Analyzing equilateral triangle compact microstrip antennas using Gaussian process regression for telemedicine and mobile biomedical imaging systems
[ X ]
Tarih
2023
Yazarlar
Dergi Başlığı
Dergi ISSN
Cilt Başlığı
Yayıncı
Springer
Erişim Hakkı
info:eu-repo/semantics/closedAccess
Özet
Antennas are vital in the internet of things (IoT) for enabling telemedicine and healthcare communication between devices and networks. They receive and transmit signals, extending range, improving efficiency, and reducing power consumption. Antennas are versatile and can be integrated into devices or added as external modules. Their flexibility and adaptability are important in applications involving humans, as they can bend and conform to the shape of the body. Overall, antennas are a crucial and adaptable component of IoT technology. The first thing that needs to be done is to determine the frequency at which the antenna should operate for the problem at hand and design an antenna that can work at those resonant frequencies. In this study, equilateral triangular-shaped compact microstrip antennas (ETMAs) were chosen, and their resonance frequencies were calculated using the Gaussian process regression method (GPR). For this purpose, 630 ETMA were simulated, and a dataset was created utilizing the antenna characteristics and resonant frequencies. Support vector machines (SVM), artificial neural networks (ANN), and GPR models were trained on the obtained data set. To validate the performance of the trained models, two ETMAs with an outer length of 50 mm and an inner slot length of 5 mm were fabricated utilizing polylactic acid (PLA) and felt-based substrates with copper tape as the conducting material. The accuracy of the resonant frequency estimation using the GPR approach for the fabricated antennas is 2.833% and 1.706% for the PLA- and felt-based antennas, respectively, when compared to the measurement results. The GPR model trained in this study has an accuracy of 0.470% and 0.662% when compared to simulations in the literature and measurement results, respectively. In addition, one of the designed antennas is in wearable form, and the other is PLA, produced with a low-cost 3D printer, allowing continuous monitoring of patients with high cancer risk. In this article, an easier and cheaper microstrip patch antenna that can be used for imaging and telemedicine applications is designed with a copper band on one flexible and one rigid substrate, and its performance is analyzed experimentally.
Açıklama
Anahtar Kelimeler
Artificial neural networks, Equilateral triangle microstrip antennas, Flexible antennas, Healthcare, Gaussian process regression, Resonant frequency estimation, Support vector machines, Telemedicine
Kaynak
Multimedia Tools and Applications
WoS Q Değeri
Q2
Scopus Q Değeri
Q1
