Semi-symbolic modeling and simulation of circuits and systems

Shrinking microelectronic circuits leads to increasing parameter variations. Verification by Monte Carlo and worst case analysis has a limited reliability or requires a very high number of simulation runs. In this paper we give an overview of the semi-symbolic simulation approach. Compared to Monte Carlo and worst case analysis, the semi-symbolic simulation needs only a single simulation run to compute all possible results for given stimuli and parameter variations. For semi-symbolic simulation we model parameter variations and tolerances by symbols. Then, we compute the transient analysis using affine arithmetic. As results of transient analysis we get the system quantities as affine expressions. The affine expressions describe the impact of the parameter variations to the quantities