Here’s a summary if you prefer to check it out first:
The quality of time domain simulation results depends on the quality of the Sparameter models used. Causality is shown to be an important parameter limiting the usefulness of a model. In this paper we demonstrate that causality problems can be classified as mathematical and physical in origin. Mathematical non-causalities are caused by discretization and truncation of S-parameters models while physical non-causalities are caused by simulations and/or measurement inaccuracies and noise.
It will be shown how mathematical non-causalities can be controlled but not eliminated. Their impact on simulation results will be discussed. Finally it will be demonstrated that physical non-causalities can be eliminated by using a carefully designed causality enforcement algorithm.