Anomaly detection for space information networks: A survey of challenges, techniques, and future directions

Space anomaly detection plays a critical role in safeguarding the integrity and reliability of space systems amid the rising tide of threats. This survey aims to deepen comprehension of space cyber threats through space threat modeling, and meticulously examine the unique challenges of space anomaly detection. The survey identifies scalability, real-time detection, limited labeled data availability, concept drift, and adversarial attacks as key challenges based on thorough literature analysis and synthesis. By extensively exploring state-of-the-art anomaly detection techniques, the study evaluates their applicability, strengths, and limitations within space networks. Going beyond analysis, a notable contribution of this work involves integrating stream-based and graph-based methods, tailored to capture the intricate temporal and structural relationships inherent in space networks. This innovative hybrid approach holds promise for heightened detection accuracy and sets the stage for future research endeavors. As space threats continue evolving in both number and sophistication, this survey timely provides insights, recommendations, and a clear roadmap for researchers, engineers, and practitioners to fortify space anomaly detection mechanisms.