A PDN Modeling Method of Conduction Immunity in Mixed Analog-Digital Integrated Circuits Based on Dominant Pole Vector Fitting Algorithm

The conduction sensitivity modeling of Mixed Analog-Digital Integrated Circuit (MADIC) mainly relies on the international standard ICIM-CI Behavioral Level Model. Its passive distribution network (PDN) has the problem that it is difficult to synthesize circuits with complex matrices when vector fitting method is used. In this paper, based on the impedance transmission characteristics of analog circuits and the theory of dominant poles, an improved vector fitting algorithm based on real poles is proposed to convert conjugate complex poles into real poles for solving. Compared with the traditional method of realizing the similarity matrix, the generated circuit achieves about 30% simplification under the premise of guaranteeing the fitting accuracy.