Kenneth K. Mei

Theory and analysis of leaky coaxial cables with periodic slots

Frequency band and coupling loss are the two important parameters of leaky coaxial cables with periodic slots. The frequency band can be predicted by analyzing the arrangement of the slots on the outer shield of the cable, but the coupling loss is not so easy to determine by classical methods. In this paper, the finite-difference time-domain (FDTD) method is used to calculate the electric field distribution in the slot cut in the outer conductor of the coaxial cable. The dyadic Greens function is then used to calculate the radiation field of the equivalent surface magnetic current densities. By these two methods, the coupling losses of the leaky coaxial cables with different periods, sizes and shapes of the slots can be accurately obtained. Some results in this paper were verified by the experimental results of leaky coaxial cables designed for railway mobile communications with a frequency band of 100-500 MHz.

Coupled azimuthal potentials for electromagnetic field problems in inhomogeneous axially symmetric media

Classical electromagnetic potential formulations are, with the exceptions of a few special cases of one-dimensional stratification, restricted to use in uniform media. A recently developed potential formulation that provides a flexible basis for numerical computation of time-harmonic field problems involving continuously and discretely inhomogeneous axially symmetric media is the topic of this paper. The formulation manifests itself in both a differential equation system and, alternately, a variational criterion. Typical numerical applications include solutions of scattering by arbitrarily shaped material bodies of revolution and radiation from inhomogeneously loaded rotationally symmetric antenna structures. Current numerical investigations by the authors, using Meis unimoment method in conjunction with both finite-difference and finite-element techniques, have shown the formulation to be highly feasible for computation of field problems having dimensions as large as several wavelengths.

Inverse scattering technique applied to remote sensing of layered media

An inverse scattering technique applied to a remote estimation of the dielectric and conductivity profile of an inaccessible layered medium is presented. The inaccessible region is illuminated by plane waves at normal incident, and the data are taken as the reflected power at a fixed remote location for a set of discrete frequencies. The problem of estimating the dielectric and conductivity profile from this set of data is posed as a nonlinear integral equation. This formulation based on reflected power is appealing for practical purpose, in that the phase information of the reflected field is not required. The equation is solved by developing a quasi-Newton iterative scheme in functional space which produces a dielectric and conductivity profile that fits the data. The Backus and Gilbert resolving-power theory is used to assess the reliability of the estimates and the resolving length of the data. Results are given for the numerical reconstruction of various dielectric and conductivity profiles from an artificial data set, together with local averages estimates and resolving kernels.

Interpolation, extrapolation, and application of the measured equation of invariance to scattering by very large cylinders

Using the conventional method of moment (MoM) calculations, a cylinder of circumferential dimension of 100 wavelengths is considered to be large. Using the measured equation of invariance (MEI) approach, a cylinder of 10000 wavelengths is within the storage capacity and numerical tolerance of a workstation. Although, the MEI has greatly reduced the storage and solution time of the matrix, its overhead to generate the matrix elements is about the same order as that of the MoM. When the target is very large, that overhead can be very time consuming. This paper presents an interpolation and extrapolation technique such that the boundary equations of the MEI for high frequencies may be predicted from those of low frequencies. It is demonstrated that in the optical limit the same set of coefficients may be used for all frequencies, which is consistent with the concept of geometric optics where the same rule is applied to all frequencies.

Comments on "A theoretical and numerical analysis of the measured equation of invariance"

In the original paper (see ibid., vol.42, no.8, p.1097-1105, 1994), the authors claimed to have found theoretical insights into the measured equations of invariance (MEI) method. Their first insight was a proof that the postulate of invariance was wrong, and their second insight led to the discovery of an optimum set of metrons. Metrons are considered to be possible induced current densities due to some unknown incident fields. They also presented a series of computational results to highlight their theories. This article points out the defects in the analyses and conclusions presented by the authors. There are two things in the paper which are basically incorrect. One is that the authors consider a zero in a numerical formulation to be an absolute zero. The other is that they assume the invariance to excitations to be the same as the invariance to metrons. Based on these assumptions, they have reached conclusions which are actually contradictory to their own calculations. This article shows where the defects of their analyses occur and why their two insights are contradictory to each other. >

Generalized Sommerfeld integrals and field expansions in two-medium half-spaces

Time harmonic modal electromagnetic fields in two-medium half-spaces are investigated. For practical and numerical considerations, the primary sources of the modal fields are chosen to be the spherical multipoles, and the potential vectors are z-directed. It is shown that modal fields of such combination are not able to represent a conventional spherical modal field. The horizontal rotating potentials are added to ensure proper representation and fast convergence. The recurrence relations which transform the spherical Hankel-Legendre functions into the Fourier-Bessel integrals are derived. The secondary fields of the Sommerfeld type are obtained for all spherical multipole sources and the added horizontally rotating potentials. The combination of the modal fields are capable of representing arbitrary electromagnetic fields resulting from radiation and scattering problems.

Radiation extraction for transmission-line interconnects

An extraction technique of radiation sources for non-uniform, transmission-lines has been introduced for the first time. The technique is to substitute solutions of the method of moments (MoM) for the lossless finite-length non-uniform transmission-line into modified transmission-line equations with dependent radiation sources to extract transmission-line parameters and the radiation source parameters.

Solving microstrip discontinuities by the measured equation of invariance

The measured equation of invariance (MEI) is a newly developed computational method which allows finite-difference (FD) or finite-element (FE) mesh to be terminated very closely to objects of interest. In this paper, the authors show how the MEI method may be applied to microstrip antennas and discontinuity problems. The authors demonstrate its use in general full-wave three-dimensional (3-D) microstrip problems, and give results for open-ended microstrip lines and microstrip bends.

Analysis of a double step microstrip discontinuity using generalized transmission line equations

In this paper, a new analysis for a double step microstrip discontinuity is presented by using generalized transmission line equations . The new analysis is based on a new concept of a finite-length transmission line integrated with its double step discontinuity so that the double step discontinuity is regarded as a finite-length nonuniform transmission line and the generalized equations could be directly implemented. Since the generalized equation coefficients are determined by dynamic numerical methods and the coefficients are invariant with the line excitation and load, the equations are dynamic rather than TEM. The interested discoveries are that the generalized equations for the whole double step structure or a partial double step structure can give us the same results and the generalized equations have broadband frequency characteristics. For the double step structure used in this paper for the analysis, the S-parameters for a frequency band from 0.5 GHz to 10 GHz can be well calculated by using only two separately generalized equations at frequency of 3 GHz and 8 GHz.

Design and calculation of the directional leaky coaxial cables

Directional leaky coaxial cables (DLCX) with periodic slots are proposed. The difference between this kind of leaky cable and the conventional LCX is that there are two rows of slots located oppositely on the outer shield of the coaxial cable. Different radiation patterns in the r-ф plane can be obtained by adjusting the relative distance between the two arrays. The radiation patterns and coupling loss are calculated by finite difference time domain method and dyadic Greens function.