High‐Isolation mmWave MIMO Antenna With Mutual Coupling Suppression and AI‐Based Performance Prediction for 5G

ABSTRACT The development of fifth‐generation (5G) wireless communication systems requires low‐latency, high‐speed, and high‐capacity data transmission, particularly in the mmWave bands. In this manuscript, a High‐Isolation mmWave MIMO Antenna with Mutual Coupling Suppression and AI‐Based Performance Prediction for 5G (MIMO‐APP‐ACVGNN) is proposed. The broadband MIMO antenna arrays were designed on a Rogers 5880 substrate, targeting high‐band 5G mmWave frequencies. The validated antenna configuration is given to the Adaptive Convergent Viewable Graph Neural Network (ACVGNN), which is used for antenna performance prediction in broadband MIMO antenna arrays. To enhance accuracy, the Superb Fairy‐wren Optimization Algorithm (SFWOA) is employed to optimize the weight parameter of ACVGNN, which precisely predicts the antenna performance. The proposed AGP‐MIMO‐ACVGNN is examined using performance metrics, such as transmission coefficient, efficiency, MSE, MAE, RMSE, R 2 , and explained variance score. The proposed AGP‐MIMO‐ACVGNN attains 7.28%, 9.68%, and 6.79% higher efficiency when compared with existing techniques such as ML‐MIMO‐HEP, MIMO‐5Gmm‐ML, and MIMO‐MIMW‐NG6G, respectively.