Model-Order Reduction Techniques for Circuits and Interconnects Simulation

The basic goal of this dissertation is to extend the use of order-reduction techniques as accurate methods for analyzing complex electronic systems. The theoretical and practical aspects of moment-matching, Krylov subspace-based methods, and rational approximations techniques are studied. The moment-matching techniques are directly applied to p-n junction device equations to accurately model the carrier dynamic in the junctions. A robust approximation technique that uses Householder orthoganalization techniques is developed to generate macromodels of electromagnetic systems, such as frequency-dependent coupled transmission lines. A pole-clustering technique with inverse distance-measure criterion is used to further reduce the models. This allows the efficient accurate simulation of frequency-dependent coupled transmission lines characterized by scattering parameters and an optimal reference system. The heterogeneous reduction techniques are woven into a unified method by using network partitioning techniques. Recursive convolution is used to speedup the transient simulation. The method does not suffer from aliasing or round-off errors caused by nonband-limited frequency responses, nor by numerical transforms of a large number of points, respectively.

Name thesis:degree_name

Ph.D.

Level thesis:degree_level

Dissertation

Discipline thesis:degree_discipline

Electrical Engineering

Grantor

University of Illinois at Urbana-Champaign

Year dc:date

2015

Author dc:creator

• Beyene, Wendemagegnehu Tsegaye

Contributors dc:contributor

• Schutt-Ainé, José E.

• Mathematics

Language dc:language

eng

Identifier

(MiAaPQ)AAI9737051

OAI identifier oai:identifier

oai:www.ideals.illinois.edu:2142/81173