PATHSHIELD

PATHSHIELD is an AI-native, privacy-preserving Intrusion Detection and Mitigation System (IDMS) developed by PinakashieldTech OÜ (PST) specifically engineered for 5G/6G hospital edge networks. As healthcare adopts mission-critical services like remote surgery and ICU telemetry, the cyberattack surface expands, requiring security that is both autonomous and network-native.

Core Technology Pillars

The framework integrates three innovative pillars to ensure a robust defense:

• Federated Transfer Learning (FTL): Enables collaborative AI model training across distributed hospital nodes without transferring raw Protected Health Information (PHI). This ensures strict compliance with GDPR and data sovereignty.
• Generative AI (GenAI): Proactively synthesizes realistic zero-day attack data to harden the system against novel threats that traditional signature-based methods miss.
• Blockchain-Based Reputation: Validates the integrity of model updates from participating nodes, defending the system against model poisoning attacks.

Implementation and Impact

The project will be validated on the Fraunhofer FOKUS testbed (TRL 7), leveraging its Open5GCore and Network Exposure Function (NEF). PST’s solution PATHSHIELD moves beyond passive monitoring to active, closed-loop enforcement. Upon detecting threats (e.g., DDoS or MitM), the system programmatically triggers the NEF APIs to re-route malicious traffic to quarantined network slices or downgrade Quality of Service (QoS) for compromised sessions all in less than 200 milliseconds.

Ultimately, PATHSHIELD provides a “security-by-design” blueprint for critical B5G infrastructure, ensuring that digital health services remain resilient, energy-efficient, and strictly private.