Rumo à Autenticação de Dispositivos IoMT através da Classificação de Sinais RF

• Autor(a) principal: Monteiro, Stéphane Joaquim Lourenço
• Data de Publicação: 2023
• Tipo de documento: Dissertação
• Idioma: eng
• Título da fonte: Repositório Científico de Acesso Aberto de Portugal (Repositórios Cientìficos)
• Texto Completo: http://hdl.handle.net/10400.22/23710
• topic: RFF; ML; CNN; IoMT

The increasing reliance on the Internet of Medical Things (IoMT) raises great concern in terms of cybersecurity, either at the device’s physical level or at the communication and transmission level. This is particularly important as these systems process very sensitive and private data, including personal health data from multiple patients such as real-time body measurements. Due to these concerns, cybersecurity mechanisms and strategies must be in place to protect these medical systems, defending them from compromising cyberattacks. Authentication is an essential cybersecurity technique for trustworthy IoMT communications. However, current authentication methods rely on upper-layer identity verification or key-based cryptography which can be inadequate to the heterogeneous Internet of Things (IoT) environments. This thesis proposes the development of a Machine Learning (ML) method that serves as a foundation for Radio Frequency Fingerprinting (RFF) in the authentication of IoMT devices in medical applications to improve the flexibility of such mechanisms. This technique allows the authentication of medical devices by their physical layer characteristics, i.e. of their emitted signal. The development of ML models serves as the foundation for RFF, allowing it to evaluate and categorise the released signal and enable RFF authentication. Multiple feature take part of the proposed decision making process of classifying the device, which then is implemented in a medical gateway, resulting in a novel IoMT technology.