Abstract
Multi-objective evolutionary algorithms have become the most efficient ap- proach for solving real-world problems, where performance depends on the convergence speed, balance between exploitation and exploration, and popu- lation diversity. Self-adaptive techniques are effective for dynamically adjust- ing hyperparameter algorithms. This paper presents an automated methodology for designing and optimizing a complementary split-ring resonator (CSRR)- incorporated microstrip patch antenna (MPA) using a fuzzy adaptive multi- objective genetic algorithm (FAMOGA). A double-ring CSRR structure etched into the ground plane introduces multi-band resonance frequencies, extending antenna coverage to the multi-bands frequencies used in internet of things (IoT) and Wi-Fi applications. The algorithm simultaneously determines the optimal CSRR ring positions, minimizes the antenna size, and satisfies design con- straints. Fuzzy logic controllers evaluate the fitness function and dynamically adapt the crossover and mutation rates, guiding the genetic algorithm toward feasible, high-performance solutions. The FAMOGA converges to the 7th gen- eration with a population of four individuals. At the three target frequencies, the optimized antenna achieves reflection coefficients of −11.08 dB, −39.25 dB, and −30.61 dB , and realized gains of 3.53 dBi, 4.16 dBi, and 6.64 dBi at 2.422 GHz , 3.646 GHz , and 5.259 GHz , correspondingly. The simulation demonstrated that FAMOGA provides a computationally efficient approach for multi-band antenna design without manual parameter tuning.