Securing Unmanned Devices in Critical Infrastructure: A Survey of Hardware, Network, and Swarm Intelligence

As Unmanned Aerial Vehicles (UAVs) become integral to critical infrastructure, ranging from precision agriculture to emergency disaster recovery, their security becomes a matter of systemic resilience. This paper provides a comprehensive thematic survey of the security landscape for unmanned devices, bridging the gap between low-level hardware vulnerabilities and high-level mission failures. We propose a multidimensional taxonomy that categorizes challenges into hardware roots of trust, swarm intelligence threats, and domain-specific applications. A primary focus is placed on the Resource–Security Paradox, where the energy cost of heavy cryptographic or AI defenses directly reduces flight endurance, creating a trade-off that adversaries exploit through battery-exhaustion attacks. Beyond standard threats, we analyze emerging risks in additive manufacturing supply chains, the “Sim-to-Real” gap in AI-driven perception, and the legal necessity of Digital Forensic Readiness (DFR) for post-incident attribution. Through a systematic review of defensive frameworks, including lightweight encryption, Mamba-KAN anomaly detection, and blockchain-anchored logging, we evaluate the effectiveness of current solutions against complex adversarial models. Finally, we identify critical research gaps, providing a roadmap for security-by-design in the next generation of critical infrastructure swarms.