Machine learning assisted design of a square loop wearable antenna for efficient medical body area networks
摘要
In recent years, wearable antennas gain significant attention for use in healthcare applications, particularly in Implantable Medical Devices (IMDs) and Medical Body Area Networks. These antennas must be small, lightweight, and body-conforming to meet the needs of these applications. As the wearable antennas play a crucial role in Medical Body Area Networks (MBAN), facilitating continuous health monitoring. Their integration in IMDs requires minimizing electromagnetic interference and optimizing radiation characteristics. As healthcare systems demand more efficient and precise devices, the development of wearable antennas becomes essential for enhanced performance. The development of such antenna has gathered substantial awareness in recent years in the telemedicine industry. This study proposes a two-dimensional square loop-based antenna design, operating at 2.4 GHz with improved radiation properties and reduced backward radiation. The antenna dimensions are 60 × 40 × 0.7 mm3, making it compact and effective for body-worn applications. The antenna design process incorporates Machine Learning (ML) techniques to minimize simulation time, increase efficiency, and enhance design accuracy. ML algorithms optimize the antenna’s performance, particularly in terms of reflection coefficient, bandwidth, and gain. The proposed antenna design has tackled the issues faced by the conventional antenna and holds promise for real-time applications in healthcare, military, sports, and identification systems. It addresses critical challenges such as Specific Absorption Rate (SAR) and efficiency, ensuring optimal performance when interacting with human body tissues. Moreover, the experimental results demonstrate that the simulated and fabricated results exhibit similar deviations, confirming that the antenna is suitable for real-world applications.