Search

This article is evaluating parallel processing in electromagnetic simulations for computationally demanding application, using developed FDTD code. A waveguide lens antenna which is complex and large in terms of wavelengths has been simulated and a prototype has been fabricated and measured. Simulated and measured radiation patterns are compared and showing good agreement. The approach of using parallel version of QuickWave-3D software on a popular multi-CPU shared-memory computer has been found suitable for demanding simulations of complex structures.

This paper studied the performance of the small-packet variant of the TCP Friendly Rate Control protocol in heterogeneous networks. The objective is to maximize system utilization, often called goodput, while maintaining fairness between the nodes. We show that by selecting the right queue size policy, it can be ensured that the bandwidth is fully utilized at a steady state. Additionally, we show that the judicious use of such a policy is beneficial for heterogeneous networks with high packet drop rates.

Fast parallel processing of the FDTD method is becoming very important, and its efficiency on different multiprocessor computer architectures should be carefully examined. Is it possible to have the same versatile parallel FDTD simulator working efficiently on a set of very different computer architectures? In contrast to distributed memory systems, shared memory systems can be built with a variety of memory configuration types (and subtypes) as well as combined with distributed memory systems. The paper focuses on viability of different shared memory architectures for parallel FDTD processing. Our analysis concerns the processing method suitable for both shared and distributed memory systems. The results of experiments comparing parallel processing efficiency on different shared memory machines are presented.

Is it possible to produce a planar dipole-like antenna with a reduced conductive area without any loss in either robustness or performance? The objective is to reduce the amount of expensive conductor to be used when applying a meshing technique to the printing of antennas. In this context, robustness means that the characteristics are maintained when the antenna is damaged, for example if it is scratched. This is particularly important for radio frequency identification tags in logistic systems. A general antenna robustness evaluation methodology, based on numerical simulations of a large number of randomly damaged antennas, is used for the antenna comparisons. The antenna performance degradation, based on the return loss (S11) at 868MHz, is monitored for some basic planar antennas. Finally, we show that it is possible to produce robust low-cost antennas using wire replacements for the solid planar antennas and thus, provided that the robustness requirement is moderate, replace the solid antenna with a thin conductor analogue., VR, STC - Sensible Things that Communicate

An ongoing European project on customization of electromagnetic simulation tool is ready to present significantly improved performance of parallel version of a professional FDTD simulator. The first stage of the work is concentrated on solution for computers with the shared memory architecture. The relevance and feasibility study of the solution for contemporary computers is shown. Possible ways to reach further performance improvements as well as solutions for distributed memory machines are discussed.

A commercial FDTD simulation software is the basis of an ongoing parallelization project. In the first phase of the project the work is aiming at optimization of the program on shared memory computers. In this paper we present the initial methods of parallelization with the achieved performance on a Sun Fire 15K NUMA computer. The achieved results are discussed, together with the possible paths to improve the speedup.

The full band Monte Carlo (MC) simulation framework is regarded as the most accurate method available to study high-field carrier transport in semiconductors. Its potential has been demonstrated in a large number of studies over the years. In this work we focus on how the quantum mechanical uncertainty at high scattering rates affects the validity of Fermi’s Golden Rule, which traditionally is the basis for the scattering handling in the MC method. Considering the uncertainty is important in for instance silicon carbide, which at moderate energies exhibits a scattering rate exceeding 1014 s- 1. The expression for time-dependent scattering rate is presented together with calculated rates for some initial states with acoustic as well as polar-optical phonon interaction. A first-order time-dependent algorithm for handling of scattering events in MC simulators is proposed., QC 20140930

The scattering probabilities for a multiband carrier wave function have been studied for hole transport in 4H-SiC. During the drift at high electric fields it is possible to find the carriers with almost equal probability in two neighboring bands. In contrast to the standard Monte Carlo procedure, a possible interference between the components of the wave function in different bands appears in the calculation of the scattering probabilities when the time between scattering events is short. An important role is played by the overlap integral, which forbids this interference in some regions of the epsilon-k (energy-wavevector) space. We compare the results obtained using the true overlap integrals calculated from the wave functions with the common assumption, in which the overlap integral is assumed equal to one. Different scattering mechanisms, such as polar-optical phonon and intervalley optical phonon are studied. The model includes a spread in energy according to the Heisenberg uncertainty principle at short times. We have calculated the final state probability distribution considering the whole Brillouin zone of 4H-SiC for a particular set of initial states., QC 20111025

A quantum mechanical extension of the full band ensemble Monte Carlo (MC) simulation method is presented. The new approach goes beyond the traditional semi-classical method generally used in MC simulations of charge transport in semiconductor materials and devices. The extension is necessary in high-field simulations of semiconductor materials with a complex unit cell, such as the hexagonal SiC polytypes or wurtzite GaN. Instead of complex unit cells the approach can also be used for super-cells, in order to understand charge transport at surfaces, around point defects, or in quantum wells., QC 20141209

The GEMS (General Monte Carlo Semiconductor) simulator is a full band ensemble program developed at Mid-Sweden University. A principal objective of the project is to create a flexible tool that can be used for the exploration of properties of new semiconductor materials as well as devices fabricated of these materials. Both cubic and hexagonal crystal symmetries can be handled with the simulator, and any material with these symmetries can be simulated if an appropriate electronic band structure together with the parameters for the scattering mechanisms is provided. During the development of GEMS a large number of studies have been performed of bulk charge transport and device characteristics. The simulated materials are 2H-, 3C-, 4H- and 6H-SiC, as well as silicon. This presentation discusses the charge transport properties of the studied SiC polytypes, and the resulting performance of the simulated devices., QC 20111102

The scattering probabilities for a multiband carrier wave function have been studied for hole transport in 4H-SiC. During the drift at high electric fields it is possible to find the carriers with almost equal probability in two neighboring bands. In contrast to the standard Monte Carlo procedure, a possible interference between the components of the wave function in different bands appears in the calculation of the scattering probabilities when the time between scattering events is short. An important role is played by the overlap integral, which forbids this interference in some regions of the epsilon-k (energy-wavevector) space. We compare the results obtained using the true overlap integrals calculated from the wave functions with the common assumption, in which the overlap integral is assumed equal to one. Different scattering mechanisms, such as polar-optical phonon and intervalley optical phonon are studied. The model includes a spread in energy according to the Heisenberg uncertainty principle at short times. We have calculated the final state probability distribution considering the whole Brillouin zone of 4H-SiC for a particular set of initial states.

A set of equations for calculating the probability for electric-field-induced interband transitions in periodic crystals (Krieger and Iafrate 1986 Phys. Rev. B 33 5494) can be used in combination with the full band Monte Carlo method to study high-field electronic transport properties in semiconductors. However, when the equations are applied to realistic cases in which the underlying band structure is obtained from numerical band structure programmes, the equations are not directly solvable because of the indeterminacy of the phases of the band structure Bloch wavefunctions. Here we discuss this problem and present a method for choosing the phases of the Bloch functions in such a way that the equations yield physically correct interband transition probabilities.

A full-band ensemble Monte Carlo simulation has been used to study the high-field carrier transport properties of 4H-SiC. The complicated band structure of 4H-SiC requires the consideration of band-to-band tunneling at high electric fields. We have used two models for the band-to-band tunneling; one is based on the overlap test and the other on the solution of the multiband Schrodinger equations. The latter simulations have only been performed for holes in the c-axis direction, since the computer capacity requirement are exceedingly high. Impact-ionization transition rates and phonon scattering rates have been calculated numerically directly from the full band structure. Coupling constants for the phonon interaction have been deduced by fitting of the simulated low-field mobility as a function of lattice temperature to experimental data. Secondary hot electrons generated as a consequence of hole-initiated impact ionization are considered in the study for both models of band-to-band tunneling. When the multiband Schrodinger equation model is used for holes in the c-axis direction, a significant change in the electron energy distribution is found, since the hole impact-ionization rate is very much increased with this model. The secondary electrons increase the average energy of the electron distribution leading to a significant increase in the electron-initiated impact-ionization coefficients. Our simulation results clearly show that both electrons and holes have to be considered in order to understand electron-initiated impact ionization in 4H-SiC

A Monte Carlo study of the transient response of single photon absorption in X-ray pixel detectors is presented. The simulation results have been combined with Monte Carlo simulation of the X-ray photon transport and absorption, and used to estimate the image properties of a detector system, including the pixel array and readout electronics. The study includes several different simulation challenges, such as full band Monte Carlo simulation of charge transport in large devices (300 mu m * 100 mu m), modelling of three-dimensional electrostatic effects using cylindrical coordinates, Monte Carlo simulation of photon transport and absorption, and a system level Monte Carlo simulation of the entire pixel detector and readout, 3rd IMACS Seminar on Monte Carlo Methods (MCM 2001), Sep 10-14, 2001, Saltzburg, Austria

The spatial resolution of scintillator-coated X-ray pixel detectors is usually limited by the isotropic light spread in the scintillator. One way to overcome this limitation is to use a pixellated scintillating layer on top of the semiconductor pixel detector. Using advanced etching and filling techniques, arrays of CsI columns have been successfully fabricated and characterized. Each CsI waveguide matches one pixel of the semiconductor detector, limiting the spatial spread of light. Another concept considered in this study is to detect the light emitted from the scintillator by diodes formed in the silicon pore walls. There is so far no knowledge regarding the theoretical limits for these two approaches, which makes the evaluation of the fabrication process difficult. In this work we present numerical calculations of the signal-to-noise ratio (SNR) for detector designs based on scintillator-filled pores in silicon. The calculations are based on separate Monte Carlo (MC) simulations of X-ray absorption and light transport in scintillator waveguides. The resulting data are used in global MC simulations of flood exposures of the detector array, from which the SNR values are obtained. Results are presented for two scintillator materials, namely CsI(Tl) and GADOX., 4th International Workshop on Radiation Imaging Detectors, Sep 08-12, 2002, Amsterdam, Netherlands

A full band Monte Carlo study of the electron transport in 3C-SiC is presented based on an ab initio band structure calculation using the local density approximation to the density functional theory. The scattering rates and impact ionization transition rates have been calculated numerically from the ab initio band structure using both energy dispersion and numerical wave functions. This approach reduces the number of empirical parameters needed to a minimum. The two empirical coupling constants used have been deduced by fitting the simulated mobility as a function of lattice temperature to experimental data. The peak velocity was found to be approximately 2.2*107 cm/s with a clear negative differential mobility above 600 kV/cm. The electron initiated impact ionization coefficients were found to be 2-10 times stronger than the reported values for the hole initiated impact ionization

In this paper we discuss different algorithms to select final states after scattering within the full band ensemble Monte Carlo (MC) simulation framework. Faster computers have made it possible to use more elaborate microscopic models for simulation of advanced semiconductor devices. Microscopic details like the band structures, the phonon dispersions and k-vector dependent deformation potentials are becoming accessible, which demand good algorithms for the selection of final states after scattering. The algorithm should also be adapted for a numerical representation of the band structure. Such an algorithm based on the rejection method is presented, which has been studied and compared in terms of efficiency and accuracy with two other algorithms for final state selection. The rejection algorithm is considered as suitable for anisotropic scattering mechanisms and when a relatively slow method is used for calculation of the wave-function overlap integral. It is shown that an accurate model of acoustic phonon scattering results in two characters of the process: one short-q process and one long-q process, very similar to a nonpolar optical intervalley process.

The high field hole transport in 4H-SiC has been studied using a full band Monte Carlo (MC) simulation model that includes band to band tunneling and allows mixing of the band states during carrier drift. Impact ionization coefficients along the c-axis direction have been extracted and compared with experimental data. It is shown that the band to band tunneling mechanism is crucial in order to explain experimental results. The carrier distribution function obtained from the MC simulations has been used to determine the breakdown luminescence spectra coming from interband transitions. Our results are in good agreement with the available experimental luminescence spectra for SiC polytypes, and the importance of including interband tunneling is clearly demonstrated., 12th International Conference on Nonequilibrium Carrier Dynamics in Semiconductors (HCIS-12), Aug, 2001, Santa Fe, New Mexico

A full band Monte Carlo study of the electron transport in 3C-SiC is presented based on ab initio band structure calculation using the Local Density Approximation (LDA) to the Density Functional Theory (DFT). The scattering rates and impact ionization transition rates have been calculated numerically from the band structure using both energy dispersion and wave functions. Coupling constants for the phonon interactions have been deduced by fitting the simulated mobility as a function of lattice temperature to experimental data. The saturation velocity was found to be approximately 2.2*107 cm/s with a visible negative differential mobility above 600 kV/cm. The electron initiated impact ionization coefficients were found to be 2 to 10 times stronger than reported values for the hole initiated impact ionization

A fully stochastic model for the imaging properties of X-ray silicon pixel detectors is presented. Both integrating and photon counting configurations have been considered, as well as scintillator-coated structures. The model is based on three levels of Monte Carlo simulations; photon transport and absorption using MCNP, full band Monte Carlo simulation of charge transport and system level Monte Carlo simulation of the imaging performance of the detector system. In the case of scintillator-coated detectors, the light scattering in the detector layers has been simulated using a Monte Carlo method. The image resolution was found to be much lower in scintillator-coated systems due to large light spread in thick scintillator layers. A comparison between integrating and photon counting readout methods shows that the image resolution can be slightly enhanced using a photon-counting readout. In addition, the proposed model has been used to study charge-sharing effects on the energy resolution in photon counting detectors. The simulation shows that charge-sharing effects are pronounced in pixel detectors with a pixel size below 170 * 170 mu m2. A pixel size of 50 * 50 mu m2 gives a highly distorted energy spectrum due to charge sharing. This negative effect can only be resolved by introducing advanced counting schemes, where neighbouring pixels communicate in order to resolve the charge sharing., 3rd International Workshop on Radiation Imaging Detectors, Sep 23-27, 2001, Orosei, Italy

Numerical simulations of microscopic and macroscopic device properties of field effect transistors in 4H- and 6H-SiC are presented. The microscopic properties have been simulated using a full band (ab initio method) Monte Carlo simulation model and the macroscopic properties have been simulated using a drift-diffusion model with transport parameters obtained from the Monte Carlo simulations. Different models for the SiC/SiO2 interface in SiC MOSFETs have been evaluated and compared with experimental data. Finally, we present a comparison of simulated device performance for MOSFETs and MESFETs in 4- and 6H-SiC technologies. Both vertical (SIT) and lateral MESFET structures have been considered

The ohmic transport of holes in p-type Al-doped 4H-SiC samples is investigated using a Monte Carlo tool based on a full-potential band structure. The temperature and doping dependence of the hole mobility and its anisotropy are calculated and discussed from a physical point of view, where we stress the importance of considering two-band conduction. Acoustic and optical phonon scattering, as well as ionized and neutral impurity scattering, have been considered. The Monte Carlo program also considers incomplete ionization of impurity atoms compatible with an Al ionization energy of 0.2 eV.

Silicon Carbide is a very interesting semiconductor material for high temperature, high frequency, and high power applications. The main reasons are its high saturation velocity, large thermal conductivity, high Schottky barriers, and high breakdown voltages. High quality 4H-SiC and 6H-SiC polytype substrates and epitaxial layers are commercially available today. An additional advantage of SiC is the native oxide that allows fabrication of MOS devices. A large effort has been devoted towards the development of high performance devices in SiC. The largest success has been for unipolar devices like Schottky diodes and different kinds of MESFETs. MOSFETs have also been fabricated in both 4H- and 6H-SiC. Unfortunately, the MOSFET performance was found to be much worse than expected, due to a very low surface mobility. Nevertheless, the technology developed is very interesting and includes possible large scale integration of digital circuits operating at very high temperatures. In this work we present numerical simulations of the device performance of different Field Effect Transistors (FETs). Both full band Monte Carlo simulations and macroscopic modeling using the drift-diffusion approach have been utilized in this work. The Monte Carlo simulations have been used to extract transport parameters and to evaluate the macroscopic models in a device configuration

In a traditional Monte Carlo (MC) model the carrier preserves its band identity during the free flight between scatterings. However, this assumption may not be valid in semiconductor materials with complex unit cell. A new model is needed where the traditional way to use classical equations during the free flight between scattering is replaced by a fully quantum mechanical model of the Bloch carrier dynamics between scattering events. In this work we present such a model along with simulated results of the hole initiated impact ionization coefficients of 4H-SiC

A full band Monte Carlo (MC) study of the high frequency performance of a 4H-SiC Short channel vertical MESFET is presented. The MC model used is based on data from a full potential band structure calculation using the local density approximation to the density functional theory. The MC results have been compared with simulations using state of the art drift-diffusion and hydrodynamic transport models. Transport parameters such as mobility, saturation velocity and energy relaxation time are extracted from MC simulations

A Monte Carlo (MC) study of the hole transport in 4H-SiC is presented using three different MC models. The three models represent different approximation levels regarding band structure and scattering formulation. The most advanced model is a completely k-vector dependent full band model while the simplest model uses three analytical bands with energy dependent scattering rates. The intermediate MC model uses a full band structure calculated using a simple k.p formulation. A comparison between the models in terms of coupling constants, scattering rate, temperature dependent mobility and saturation velocity is presented, 2nd IMACS Seminar on Monte Carlos Methods, Jun 07-11, 1999, Varna, Bulgaria

The 4H-SiC static induction transistor (SIT) is a very competitive device for high frequency and high power applications (3-6 GHz range). The large breakdown voltage and the high thermal conductivity of 4H-SiC allow transistors with extremely high current density at high voltages. The SIT transistor shows better output power capabilities but the unity current-gain frequency is lower compared to a MESFET device. In this work we show, using a very accurate numerical model, that a compromise between the features given by the SIT structure and the ordinary MESFET structure can be obtained using the vertical MESFET structure. The device dimension has been selected very aggressively to demonstrate the performance of an optimized technology. We also present results from drift-diffusion simulations of devices, using transport parameters obtained from the Monte Carlo simulation. The simulations indicate that 2H-SiC is superior to both 4H and 6H-SiC for vertical devices. For lateral devices, 2H-SiC is slightly faster compared to an identical 4H-SiC device, 11th European Conference on Diamond, Diamond-like Materials, Carbon Nanotubes, Nitrides and Silicon Carbide (Diamond 2000), Sep 03-08, 2000, Oporto, Portugal

The ohmic transport of holes in p-type aluminum-doped 4H-SiC samples is investigated using a Monte Carlo (MC) tool based on a full-potential band structure. The temperature and doping dependence of the hole mobility and its anisotropy are calculated and discussed from a physical point of view, where we stress the importance of considering two-band conduction. Acoustic and optical phonon scattering, as well as ionized and neutral impurity scattering, have been considered. The MC program considers incomplete ionization of impurity atoms, and we assume an impurity level with the ionization energy 0.2 eV, corresponding to Al-doped samples. © 2001 Published by Elsevier Science B.V, The paper was also presented at the following conference: Spring Meeting of the European-Materials-Research-Society STRASBOURG, FRANCE, JUN 05-08, 2001 European Mat Res Soc

The imaging properties of X-ray pixel detectors depend on the quantum efficiency for X-rays, the generated signal for each X-ray photon and the distribution of the generated signal between different pixels. In a scintillator coated device the signal is generated both by X-ray photons captured in the scintillator and by X-ray photons captured directly in the semiconductor. Hence, the signal-to-noise ratio (SNR) in the image is then a function of the number of photons captured in each of these processes, and the yield of each process, in terms of electron-hole pairs (EHPs) produced in the semiconductor. The full process from the absorption of the X-ray photon to the final signal read out from the detector has been simulated with a combination of the Monte Carlo program MCNP and the commercial carrier transport simulation tool MEDICI. An in house program calculating the light transport between the scintillator and the semiconductor serves as a link

The high field transport properties of 4H-SiC have been studied using a bipolar full band ensemble Monte Carlo model. The impact ionization transition rates and the phonon scattering rates for both electrons and holes have been calculated numerically from the full band structure. The simulation results show a large anisotropy in the impact ionization coefficients for both electrons and holes.