Main Technical Results Aligned with Project Objectives

1. Policy‑Driven Security, Privacy, and Trust Framework (Objective 1)
   A fully operational, context-aware, policy-based security architecture was implemented for multi-tenant B5G/6G infrastructures. It integrates dynamic adaptation to regulatory requirements, encryption key management, and anomaly detection to guarantee data integrity and computation security, even in the presence of malicious tenants.

2. Optimal Data Fragmentation and Privacy-Preserving Representations (Objective 1)
   Advanced data fragmentation, storage distribution, anonymization, and obfuscation techniques were deployed. These mechanisms ensure sensitive data remains secure while preserving utility for AI/ML training. Representation learning algorithms were applied to identify malicious network flows and enhance threat detection.

3. Federated Learning Combined with Blockchain for System-Level Security (Objective 1)
   A federated learning architecture integrated with blockchain was implemented to provide security in server-less 6G networks. Homomorphic encryption, secure key aggregation, and zero-knowledge proofs were used to protect local models and guarantee correct participation of agents. Reinforcement learning and transfer learning techniques enabled detection of malicious nodes in V2X automotive scenarios.

4. Decentralized Blockchain Smart Contract Platform for Network Slicing (Objective 2)
   A Hyperledger Fabric-based smart contract platform was deployed to manage decentralized transactions for network slicing in multi-tenant environments. Private and public channels were created for secure slice negotiation and management without relying on a central trusted entity.

5. Integrated NextGen OSS with Secure Orchestration APIs (Objective 2)
   The NextGen OSS platform was deployed to orchestrate multi-tenant slices, integrate blockchain smart contracts, and enforce security policies. It included real-time processing modules, AI/ML capabilities for threat detection, automated workflows, and APIs for interoperability with 5G core and RAN simulators (OAI and UERANSIM).

6. AI-Driven Threat Detection and Fusion Algorithms (Objective 1)
   Machine learning models, including ARIMA, Prophet, Isolation Forest, K-means clustering, and representation-based fusion algorithms, were applied for anomaly and threat detection. These models processed logs from gNodeBs and other network components, enabling real-time detection of signaling storms, DoS attempts, and malicious behaviors.

7. Use Case Demonstrations Validating Core Solutions (Objectives 1 & 2)
   Three fully validated use cases demonstrated the integration of the project’s technologies:

   • UC1: Secure user registration and query with data recorded simultaneously on blockchain and NextGen OSS, combined with anomaly detection.

   • UC2: Automated network slice creation and inventory management, including verification of resource availability and storage of slice configurations on blockchain and OSS.

   • UC3: Real-time RAN anomaly detection using predictive models and clustering to identify unauthorized access, signal storms, and misconfigured connections.
     The integrated deployment utilized AWS Kubernetes infrastructure, CI/CD pipelines, and containerized modules for scalable and reproducible validation.

8. Dissemination, Exploitation, and Industrial Outreach (Objective 3)
   Project results were widely disseminated through international publications, workshops, and high-visibility demonstrations at MWC 2025 and Smart City Expo 2023. Industrial partners were actively engaged, creating opportunities for knowledge transfer, early adoption of technologies, and future collaborations in secure 6G network deployment.