Use of regressive method for RF MEMS test and diagnosis
Résumé
The test of RF equipments at their ever-increasing operating frequency range is time consuming and is therefore very expensive. This paper introduces a preliminary study of the method combining low frequency test stimuli and regression-based techniques for off-line and on-line low cost testing of RF MEMS switch. The study includes the modelling of RF MEMS switches as well as some initial ideas how to test and diagnose their functionality. The proposed approach is based on the principle of alternate test that replaces conventional specification-based testing procedures. The basic idea is to apply a low frequency alternate test signal to the RF device and to automatically extract the characteristics from the component response. These features are utilized in a regression process to predict the RF conventional specifications parameters like s-parameters. Preliminary validation results obtained on a capacitive RF switch are encouraging. An evaluation of the feasibility of implementing a built-in-test based on the proposed approach is presented.