Design and Optimization of Stacked Wideband On-Body Antenna with Parasitic Elements and Defected Ground Structure for...

The ability to measure vital signs using electromagnetic waves has been extensively investigated as a less intrusive method capable of assessing different biosignal sources while using a single device. On-body antennas, when directly coupled to the human body, offer a comfortable and effective alternative for daily monitoring. Nonetheless, on-body antennas are challenging to design primarily due to the high dielectric constant of body tissues. While the simulation process may often include a body model, a unique model cannot account for inter-individual variability, leading to discrepancies in measured antenna parameters. A potential solution is to increase the antenna's bandwidth, guaranteeing the antenna's impedance matching and robustness for all users. This work describes a new on-body microstrip antenna having a stacked structure with parasitic elements, designed and optimized using artificial intelligence (AI). By using an AI-driven design approach, a self-adaptive Bayesian neural network surrogate-model-assisted differential evolution for antenna optimization (SB-SADEA) method to be specific, and a stacked structure having parasitic elements and a defected ground structure with 27 tuneable design parameters, the simulated impedance bandwidth of the on-body antenna was successfully enhanced from 150 MHz to 1.3 GHz, while employing a single and simplified body model in the simulation process. Furthermore, the impact of inter-individual variability on the measured S-parameters was analyzed. The measured results relative to ten subjects revealed that for certain subjects, the SB-SADEA-optimized antenna's bandwidth reached 1.6 GHz.