arXiv Paper Proposes Temporal and Multimodal Deep Learning for LEO Satellite Cyberattack Detection

A new arXiv preprint presents a deep learning approach for spotting cyberattacks in Low-Earth Orbit satellite communication networks, which face constantly shifting and complex conditions. The method combines temporal and multimodal modeling, departing from standard network intrusion detection techniques built for more static terrestrial environments. The work is a cross-listed submission and has not yet been peer reviewed.