Quad port MIMO antenna design for WLAN and N78 application using regression machine learning validation

Abstract

This paper presents the design and analysis of a compact four-port MIMO antenna intended for 5G communication systems, specifically targeting the n78 band (3.5 GHz) and the WLAN band (5.25 GHz). The proposed antenna has dimensions of 88.2 mm × 88.2 mm (1.03λ 0 × 1.03λ 0 ). It achieves peak gains of 6.22 dB and 6.55 dB at its respective resonance frequencies, with a high isolation of − 28 dB between ports. The antenna covers bandwidths of 0.3 GHz (8.70%) and 0.37 GHz (7.07%) at 3.5 GHz and 5.25 GHz, respectively, with an Envelope Correlation Coefficient (ECC) of 0.006 and Diversity Gain (DG) of 9.99 dB, ensuring excellent MIMO performance. The antenna is simulated using CST Studio Suite and designed to offer high gain and low return loss, suitable for 5G and WLAN applications. Additionally, an RLC equivalent circuit model is developed in the ADS Agilent software, where the resonance frequencies closely match those obtained from full-wave simulations. To further validate and predict the antenna’s performance characteristics, a supervised machine learning (ML) regression model is applied. The model’s accuracy is evaluated using several performance metrics, including the variance score, R-squared (R 2 ), mean squared error (MSE), mean absolute error (MAE), root mean square error (RMSE), and mean squared logarithmic error (MSLE).