A methodology for analog circuit macromodeling

This work describes a complete framework for generation of compact analog circuit macromodels which significantly reduce the model complexity while still capturing the dominant linear and nonlinear response of the circuit. The technique is applicable to a broad class of circuits that exhibit weakly nonlinear behavior such as mixers, RF power amplifiers and switched-capacitor circuits. The Volterra-based circuit models are first characterized using a combination of SPICE simulation and efficient numerical fitting techniques. The complexity of the extracted circuit models is further reduced by model order reduction techniques while maintaining a high degree of accuracy. The efficacy of our macromodeling methodology is verified by comparison with SPICE simulations. The efficiency of our macromodels makes them suitable for whole-system verification and high-level design analysis.