Design and Comparative Analysis of Microstrip Antenna Using Neural Networks

In this paper, a methodology is proposed to design the pattern of a microstrip antenna using the approach of generating synthetic data based on Generative Adversarial Networks (GANs) and Multilayer Perceptron (MLP). Initially, we provide the model with a dataset developed with various geometries associated with the RO303 substrate. With this, the generator will be able to synthesize new samples that resemble the training set, and it can learn the patterns and relationships between the geometries and their characteristics. One of the main advantages of this approach is the ability to quickly explore a wide range of antenna designs. By using GANs, we can generate multiple antenna geometries in a short period, enabling an efficient search for the best solutions that meet the specific project requirements.