Cognitive Cyber-Resilient C2 Link Management for Advanced Air Mobility Platforms

Abstract

Advanced Air Mobility (AAM) platforms rely on heterogeneous command and control (C2) links-LTE and SAT-COM-yet current frameworks prioritize connectivity over integrity, leaving them critically vulnerable to coordinated jamming, spoofing, and replay attacks. As AAM operations scale toward dense urban airspace, the absence of trust-aware link management introduces systemic risks that SNR-based switching policies are fundamentally ill-equipped to address. We propose a cognitive cyber-resilient C2 link management framework that elevates link trust above raw uptime by integrating multi-layered threat intelligence with adaptive decision-making. A hybrid BiLSTM+TCN anomaly detector achieves 89.69% accuracy with perfect recall $(\text{AUC} =1.000)$ across jamming and replay attack classes, enabling early and reliable threat identification under realistic channel conditions. Guided by dynamic threat probability estimates, RF fingerprint authentication, and a Watchdog safety layer, a Dueling Deep Q-Network (Dueling-DQN) agent selects defensive link-switching actions that preserve mission continuity without sacrificing security posture. The framework is implemented and validated in OMNeT++ 6.3 under five independent random seeds, demonstrating complete elimination of C2 hijack events across all single-vector attack scenarios. In contrast, the SNR-threshold baseline averaged two hijack events per scenario, confirming the inadequacy of reactive, signal-qualityonly approaches. These results establish a strong foundation for trust-centric, AI-driven C2 link resilience in next-generation AAM deployments.