Inverse Design of Rectangular Microstrip Patch Antenna using Neural Network Combining with Time-Domain Representation of S-parameters

In this study, a multilayer perceptron (MLP) neural network is utilized to inversely design a rectangular microstrip patch antenna. The MLP network maps the input features, including the resonant frequency, directivity, realized gain, and time-domain representation of the S-parameters, to the output that is the antenna geometrical parameters. Especially with the time-domain representation of S-parameters added into the input features, the prediction error of the antenna geometrical parameters can be decreased dramatically. The mean square error (MSE) and mean absolute error (MAE) are 0.00432% and 0.484% for the test dataset, respectively. The promising results verified that incorporating the time-domain representation of S-parameters into the features could effectively improve the prediction accuracy of the MLP network for inverse antenna design.