamics of Gaussian and super-Gaussian chirped solitons which propagates through birefringent optical fibers that is governed by the dispersion-managed vector nonlinear Schrödinger's equation. The waveform deviates from that of a classical soliton. The periodically changing strong chirp of such a soliton reduces the effective nonlinearity that is necessary for balancing the local dispersion. This study is extended to obtain the adiabatic evolution of the parameters of such a soliton in presence of perturbation terms.

A numerical study of split cylindrical dielectric resonator antennas on a conducting ground plane excited by a coaxial probe is presented. The numerical solution is based on the method of moments for a body of revolution coupled to a wire. We consider in this study bandwidth enhancement for dielectric resonators excited in the HEM₁₁ and HEM₁₂ modes for the split dielectric cylinder. A wideband performance of about 35% has been achieved for the antenna and experimental measurements have verified this finding.

A novel definition of pulse propagation velocity is introduced. It is shown that the present definition does not lead to confusing results such as complex velocity or velocity exceeding the light velocity in the vacuum. Also shown are the parallels of this definition to the classical and quantumm echanics conceptions. Using the present definition reveals certain analogies between electromagnetic pulse propagation in the classical physics and de Broglie wave-packets propagation in the quantumm echanics, thus adding support to its validity.

In this paper a statistical model of the excitation of a conjugate-matched two-wire transmission line in a lossy half space by an electromagnetic (EM) wave is developed. The EM field, radiating in the air, is obliquely incident to the interface defined by the lossy medium and air. Three different orientations of the transmission line for horizontal and vertical polarization of the EM field are examined. The objective is to derive analytic formulas for the probability density function (pdf) and cumulative distribution function (cdf) of the induced near-end and far-end voltage magnitudes in each case, taking into consideration the statistical behavior of the amplitude of the incident electric field vector and the angle of incidence as well. Consequently, the mean values as well as the typical deviation values are presented and the contribution of each one of the parameters is discussed in detail. Finally, a chi-square goodness-of-fit test is applied in order to fit the distribution of the induced voltage with one of the known distributions.

The finite-difference time-domain (FDTD) method is applied to the probe-fed square patch microstrip antenna stacked a parasitic patch for high gain. The input impedance, the directivity, the far field radiation patterns and the near field distributions are calculated and the relation between the antenna structure and the high gain is investigated The calculated input impedance and radiation patterns agree well with the experimental values. When the size of parasitic patch is nearly equal to the fed patch and the distance between the fed patch and the parasitic patch is about a half wavelength, the maximum gain of 9.43 dBi is obtained. In this case, the region between the fed patch and the parasitic patch forms a resonator. Then, the amplitude of current distribution on the parasitic patch becomes large and its phase is opposite to the current on the fed patch. The amplitude of electromagnetic fields of the space between the patches are increased.

Modeling of long range propagation of collimated wavepackets poses some major difficulties with the conventional FDTD scheme. The difficulties arise from the vast computer resources needed to discretize the entire region of interest and the accumulation of numerical dispersion error. As a means for circumventing these difficulties, the moving frame FDTD approach is in this work. In this approach, the computational grid size is limited to the order of the pulse length, and it and moves along with the pulse. The issues discussed in conjunction with this method are those of numerical dispersion, which is shown to be reduced substantially compared with the stationary formulation, numerical stability, and absorbing boundary conditions at the leading, trailing and side boundaries, Numerical results of pulsed beam propagation in both homogeneous and plane stratified media are shown, and the capability of the method is demonstrated with propagation distances exceeding the order of 10⁴ pulse lengths.