Propagation-Aware Supervisory Control for Resilient Autonomous Aerial Teams

Autonomous aerial teams operating in contested environments must preserve both collective coherence and mission-reference correctness when coordination data become unreliable. This study develops Propagation-Aware Stabilization (PAS), a supervisory control layer combining independently cross-checked anomaly scoring, continuous coupling modulation, selective link interruption, and adaptive damping. A reproducible two-dimensional point-mass study compared fixed consensus, weighted mean-subsequence-reduced consensus (W-MSR), PAS ablations, and full PAS across 3,750 Monte Carlo trials. The advanced campaign incorporated moving communication topology, directed burst loss, bounded communication delay, correlated receiver bias, progressive attrition, overt spoofing, stealth spoofing, and shared cross-channel compromise. Under overt spoofing, PAS reduced peak formation error by 94.2% relative to fixed consensus and 89.4% relative to W-MSR. It reduced mission-centroid drift by 98.6% and 96.7%, respectively, and achieved 98% mission success. Under stealth spoofing and dual-channel compromise, however, mission success fell to 8% and 0%, despite comparatively coherent formations. The results demonstrate that PAS can contain detectable corrupted influence but also establish an important failure boundary: formation coherence does not guarantee mission correctness when transmitted state and its nominally independent cross-check share a failure mode. The paper proposes separate coherence and mission-correctness monitoring, diversity-preserving trust anchors, and hybrid fallback logic as requirements for further development. This work is an unclassified, platform-agnostic computational proof of concept using synthetic data. It does not claim operational effectiveness, technology readiness, or validation for a particular aircraft or mission.