Compact F-slot antenna architecture for high-efficiency millimeter-wave operation in certain 5G and prospective 6G wireless networks

This paper presents a compact microstrip antenna array with an F-shaped slot, designed to meet millimeter-wave frequency requirements for emerging 5G and 6G wireless networks. The proposed array measures 14 × 6 mm 2 and is implemented on a Rogers RT5880 substrate with a 50 Ω stepped feed line for impedance matching. Operating in the 37.5–38.5 GHz range, it offers a 1 GHz bandwidth and a peak gain of 10 dBi. To enhance design efficiency, supervised machine learning techniques-Gaussian Process Regression (GPR), Support Vector Machines (SVM), and Linear Regression (LR) were applied for accurate Return Loss (S11) prediction. GPR outperformed others with R 2 = 0.99, MAE = 0.23, MSE = 0.085, RMSE = 0.291, and MAPA = 99.89%. This ML-driven approach reduced simulation time and accelerated prototyping. Combining compactness, wide bandwidth, and high gain, the proposed antenna array is a strong candidate for next-generation high-frequency communication infrastructures.