A practical large-signal global modeling simulation of a microwave amplifier using artificial neural network

We present a new technique to obtain large-signal global modeling simulation of a MMIC amplifier. The active device is modeled with a neural network trained with data obtained from a full hydrodynamic model. This neural network describes the nonlinearities of the equivalent circuit parameters of a MESFET implemented in an extended Finite Difference Time Domain (FDTD) mesh. We successfully represented the transistor characteristics with a one-hidden-layer neural network whose inputs are the gate voltage V_gs, and the drain voltage V_ds. Small-signal simulation is performed and validated by comparison with HP-Libra. Then, the large signal behavior is obtained, which demonstrates the successful use of artificial neural network (ANN) in the FDTD marching time algorithm