Artificial Neural Network Based Selection Method for Antenna Parameters in Microwave Application

The usage of machine learning in the field of microwave engineering has increased. This project proposes a design optimized model in order to achieve parameter optimization of 2 × 2 square patch antenna using Artificial Neural Network (ANN) Technique. An artificial neural network was trained to get the optimum value of the antenna parameters in less time. The input of the model is length, width of the square patch and range of antenna parameter value, the output is S-parameters, VSWR. This was made a perfect choice for communication systems engineers due to its low weight, low cost. The formulations are used to calculate the neural network cost function. The proposed neural network can tune and get the best value for the parameter. Then with the help of a machine learning algorithm, geometrical parameters of the patch antenna were designed and optimized. The substrate called fr4 is used. Hence a 2 × 2 square patch antenna element was optimized, designed, and developed.