Once confined to very specific applications, satellite navigation systems, also known as GNSS (Global Navigation Satellite Systems), are now widespread and can be found in numerous devices, including some sensitive systems such as autonomous vehicles and military applications. Alongside the rapid expansion of this technology, the frequency and complexity of attacks against these systems have also increased through techniques known as jamming and spoofing. To mitigate equipment vulnerabilities, most receivers incorporate algorithms to detect such attacks by analyzing various signals and metrics provided by the system. The LS2PR laboratory at LETI is developing an innovative approach to address this issue using artificial intelligence (AI) techniques to detect and classify such attacks with very short response times. However, the effectiveness of these techniques depends on the quality of the database used for training, which is further complicated by the difficulty of implementing attacks involving complex equipment. Leveraging the knowledge of CEA and its partner teams, the candidate should: Conduct a state-of-the-art review of different types of GNSS attacks. Acquire or enhance their knowledge of the use of relevant systems (GNSS receivers, signal generators like Skydel). Define a testing protocol that considers feasibility constraints and the learning objectives (learning biases, etc.). Implement the testing setup, including equipment setup, configuration, and programming. Collect data. Perform post-processing and statistical analysis of the collected data to validate their suitability for machine learning. Prepare reports and participate in project meetings. Publish their work in high-quality conferences and journals
English Intermediate
Talent impulse, the scientific and technical job board of CEA's Technology Research Division
© Copyright 2023 – CEA – TALENT IMPULSE - All rights reserved