A guaranteed-convergence framework for passivity enforcement of linear macromodels

Passivity enforcement is a key step in the extraction of linear macromodels of electrical interconnects and packages for Signal and Power Integrity applications. Most state-of-the-art techniques for passivity enforcement are based on suboptimal or approximate formulations that do not guarantee convergence. We introduce in this paper a new rigorous framework that casts passivity enforcement as a convex non-smooth optimization problem. Thanks to convexity, we are able to prove convergence to the optimal solution within a finite number of steps. The effectiveness of this approach is demonstrated through various numerical examples.