Defense Against Aerial Swarm Attacks: A Stackelberg Game-Theoretic Approach

This paper presents a novel deception-aware Bayesian stackelberg game-theoretic framework for defending critical infrastructure against a swarm of uncrewed aircraft systems (UAS) comprising real and decoy attackers. By integrating probabilistic sensing and cognitive bias modeling, the proposed framework enables real-time, deception-aware interception of the attackers. Key contributions of this work include a confidence-weighted task assignment strategy, prospect-theoretic decision-making to reduce false engagements, and collision-aware pursuit under 3D constraints. Simulation results demonstrate the effectiveness of the approach across parameter variations, achieving maximum interception rate of 80% and robustness to engagements with decoy attackers. This work advances real-time UAS defense by bridging game theory, control, and cognitive modeling, providing a validated solution for real-time UAS swarm defense.