Flnite-difierence Time-domain Research Articles

COMPARISON OF RADAR AND THERMOACOUSTIC TECHNIQUE IN MICROWAVE BREAST IMAGING

Microwave radar and microwave-induced thermoacoustic technique exploit the contrast in the permittivity and conductivity between malignant and healthy tissue. They have emerged as promising techniques for detecting breast cancers. This paper compares the imaging capability of these techniques in the presence of homogeneous and heterogeneous breast tissue. Relying on the data from the flnite-difierence time-domain simulations, the study shows that both techniques are capable of imaging homogeneous objects. In the presence of electromagnetic dispersion and heterogeneity, radar signals sufier from strong dispersion and multiple scattering, which decorrelate the signals with the scatterers. The microwave- induced thermoacoustic technique takes the advantage of breast being acoustically homogeneous and is capable of generating high-quality images.

APPLICATIONS OF A THREE-DIMENSIONAL FDTD METHOD WITH WEAKLY CONDITIONAL STABILITY TO THE ANALYSIS OF MICROSTRIP FILTERS WITH FINE SCALE STRUCTURES

In three-dimensional space, the hybrid implicit-explicit flnite-difierence time-domain (HIE-FDTD) method is weakly condi- tionally stable, only determined by two space-discretizations, which is very useful for problems with flne structures in one direction. Its nu- merical dispersion errors with nonuniform cells are discussed and com- pared in this paper. To enlarge the applicable fleld of the HIE-FDTD method to open space, the absorbing boundary conditions (ABCs) for this method are also introduced and applied. Two microstrip fllters with flne scale structures in one direction are solved by the HIE-FDTD method. Conventional FDTD method and alternating-direction im- plicit FDTD (ADI-FDTD) method are also used for comparing. Re- sults analyzed by the HIE-FDTD method agree well with those from conventional FDTD, and the required central process unit (CPU) time is much less than that of the ADI-FDTD method.

FDTD METHOD INVESTIGATION ON THE POLARIMETRIC SCATTERING FROM 2-D ROUGH SURFACE

A polarimetric scattering from two-dimensional (2-D) rough surface is presented by the flnite-difierence time-domain (FDTD) algorithm. The FDTD calculations with sinusoidal and pulsed plane wave excitations are performed. As the sinusoidal FDTD is concerned, it is convenient to obtain the scattered angular distribution of normalized radar cross section (NRCS) from rough surface for a single frequency. And the advantage of pulsed FDTD is to calculate the frequency distribution of NRCS from rough surface in a scattered direction of interest. A single frequency scattering from rough surface by sinusoidal FDTD is validated by the result of Kirchhofi Approximation (KA). And the frequency response of rough surface by pulsed FDTD is verifled by that of sinusoidal FDTD, which requires an individual FDTD run for every frequency. To save computation time, the MPI-based parallel FDTD method is adopted. And the computation time of parallel FDTD algorithm is dramatically reduced compared to a single-process implementation. Finally, the polarimetric scattering of rough surface with the sinusoidal and pulsed FDTD illumination are presented and analyzed for difierent polarizations.

WEAKLY CONDITIONALLY STABLE AND UNCONDITIONALLY STABLE FDTD SCHEMES FOR 3D MAXWELL'S EQUATIONS

To overcome the Courant limit on the time step size of the conventional flnite-difierence time-domain (FDTD) method, some weakly conditionally stable and unconditionally stable FDTD methods have been developed recently. To analyze the relations between these methods theoretically, they are all viewed as approximations of the conventional FDTD scheme in present discussion. The errors between these methods and the conventional FDTD method are presented analytically, and the numerical performances, including computation accuracy, e-ciency, and memory requirements, are discussed, by comparing with those of the conventional FDTD method.

SAR AND TEMPERATURE ELEVATION IN A MULTI-LAYERED HUMAN HEAD MODEL DUE TO AN OBLIQUELY INCIDENT PLANE WAVE

In this paper, the Speciflc Absorption Rate (SAR) distribution in a human head irradiated by an obliquely incident electromagnetic plane wave is studied. A simple planar multi-layered structure is used to model the human head. Moreover, the steady state thermal elevation distribution is calculated by solving the bioheat equation using the flnite difierence time domain (FDTD) method. Both types of wave polarization (perpendicular and parallel) will be discussed. The obtained results conflrm the importance of performing a thermal analysis along with the dosimetric one. It is found that the induced temperature elevation in the brain region, in all the examined conditions, never exceeds 0.4 - C. This value is well below the threshold for the induction of adverse thermal efiects to the neurons which is 3.5 - C.

TIME-DOMAIN INVERSE SCATTERING OF A TWO-DIMENSIONAL METALLIC CYLINDER IN SLAB MEDIUM USING ASYNCHRONOUS PARTICLE SWARM OPTIMIZATION

This paper presents asynchronous particle swarm opti- mization (APSO) applied to the time-domain inverse scattering prob- lems of two-dimensional metallic cylinder buried in slab medium. For this study the flnite-difierence time-domain (FDTD) is employed for the analysis of the forward scattering part, while for the APSO is ap- plied for the reconstruction of the two-dimensional metallic cylinder buried in slab medium, which includes of the location and shape the metallic cylinder. For the forward scattering, conceptually several elec- tromagnetic pulses are launched to illuminate the unknown scatterers, and then the scattered electromagnetic flelds around are measured. In order to e-ciently describe the details of the cylinder shape, sub- gridding technique is implemented in the flnite difierence time domain method. Then, the simulated EM flelds are used for inverse scattering, in which APSO is employed to transform the inverse scattering prob- lem into optimization problem. By comparing the simulated scattered flelds and the calculated scattered flelds, the shape and location of the metallic cylinder are reconstructed. In addition, the efiects of Gaussian noises on imaging reconstruction are also investigated.

Accurate Evaluation of RF Coil-tissue Interactions Using a Hybrid FDTD-MoM Method

A comprehensive consideration of loading efiects is essential to the design/analysis of high-frequency MRI radio frequency (RF) coils. In this study, a hybrid flnite difierence time domain (FDTD)-method of moments (MoM) simulation method is implemented for the accurate evaluation of RF coil-tissue interaction. Using this approach, the RF coil is modelled using the MoM algorithm, and the biological load is handled with the FDTD algorithm. Efiective commu- nication between these two approaches is realized through a Huygence equivalent surface. The algorithm has been successfully validated and applied to a loaded coil study and the simulation demonstrated the potential of the hybrid algorithm for high frequency RF coil designs.

FDTD MODELING AND SIMULATION OF MICROWAVE HEATING OF IN-SHELL EGGS

Considering microwaves as a viable alternative for the pasteurization of In-shell eggs, preliminary trials performed had conflrmed that microwave at 2450MHz can be successfully used to raise the temperature of in-shell eggs to the required pasteurization temperatures in a few minutes. Based on these trials a flnite difierence time domain (FDTD) model was developed using C language and MATLAB to simulate the E fleld and power distribution in lossy dielectric media like that of the egg components (egg white and yolk) taking into consideration the complex shape, dielectric properties and heterogeneous composition of the in-shell egg. This can be used to assist in the design and development of an industrial microwave in- shell eggs pasteurization unit.

The Analysis and Design of High Power Millimeter Wave Pulse Detector for 2 mm Frequency Band

The research progress of a high power millimeter wave pulse detector for 2mm frequency band is presented here. This power detector, so-called resistive sensor, is composed of a 2mm standard waveguide and a semiconductor sensing sample within it. At flrst, the principle of the detector is analyzed theoretically, and the sensitivity of the detector in the warm-electron region is derived. Then the structural parameters of the detector are calculated and optimized to make the frequency response ∞at over the waveguide frequency band. A three-dimensional flnite difierence time domain (FDTD) method is used here. The results show that, while the biased voltage on the detecting element is 10V, the sensitivity of the optimized detector is about 6V/kW with a ∞uctuation less than 27% in the frequency range 113{173GHz, especially less than 9.8% in the frequency range 130{160GHz. Compared with the diode detectors in 2mm frequency band, the designed semiconductor detector shows better performance in pulse power measurement.

TRANSIENT RESPONSES OF SOME ANTENNAS UNDER THE IMPACT OF AN INTENTIONALLY INCIDENT HIGH-POWER ELECTROMAGNETIC PULSE

An improved flnite difierence time domain (FDTD) method is employed for fast capturing transient responses of reconflgurable and multiple-input-multiple-output (MIMO) antennas under the impact of an intentional high-power electromagnetic pulse (EMP) but with difierent waveforms, respectively, where lumped element and sub-cellular thin-wire algorithms and coaxial feed model are integrated together for handling such three-dimensional antennas used for wireless communication. Parametric studies are carried out to show efiects of high-power EMP waveforms, its polarization state and incident direction on the transient coupled voltages on the coaxial feed line and across the diodes, with su-cient information obtained for understanding the interaction between the EMP and the antennas.

A COMPACT POLARIZATION BEAM SPLITTER BASED ON A MULTIMODE PHOTONIC CRYSTAL WAVEGUIDE WITH AN INTERNAL PHOTONIC CRYSTAL SECTION

We present the design and simulation of an ultra-compact polarization beam splitter (PBS) by combining a photonic crystal (PhC) multimode waveguide and an internal PhC section. The PhC multimode waveguide is designed to collect the powers re∞ected by or transmitted through the internal PhC structure which serves as a polarization sensitive scatterer. Plane wave expansion (PWE) method is used to calculate the band structure and the flnite-difierence time- domain (FDTD) method is employed to obtain the spectrum response. The simulation results show that the present design can give an ultra- compact PBS with high extinction ratio over a broad bandwidth.

QUANTITATIVE ESTIMATION OF SCATTERING WAVES IN CYLINDER-BODY MODEL FOR BODY AREA NETWORKS: COMPARISON OF ANALYSES WITH UNIFROM CYLINDER- AND SLAB-BODY MODELS

This paper estimates separately the components of scattering waves generated in cylinder-body model for body area networks. For the evaluation, scattering fleld formulations in relation to uniform cylinder- and slab-body models are provided, and the reliability of the analyses is testifled by the comparison with results computed by the flnite-difierence time-domain (FDTD) method. Creeping waves, cylinder leaky waves, and cylinder guided waves, which are created only in cylindrical structure, are extracted quantitatively by contrasting the scattering waves that are calculated with the two body models. In addition to the extracted waves, other components of scattering waves such as re∞ected waves, transmitted waves, surface waves, leaky waves, and guided waves also are examined. From evaluations with various operating frequencies and thicknesses of the body model, it is conflrmed that re∞ected waves have the most in∞uence on electrical characteristics of a source. Moreover creeping

A Combined Cavity with Improved Performance under Simultaneous Resonance of Sub-cavities

We insert a dielectric rod in the point cavity or increase the volume of the point cavity in the combined system to shift the resonant frequency of the point cavity to be the same as that of the feedback cavity. These systems are advantageous over our previous design in that the feedback cavity is uniform and no stray scattering exists in the cavity. Simulations through flnite-difierence time-domain (FDTD) method demonstrate that, in double resonance, the maximum quality factor of the new structure is promoted by 34.7%, and the maximum localized fleld intensity in the point defect cavity is promoted by 314% over that in the earlier structure.

移动坐标系FDTD方法在脉冲波传播中的应用

The moving coordinate frame flnite-difierence time-domain (FDTD) method is developed to simulate the propagation of electromagnetic pulses over long distances. In this paper, this method is extended to three dimensional cases. The discretized form of moving coordinate frame FDTD in three dimensions is given, and is applied to calculate propagation of difierent pulses in rectangular waveguides. The widths of pulses in propagation direction increase when pluses propagate in waveguide. Dispersion phenomena appear, and slow parts of the pulses move to the tail of pulses. By comparing the results of both FDTD method and moving coordinate frame FDTD method, this method is numerically validated. Relative difierence between two methods is less than 0.1%, the simulation results are shown to be accurate. With the increasing applications on communication and detection, there is an ever-increasing in- terest in the propagation of pulses over long distance. The flnite-difierence time-domain (FDTD) method (1) is widely used in electromagnetic simulation. Because the FDTD method is a full- wave method, when simulating propagation of electromagnetic pulses over very long distance, the computational burden quickly becomes too large to afiord.This is the major di-culty why FDTD method can not be applied to simulate long distance propagation of pulses directly. In order to overcome this di-culty, moving coordinate is combined with FDTD method. Models of one di- mension as electromagnetic pulse propagation over large distance in medium (2) and acoustic wave propagation in waveguide has been studied (3) approach this method. Coordinate of the direction in which pulse's propagation is transformed into moving coordinate, which is 'frozen' with pulses, and the pulses are in small calculation zones when simulating (as shown in Fig. 1). Moving grid whose size is limited to the order of the pulse length is created, and moves along the approach of the center-of-mass of the wavepacket. This method takes advantage of the fact that when an electromagnetic pulse propagates over long distance, the signiflcant pulse energy exists only over a small part of the propagation path at any time. The FDTD meshes (in the direction of the propagation path) need only to be large enough to contain the pulses while using this method to simulate the propagation of pulses. This method is useful in particular for tracking well-collimated short pulse beam, referred to as pulsed beams or localized waves. In this paper, the models are extended to simulate electromagnetic pulses propagating in waveg- uide over large distance and results are compared with the results using FDTD approach in order to testify the validity and feasibility of the moving coordinate FDTD method.

The Feasibility of Numerical Calculations of Vias Using the Matrix-Penciled Moment Method

The feasibility of numerical calculations of a through hole via using the Matrix- Penciled Moment Method (MP-MM) proposed by Show-Gwo Hsu and Ruey-Beei Wu is investi- gated. It shows that the MP-MM can't be used to the via which is near the source port, and there is a lower bounded operating wavelength beyond which MP-MM is not applicable. The conclusions are proven by both theory and the simulation results gotten by HFSS. The miniaturization and large scale integration of electronic devices places increasing demand on the multilayer interconnection geometry. Design of high-speed electronic systems requires careful consideration of the high frequency problems due to parasitic efiects from these interconnects. In multi-layer structures, there are parallel lines for signal transmission as well as vias to connect the signal lines in difierent layers. Several studies involving the electrical characterization and design of the signal lines have been reported in previous literature. The equivalent circuits for some basic via structures have been presented by several investigators on the basis of quasi-static analysis (1). Using the flnite-difierence time-domain (FD-TD) method (2), presented a full wave analysis for the propagation characteristics of a through hole via. In (3), three-dimensional mode matching technique was used to model the via hole grounds in microstrip. In (4,5), flnite-elment time-domain method was used to analyze through hole. In the last decade, Show-Gwo Hsu and Ruey-Beei Wu proposed full wave characterization of a through hole via using the Matrix-Penciled Moment Method (6). Based on the Pocklinton Equation, the method calculates the current distributions on via discontinuities by MoM, extracts poles by using Matrix-penciled method, gets the complex coe-cient of every mode by solving Vandermonde equations, and flnally, obtains the scattering parameters of via from superposition principle. The analytical and computational process is more complex than any other computational electromag- netic methods, but it has predominance in computational accuracy and speed for using the right basis and testing function and transforming the three dimensional problem to one dimensional by equivalent thin wire approximation. The MoM is a computational electromagnetic method with high computational accuracy in most scattering and antenna radiation problem, but the method in Show-Gwo Hsu's literature contained two equivalences: 1) The shape of the microstrip/strip lines in horizontal segment (shown in Figure 1 is ignored and treated as circular cylinder. This equivalence is accurate when the length of transmission lines is inflnite, but in actual microwave circuits, due to the restriction of trace and the measure of microwave circuit boards, the microstrip/strip line can't extend to inflnity. The equivalence theory and its limitation will be presented in Sections 3 and 4, respectively. 2) MP-MM is based on Pocklinton Equation obtained by considering the antenna as thin wire. Section 5 will discuss the condition in which the transmission line can be seen as thin wire and propose a lower bounded operating wavelength of this method. There is no question about the predominance of this method in computing the electromagnetic characterizations of via, but it is not a general-purposed method for it is not based on the real structure and rigorous equation which has no approximation. At present, most literatures about this method introduce its principles (6) and/or some partic- ular calculational examples (7), on the contrary, neglecting the investigation about the application spectrum in which the method can be used, that is the emphasis of this paper. The structure investigated in this paper is a through via with one ground plane. The tool used for simulation is HFSS.

FDTD Study of a Novel Terahertz Emitter with Electrical Field Enhancement Using Surface Plasmon Resonance

In this work, the flnite-difierence time-domain (FDTD) technique is used to study novel terahertz (THz) emitter structures. The proposed THz antenna requires less pump power of the femtosecond laser pulse whilst provides higher THz output power. This is achieved by the enhancement of the localized electric fleld in the THz emitter This electric fleld enhancement is found to have two origins: One is owing to the enhancement of the static electric fleld of the bias voltage, and the other is the enhancement of the electric fleld of the incoming femtolaser pulse. The latter enhancement is caused by the interaction of the pump laser and the surface plasmon resonance at the conical gold structure of the photoconductive emitter. This new terahertz emitter could lead to new applications where high-power and broadband terahertz sources are needed.

Numerical Simulation of EM Environment and Human Exposure When Using RFID Devices

In this work, RFID readers and active tags, operating at 2.4GHz and 5.8GHz, are simulated to calculate the electromagnetic fleld distribution generated in their surroundings, considering the presence of the user's body and possible scattering obstacles, and including the analysis of the speciflc absorption rate (SAR) in the human models (which can be very close to the tag). The flnite-difierence time-domain (FDTD) method and the flnite elements (FE) method are used. Results show that signiflcant fleld level can be found in regions far from the tag-reader direction. This fact could constitute a risk because of the possible presence of general public and eventual eavesdroppers. Nevertheless, the calculated values of SAR are always below the basic restrictions.

GPU-ACCELERATED FUNDAMENTAL ADI-FDTD WITH COMPLEX FREQUENCY SHIFTED CONVOLUTIONAL PERFECTLY MATCHED LAYER

This paper presents the graphics processing unit (GPU) accelerated fundamental alternating-direction-implicit flnite-difierence time-domain (FADI-FDTD) with complex frequency shifted convolu- tional perfectly matched layer (CFS-CPML). The compact matrix form of the conventional ADI-FDTD method with CFS-CPML is formu- lated into FADI-FDTD with its right-hand-sides free of matrix oper- ators, resulting in simpler and more concise update equations. Using Compute Unifled Device Architecture (CUDA), the FADI-FDTD with CFS-CPML is further incorporated into the GPU to exploit data par- allelism. Numerical results show that a much higher e-ciency gain of up to 15 times can be achieved.

APPLICATION OF THE IMPROVED MATRIX TYPE FDTD METHOD FOR ACTIVE ANTENNA ANALYSIS

An improved flnite-difierence time-domain (FDTD) method has been extended to analyze the antennas with complicated lumped/active networks in this paper. The improved FDTD method is based on a novel integral transform and the matrix theory. Combining the novel integral transform with Kirchhofi's circuit laws, the hybrid networks comprised of high order linear and nonlinear elements with arbitrary connection can be modeled by a stable matrix equation. An efiective model is built for linear lumped networks including internal independent sources. A wire antenna loaded with wideband match network and a two-element active patch antenna loaded with Gunn diodes are analyzed by the developed techniques. The analysis results indicate that the improved matrix-type FDTD method is not only sta- ble and accurate, but also time-saving in simulating the complicated hybrid networks.

ENHANCEMENT OF BLUE LIGHT EMISSION USING SURFACE PLASMONS COUPLING WITH QUANTUM WELLS

dimension flnite-difierence time-domain (FDTD) method is used to simulate the enhanced blue light emission of gallium nitride light emitting diode (GaN-LED) using the surface-plasmons (SPs) cou- pling with the quantum wells. The numerical simulation results demon- strate that when the silver fllm is coated on GaN-LED, the excited SPs play a key role in the enhanced blue light emission, and the en- hancement depends on the geometries of GaN-LED and silver fllm. An enhancement factor is given to describe the enhancement efiect of light emission. By changing the structure parameters of GaN-LED and sil- ver fllm, the enhanced peak of the light emission in the visible region can be controlled. Under the optimal parameters, about 17 times en- hancement at 460nm can be obtained, and the enhancement efiect is evidently demonstrated by the SPs fleld distribution.