Design of 2X2 Microstrip Patch Antenna Array and Optimization of Bandwidth using Efficient Machine Learning Technique

Abstract

Microstrip patch antenna plays key role in the wireless communication. The research is going on to design and optimization of the antenna for various advance application such as 5G and IoT. Artificial intelligence based techniques such as machine learning (ML) is capable to optimize the parameter values and make prediction model. This paper presents a design of 2×2 microstrip patch array antenna and optimization of bandwidth using efficient machine learning technique. The copper material is used to design the top and patch and FR4 Epoxy substrate is used to design the bottom of the antenna. The random forest regressor machine learning technique is used to optimize the antenna parameter such as bandwidth. The antenna designing is performed using the CST software and optimization is performed using the Python spyder 3.7 software. Simulation results show that the bandwidth is achieved 172.11 MHz, return loss is -46.61 dB and resonant frequencies are 4.915GHz and 6.018GHz. The optimization of performance is calculated in terms of accuracy, mean absolute error and mean squared error. Proposed random forest ML technique is achieved 99.56% accuracy in the antenna parameters prediction model.