Propagation of electromagnetic waves in three-dimensional (3D) space is studied using the Transmission-Line Modelling (TLM) method. The theoretical foundation, generalization and more advanced versions of the symmetrical condensed node (SCN), a basic structure for 3D TLM modelling, are presented. A unified approach to the derivation of TLM schemes capable of modelling nonuniform anisotropic materials on an arbitrarily graded mesh is introduced. The parameters and the scattering properties for a general symmetrical condensed node (GSCN) constructed from link lines with arbitrary characteristic impedances and up to six stubs are described. For the first time in 3D TLM modelling it is shown that stubs can be removed completely from the SCN, whilst still allowing general inhomogeneous problems to be modelled. This development, described as the symmetrical super-condensed node (SSCN), contributes significant reductions in memory storage and computer run-time in TLM simulations. In order to assess accuracy and numerical dispersion in the various TLM schemes based on the GSCN, dispersion relations are studied. A general approach for the analytical expansion of the general dispersion relation is introduced, enabling the derivation of dispersion relations in algebraic form for all currently available nodes. Based on the results obtained from the dispersion analysis, work aimed at constructing an optimal node is described. A class of adaptable symmetrical condensed nodes (ASCN) is derived, with superior propagation characteristics compared to other known TLM nodes.

No description available.

Thesis (M.Eng.)--Memorial University of Newfoundland, 2001. / Restricted until June 2002. Bibliography: leaves 96-98.

Thesis (M.S. in Electrical Engineering)--Naval Postgraduate School, June 2003. / Thesis advisor(s): Douglas J. Fouts, Phillip E. Pace. Includes bibliographical references (p. 143-144). Also available online.

Protein function is the ultimate expression of the genetic code of every organism, and determining which proteins interact helps reveal their functions. MatrixMatchMaker (MMM) is a computational method of predicting protein-protein interactions that works by detecting co-evolution between pairs of proteins. Although MMM has several advanced features compared to other co-evolution-based methods, these come at the cost of high computation, and so the goal of this research is to improve the performance of MMM. First we redefine the computational problem posed by the method, and then develop a new algorithm to solve it, achieving a total speedup of 570x over the existing MMM algorithm for a biologically meaningful data set. We also develop hardware which has not yet succeeded in further improving the performance of MMM, but could serve as a platform that could lead to further gains.

The complex nature of solving microwave circuit problems with their many varied boundary conditions, precludes direct analysis utilising the basic laws involved. This thesis describes methods that have been developed to determine the defining parameters of such problems. The methods are based on numerical techniques, using a digital computer to perform the calculations. The finite difference and element techniques are reviewed briefly and extensions made into the analysis of three dimensional configurations. The difficulties of such methods are also discussed. The evolution of the steady state transmission line element method from the finite difference/element techniques is shown to possess many distinct advantages over the more conventional techniques, notably that of the ease in which it is formulated. Examples of rectangular, circular and elliptical-wave guide analysis are shown, and comparisons formed with finite difference/element analysis where necessary. Proposals are also introduced whereby the method may be utilised to provide a wide range of microwave characteristics, with little or no alterations to existing procedures. The adaptation to other fields of interest, such as those associated with structural or fluid dynamics was also briefly noted.

Protein function is the ultimate expression of the genetic code of every organism, and determining which proteins interact helps reveal their functions. MatrixMatchMaker (MMM) is a computational method of predicting protein-protein interactions that works by detecting co-evolution between pairs of proteins. Although MMM has several advanced features compared to other co-evolution-based methods, these come at the cost of high computation, and so the goal of this research is to improve the performance of MMM. First we redefine the computational problem posed by the method, and then develop a new algorithm to solve it, achieving a total speedup of 570x over the existing MMM algorithm for a biologically meaningful data set. We also develop hardware which has not yet succeeded in further improving the performance of MMM, but could serve as a platform that could lead to further gains.

Thesis (M.S.)--Virginia Polytechnic Institute and State University, 1990. / Vita. Abstract. Includes bibliographical references (leaves 74-75). Also available via the Internet.

Thesis (M.S.)--Virginia Polytechnic Institute and State University, 1994. / Vita. Abstract. Includes bibliographical references (leaves 95-98). Also available via the Internet.

Thesis (M.S.)--Virginia Polytechnic Institute and State University, 1994. / Vita. Abstract. Includes bibliographical references (leaves 126-128). Also available via the Internet.