Optimizing a Dual-band Antenna for IoT Systems by Machine Learning Algorithms

The IoT system has developed strongly with the appearance of many compact devices connected through diverse transmission technologies and using different popular frequency bands. However, designing and optimizing compact, multi-band antennas for IoT systems using traditional methods requires significant time and effort. Machine Learning (ML) is emerging as a new trend in antenna research, supporting the design and optimization process. Current research often focuses only on optimizing the performance of single-band antennas using traditional structures such as circular and rectangular patch antennas. It requires manual calculation of antenna size parameters. This paper proposed a new method of optimizing the antenna's size parameter, frequency, and bandwidth called Machine Learning Antenna (MLA). MLA uses machine learning algorithms to calculate the size of the dual-band antenna from the desired resonant frequencies and bandwidths with a mean squared error (MSE) of above 0.94. The dual-band antenna for the IoT system is designed using the MLA method with a small size of 30 mm x 40 mm, using an FR-4 substrate with a thickness of 1.6 mm. This antenna uses Defected Ground Structure (DGS) and shorting pin methods to reduce the size, operating flexibly on the 2.4 GHz, 5 GHz, and 6 GHz frequency bands, suitable for IoT, IoV, WiFi 6, WiFi 7, 4G, 5G systems.