Advances in Artificial Neural Network Assisted Microwave Sensors for Material Characterization

This article reviews the current advances in planar microwave sensors assisted by artificial neural networks (ANNs) for various dielectric material characterization applications. The architecture of ANNs and intelligent sensing systems are introduced. The ability of ANNs to improve the accuracy of microwave sensors in measuring samples with multiple components and under various ambient conditions is discussed. ANNs enable the determination of complex and nonlinear relationships between sensing data and desired output data, such as complex permittivity, concentrations, and defect detection. More advanced neural network algorithms facilitate the training of more complex datasets. Finally, the current challenges and future prospects of ANN-integrated microwave sensors are discussed. It is envisaged that the summarized research efforts can provide valuable insights for scientists and engineers aiming to implement ANNs in microwave sensors.