TinyML-Driven Cybersecurity for Autonomous Spacecraft: Latency-Accuracy Analysis for SPARTA RF and Cyber Threat Detection
arXiv:2606.05779v1 Announce Type: cross Abstract: Autonomous spacecraft require rapid, lightweight, and reliable onboard detection of cyber-RF threats. Using the SPARTA attack model, we analyze the latency-accuracy trade-offs of TinyML-compatible classical models -- Random Forest, Logistic Regression, SVM, and MLP -- for detecting uplink jamming, Fake-NR spoofing, payload manipulation, ground-segment compromise, and unauthorized command injection. We present a physics-informed theoretical analysis of each model's computational complexity, VC dimension, Lipschitz continuity, and latency scaling
The increasing reliance on autonomous spacecraft and the escalating cyber threat landscape necessitate advanced, on-board defensive capabilities.
This research provides a framework for integrating TinyML into critical space infrastructure, addressing the unique computational and security constraints of autonomous systems in orbit.
The feasibility of deploying highly efficient, AI-driven cybersecurity directly on resource-constrained spacecraft is significantly advanced, enabling real-time threat detection without ground interaction.
- · Spacecraft manufacturers
- · Defense contractors
- · AI/ML chip developers
- · Satellite operators
- · Adversarial space powers
- · Traditional ground-based security models
Autonomous spacecraft will become more resilient to diverse cyber and RF threats, improving mission success rates.
This will accelerate the development and deployment of more complex autonomous space missions, reducing operational costs and increasing strategic capabilities.
The integration of TinyML for cybersecurity in space could lead to analogous applications in other remote, resource-constrained critical infrastructure sectors on Earth.
This signal links to a primary source. Continuum Brief monitors and indexes it as part of the live intelligence stream — we do not republish source content.
Read at arXiv cs.AI