XG-boost-based optimization of corrugated arm MEMS switch for improved radio frequency performance

Radio frequency micro electro mechanical (RF MEMS) Switches with structural changes in membrane for lower pull-down voltages, high isolation and low insertion loss are required for switching over several GHz range. A corrugated arm RF MEMS switch for a low-actuation voltage with low residual stress has already been fabricated and tested for mechanical characteristics. The goal of this paper is to make the corrugated arm switch suitable for RF applications by modelling the switch in both UP and DOWN states using XG-Boost approach. Using XG boost approach the desired gap(g) and frequency of operation(f) for better Radio Frequency parameters such as isolation loss and insertion loss is predicted and reported. Time-consuming optimization methods to predict the S-parameters in customized EM simulators such as HFSS, CST have been overcome using the proposed XG Boost regression model. In this paper, the proposed regressor approach reduces the design time of the switch by 99.4048% compared to the conventional EM simulator which takes 5.33 h in an Intel-Core(2.4 GHz) CPU with 8 GB of RAM. Validation is done by comparing with the HFSS model and good agreement is obtained. The simulation of the proposed switch results in an insertion loss of − 0.12 dB and isolation of − 57.7 dB at 40 GHz. The proposed XGBoost Regressor model predicts the RF performance of the switch with the accuracy of 99.98%.