Predicting the Maximum Achievable Antenna Bandwidth and Efficiency Using Machine Learning: a Terminal-Integrated Meander IFA Case Study

In this paper, an approach based on Machine Learning (ML) to predict the maximum achievable performance (fractional bandwidth and total efficiency) of a printed Inverted F-antennas (IFAs) integrated into compact IoT terminals is proposed. This original approach relies on the use of a Multi-Layer Perceptron (MLP) artificial neural network (ANN), which is trained using data from numerical simulations that take into account the constraints of practical implementations. The effectiveness of the approach is demonstrated through comparisons with numerical and experimental results, as well as with theoretical results available in the literature. The obtained results show that the proposed supervised regression ML model is capable of predicting the maximum achievable fractional bandwidth and total efficiency, with an accuracy of 95.9% and 98.6%, respectively.