Design and optimization of miniature antennas for biomedical applications using artificial neural networks

Abstract

Biomedical antennas play a vital role in modern healthcare technologies, supporting implantable devices, wearable sensors, and telemedicine through reliable wireless communication. These antennas require compact size, high efficiency, and strict specific absorption rate compliance for safe operation. This thesis introduces an i nverseartificial neural network approach combined with multi-objective optimization to enhance the design process of biomedical antennas. The proposed method enables direct prediction of antenna geometries from target performance metrics, reducing computational time while maintaining high accuracy. Case studies on ultra-wideband and dual-band antennas demonstrate the effectiveness of the proposed models in achieving miniaturization, impedance matching, and specific absorption rate optimization.