A study by Luís Miguel Pires, lecturer at the School of Engineering and Technologies, has been published in an international scientific journal.
IPLUSO has strengthened its presence in international scientific research with the publication of a new study by Professor and Researcher Luís Miguel Pires, a lecturer at the School of Engineering and Technologies (EET), in the Electronics (MDPI) journal. The publication highlights the institution's commitment to advancing knowledge in telecommunications, artificial intelligence and future 6G networks.
The scientific paper, entitled Physics-Supported Linear and Nonlinear Dimensionality Reduction for Supervised Adaptive Channel Selection in Hybrid RF-FSO-THz Communication Systems, presents a novel framework that combines physics-based models with machine learning techniques to support the adaptive selection of the most suitable communication channel in hybrid Radio Frequency (RF), Free Space Optical (FSO) and Terahertz (THz) communication systems.
Developed by Luís Miguel Pires, the research focuses on technologies considered essential for the next generation of mobile networks (6G) and the Internet of Things (IoT), addressing the connectivity, efficiency and performance challenges of future communication infrastructures.
As part of the study, a physics-based simulation model was developed to represent different environmental conditions, including rain, fog and high humidity. This model was then used to generate a dataset containing thousands of samples, enabling the evaluation of dimensionality reduction techniques and supervised machine learning algorithms.
The results show that Principal Component Analysis (PCA) is capable of maintaining accuracy levels very close to those achieved using the full set of variables, while significantly reducing computational complexity and preserving a high degree of physical interpretability.
