Compact LoRa Patch Antenna Optimization Using Dual Random Starfish Aggregation Coupled Transformer Network for Vital Sign Detection in Breast Cancer WBANs
摘要
The rapid advancement of Wireless Body Area Networks (WBANs) has created a growing demand for compact, efficient, and reliable antenna systems to support continuous health monitoring, particularly for breast cancer applications. Recent methods, including CPW-fed patch antennas, artificial neural network (ANN)-driven models, and wearable textile antennas, have improved antenna design automation and flexibility. However, challenges such as signal distortion from body proximity, gain reduction under bending, Specific Absorption Rate (SAR) compliance, and lack of adaptive tuning continue to limit practical deployment. To overcome these limitations, this study presents a compact LoRa patch antenna optimized using a novel Dual Random Starfish Aggregation Coupled Transformer Network (Dual-Ran-SACTN) framework. This system combines the Starfish Optimization Algorithm (SFOA), a Random-Coupled Neural Network (RCCN), and a Dual-Aggregation Transformer Network (DuAT) to enhance convergence speed and learning efficiency. The antenna, designed in CST Microwave Studio, measures only 80 × 60 mm2 (0.23 λ₀ × 0.17 λ₀), offering a lightweight and wearable structure for continuous vital sign monitoring. The proposed model exhibits a bidirectional radiation pattern in the E-plane and an omnidirectional pattern in the H-plane, achieving a peak gain of 2.12 dBi and a high radiation efficiency of 99.8% at 868 MHz. Additionally, the design maintains low SAR and stable performance under bending, making it robust for wearable WBAN applications. This work offers a real-time, energy-efficient solution for intelligent breast cancer monitoring through adaptive antenna optimization. This model supports practical applications such as continuous breast cancer monitoring, wearable health diagnostics, and real-time WBAN-based physiological signal tracking.