| Publication Type |
dissertation |
| School or College |
College of Engineering |
| Department |
Electrical & Computer Engineering |
| Author |
Smith, Steven Michael |
| Title |
Stochastic finite-difference time-domain |
| Date |
2011-05 |
| Description |
This dissertation presents the derivation of an approximate method to determine the mean and the variance of electromagnetic fields in the body using the Finite-Difference Time-Domain (FDTD) method. Unlike Monte Carlo analysis, which requires repeated FDTD simulations, this method directly computes the variance of the fields at every point in space at every sample of time in the simulation. This Stochastic FDTD simulation (S-FDTD) has at its root a new wave called the Variance wave, which is computed in the time domain along with the mean properties of the model space in the FDTD simulation. The Variance wave depends on the electromagnetic fields, the reflections and transmission though the different dielectrics, and the variances of the electrical properties of the surrounding materials. Like the electromagnetic fields, the Variance wave begins at zero (there is no variance before the source is turned on) and is computed in the time domain until all fields reach steady state. This process is performed in a fraction of the time of a Monte Carlo simulation and yields the first two statistical parameters (mean and variance). The mean of the field is computed using the traditional FDTD equations. Variance is computed by approximating the correlation coefficients between the constituitive properties and the use of the S-FDTD equations. The impetus for this work was the simulation time it takes to perform 3D Specific Absorption Rate (SAR) FDTD analysis of the human head model for cell phone power absorption in the human head due to the proximity of a cell phone being used. In many instances, Monte Carlo analysis is not performed due to the lengthy simulation times required. With the development of S-FDTD, these statistical analyses could be performed providing valuable statistical information with this information being provided in a small fraction of the time it would take to perform a Monte Carlo analysis. |
| Type |
Text |
| Publisher |
University of Utah |
| Subject |
Delta method; EM simulation; FDTD; Finite-difference time-domain; Stochastic |
| Dissertation Institution |
University of Utah |
| Dissertation Name |
Doctor of Philosophy |
| Language |
eng |
| Rights Management |
Copyright © Steven Michael Smith 2011 |
| Format Medium |
application/pdf |
| Format Extent |
6,606,079 bytes |
| Identifier |
us-etd3,35953 |
| Source |
Original housed in Marriott Library Special Collections, QP6.5 2011 .S65 |
| ARK |
ark:/87278/s6dz0q3x |
| Setname |
ir_etd |
| ID |
194672 |
| Reference URL |
https://collections.lib.utah.edu/ark:/87278/s6dz0q3x |
