Innovative Autonomous Design of Multi-Band Reconfigurable Antenna Using Genetic Algorithms

This paper presents a novel approach to designing reconfigurable antenna using genetic algorithms (GAs) and accurate PIN diode modeling. The method integrates TRL calibration for precise diode characterization, equivalent circuit modeling, and a multi-objective GA implemented in CST Microwave Studio. A pixelization-based optimization technique is employed to determine optimal diode positions in a Defected Ground Structure (DGS). The proposed antenna operates in four distinct modes, achieved through strategic placement of two PIN diodes. Capacitive slot integration eliminates the need for separate bias circuits, making the design simpler and less complex. The results demonstrate good agreement between simulated and measured performance metrics, with maximum gains ranging from 3 to 4.34 dBi and Half-Power Beamwidths (HPBW) exceeding 100 degrees across all frequencies. This automated design methodology accelerates optimization, improves the antenna adaptability and performance, and ensures its compatibility with modern wireless communication systems.