Collaborative intrusion detection system for unmanned aerial vehicles swarm security

• Ano de defesa: 2024
• Autor(a) principal: Silva, Leandro Marcos da
• Tipo de documento: Dissertação
• Tipo de acesso: Acesso aberto
• Idioma: eng
• Instituição de defesa: Biblioteca Digitais de Teses e Dissertações da USP
• Palavras-chave em Português: Aprendizado de máquina; Colaborativo; Collaborative; Cybersecurity; Detecção de intrusão; Intrusion detection; Machine learning; Segurança cibernética; Unmanned Aerial vehicle; Veículo aéreo não tripulado
• Link de acesso: https://www.teses.usp.br/teses/disponiveis/55/55134/tde-17042025-160925/

Unmanned Aerial Vehicles (UAVs), also known as drones, have become increasingly popular in various applications, including military, civil, and commercial. However, these vehicles are vulnerable to cyberattacks that can compromise security and privacy and cause physical harm. These attacks can involve signal interception, unauthorized access, data theft, and even remote control of the UAV, among others. Therefore, UAV manufacturers and users need to be aware of the threats and take appropriate measures to raise levels of security and integrity. One of the preventive measures is the use of Intrusion Detection System (IDS), which monitors system settings, data files, and network transmission to identify any abnormal behavior. Upon detection, the IDS notifies the ground control station so that a decision can be made. UAV IDS generally focus attacks on a specific data source without mentioning applying to a swarm. In this sense, this work presents the REMY collaboRactive intrusion dEtection system for unManned aerial vehicles swarm SecuritY. REMY detects attacks on the UAV network and in-flight anomalies, applying machine learning techniques through supervised and unsupervised learning. The threats identified in the network are blackhole, grayhole, and flooding, and in flight, GPS spoofing and jamming. Federated learning is also present in REMY to ensure data privacy and training collaboration between UAVs. Furthermore, geographic and physical characteristics are considered to make the IDS operation independent of the hardware. Finally, aspects of a lightweight IDS are considered in the development, such as low energy consumption and the use of hardware resources.