A causal model for linear RF systems developed from frequency-domain measured data

Abstract

With the ever-growing complexity of interconnect networks, models developed from measured data or data from 3-D electromagnetic simulators are increasingly becoming essential. It is to this end that the current contribution is directed. In particular, it focuses on the development of a model via a Fourier series expansion (FSE) approach. Its primary advantage is that the response in the time domain can be explicitly obtained in a simple form for an arbitrary input using only a set of FSE coefficients. Also, it guarantees causality without requiring a numerical implementation of a Hilbert transform. The end result is a causal and stable time-domain representation of a system that may subsequently be used in a time-domain simulator such as SPICE.