A REVIEW PAPER ON MICROSTRIP PATCH ANTENNA DESIGN USING NEURAL NETWORKS

Abstract

In the microwave community, artificial neural network (ANN) modelling has recently acquired huge attention across the globe, especially in the analysis and synthesis of microstrip antennas (MSAs) due to their generalisation and adaptability characteristics. A trained neural model, that is almost identical to its measured as well as simulated equivalent, determines the response very fast. It also completely bypasses the repeated use of traditional models, as these models demand to be re-discretized for any small geometry change, which is a time-consuming exercise itself. The ultimate objective of this survey paper is to comprehensively study this emerging field for both analysis as well as the synthesis of the MSAs. Some of the untouched cases, that are basically required for the antenna designers to be resolved, are also observed during the review procedure. Unique and efficient solutions rooted on the neural networks are suggested for these instances. The existing proposed neural methods are also verified by the prototypes' fabrication and characterization.