Nonlinear Behavioral Modeling of Passive RFID-Transponder-Frontends

Abstract

A nonlinear RFID frontend behavioral model has been developed. By using this model the simulation time for inlay optimizations can be decreased considerably and the models hide the IC manufacturer's intellectual properties. The inherent model order reduction is based on nonlinear state-space mapping using derivative coordinates and harmonic-balance simulations. Feedforward multi-layer-perceptron artificial-neural-networks have been used for the nonlinear multivariate system mapping. The behavioral modeling of RF power rectification, RF voltage limiting and backscatter modulation is demonstrated for a modern passive UHF-RFID CMOS frontend.