RF varactor design based on evolutionary algorithms

This paper introduces an optimization methodology for the design of RF varactors. The characterization of the varactor behaviour is supported by a set of equations based on technological parameters, granting the accuracy of the results, as well as the adaptability of the model to any technology. The varactor design is achieved through the implementation of a Genetic Algorithms (GA) optimization methodology, which is able to deal with continuous and/or discrete variables, making possible to suit both technological and layout constraints. A set of working examples for UMC130 technology are addressed. The results presented, spotlight the potential of varactor analytical model, combined with a GA optimization procedure, when integrated in optimization design tools. The accuracy of the results is checked against HSPICE simulator.