Mycelial Federated Intrusion Detect System

Modern aircraft are evolving into highly intelligent, interconnected systems, leveraging Electronic Flight Bags (EFBs), Aircraft Interface Devices (AIDs), and real-time data exchange for enhanced operational efficiency. However, this digital transformation expands the attack surface, exposing aviation networks to sophisticated cyber threats that traditional security models—reliant on centralized monitoring and predefined attack signatures—fail to address. The Mycelial Federated Intrusion Detection System (MyFIDES), powered by the AeroShield Federated Learning System (AFLS), introduces a decentralized, AI-driven cybersecurity framework designed to detect and mitigate emerging threats in networked aircraft environments. MyFIDES leverages federated learning (FL) to train machine learning models directly on EFBs and AIDs without transmitting sensitive operational data to a centralized server. This approach preserves data privacy, reduces reliance on continuous cloud connectivity, and enables real-time anomaly detection even in low-bandwidth aviation environments. MyFIDES will integrate a lightweight intrusion detection agent with real-time streaming analytics, utilizing Apache Flink and MQTT-based communication to detect, classify, and respond to threats autonomously. The system will feature tamper-proof event logging, secure aircraft-to-ground (A2G) and aircraft-to-aircraft (A2A) communication, and a user-friendly cybersecurity dashboard for airline security teams. The Phase I project will establish MyFIDES’ feasibility through a simulated EFB/AID environment, hosted on DigitalOcean cloud resources and commodity hardware. The system will be evaluated against multiple attack scenarios, including GPS spoofing, malware injection, and unauthorized access attempts. In collaboration with Jackson State University and industry experts, MyFIDES will demonstrate a scalable, adaptive cybersecurity solution that enhances aviation resilience against next-generation cyber threats