Fundamental aspects of large switched-capacitor networks

Fast image-processing networks using reciprocal resistive elements can be implemented using switched-capacitor (SC) techniques. This paper establishes a rigorous theoretical basis for mapping real VLSI SC networks to ideal SC networks and globally asymptotically stable discrete-time systems. These new results encompass several general classes of practical and useful SC networks, and include stability and uniqueness of the networks after general switch-closing events, mapping theorems and rules, and stability of the associated discrete-time systems. We also prove several general results based on an energy viewpoint, such as energy transfer efficiency and uniqueness of transferred source charges