Automated Microwave Amplifier Synthesis Design via Dynamic Bayesian Optimization and Tuning Space Mapping

Traditional microwave amplifier (MWA) design relies on manual topology selection and iterative parameter tuning, which is both time-consuming and susceptible to local optima. The presence of multiple, often conflicting, design objectives further increases the complexity of the process. This work proposes a fully automated MWA design framework that integrates three key innovations: 1) a topology generation algorithm combining sequential search with Bayesian optimization (BO) to efficiently explore matching networks; 2) a hybrid dynamic step BO (HDSBO) scheme that accelerates high-dimensional parameter tuning while balancing global and local search; and 3) the tuning space mapping (TSM) technique that bridges schematic and layout responses with only a few electromagnetic (EM) simulations. The proposed flow is validated on a low-noise amplifier (LNA). Compared with conventional workflows, the method significantly reduces design iterations and simulation cost, enabling automated and reliable design.