Periodic Reporting for period 1 - SAFE-6G (A Smart and Adaptive Framework for Enhancing Trust in 6G Networks)
Okres sprawozdawczy: 2024-01-01 do 2025-06-30
Problem Statement and Need for SAFE-6G
Current operator-centric 4G/5G architectures centralise key functions, creating bottlenecks as billions of devices and diverse services — from XR to mission-critical IoT — join the network. Europe’s 6G vision calls for human-centric, open, and sovereign systems with trustworthiness — safety, security, privacy, resilience, and reliability — built in from the start, without compromising agility or innovation.
As network functions move to cloud-native, virtualised, and disaggregated deployments across the edge–cloud continuum, the attack surface grows. Open interfaces, third-party apps, and federated deployments bring innovation but also risks, including untested components, data exposure, and complex governance. Existing 5G security frameworks, designed for centralised architectures, are insufficient. SAFE‑6G addresses this by treating security as holistic trustworthiness, balancing protection with performance, using cognitive automation, and giving users control over trust policies and data ownership.
SAFE-6G re-architects the 6G core into User Service Nodes (USN) and Network Service Nodes (NSN), enabling per-user instantiation of services, policies, and trust functions as modular, cloud-native microservices across the edge-to-cloud continuum. A cognitive coordinator, powered by explainable AI/ML, orchestrates intent-driven configuration, third-party innovation, and seamless trusted service mobility, making trust a measurable Key Value Indicator (KVI).
Five trust functions — Safety, Security, Privacy, Resilience, and Reliability — are dynamically balanced by the coordinator, supported by federated AI and an xAI-enhanced MLOps framework. Users interact through a chatbot that translates natural-language intents into trust configurations with transparent explanations of AI decisions.
Pathway to Impact
- Define 6G trustworthiness requirements and deliver a reference architecture, validated through immersive/metaverse use‑cases.
- Design cognitive coordination for trust, integrating federated and explainable AI, ethical AI certification, and differential privacy.
- Implement the five trust functions with zero‑trust principles, blockchain‑enabled credentials, advanced anomaly detection, and zero‑touch lifecycle automation.
- Adopt a cloud‑native, MetaOS‑compatible architecture for interoperability, scalability, and sustainable integration into future deployments.
- Validate via metaverse pilots, applying rigorous KPI/KVI frameworks under diverse threat models.
- Maximise adoption and impact through active dissemination, standardisation input, capacity building, and engagement with industry, regulators, and open‑source communities.
Expected Scale and Significance
By delivering a blueprint and validated prototype for natively trustworthy, user‑centric 6G, SAFE‑6G is poised to:
- Reduce vulnerabilities in multi‑stakeholder edge–cloud deployments;
- Empower users and verticals to control trust policies and data governance;
- Lower barriers to innovation via open, secure API ecosystems;
- Provide measurable trustworthiness KPIs/KVIs that can inform EU policy and standardisation;
- Contribute to Europe’s leadership in secure, human‑centric network technology.
Current operator-centric 4G/5G architectures centralise key functions, creating bottlenecks as billions of devices and diverse services — from XR to mission-critical IoT — join the network. Europe’s 6G vision calls for human-centric, open, and sovereign systems with trustworthiness — safety, security, privacy, resilience, and reliability — built in from the start, without compromising agility or innovation.
As network functions move to cloud-native, virtualised, and disaggregated deployments across the edge–cloud continuum, the attack surface grows. Open interfaces, third-party apps, and federated deployments bring innovation but also risks, including untested components, data exposure, and complex governance. Existing 5G security frameworks, designed for centralised architectures, are insufficient. SAFE‑6G addresses this by treating security as holistic trustworthiness, balancing protection with performance, using cognitive automation, and giving users control over trust policies and data ownership.
SAFE-6G re-architects the 6G core into User Service Nodes (USN) and Network Service Nodes (NSN), enabling per-user instantiation of services, policies, and trust functions as modular, cloud-native microservices across the edge-to-cloud continuum. A cognitive coordinator, powered by explainable AI/ML, orchestrates intent-driven configuration, third-party innovation, and seamless trusted service mobility, making trust a measurable Key Value Indicator (KVI).
Five trust functions — Safety, Security, Privacy, Resilience, and Reliability — are dynamically balanced by the coordinator, supported by federated AI and an xAI-enhanced MLOps framework. Users interact through a chatbot that translates natural-language intents into trust configurations with transparent explanations of AI decisions.
Pathway to Impact
- Define 6G trustworthiness requirements and deliver a reference architecture, validated through immersive/metaverse use‑cases.
- Design cognitive coordination for trust, integrating federated and explainable AI, ethical AI certification, and differential privacy.
- Implement the five trust functions with zero‑trust principles, blockchain‑enabled credentials, advanced anomaly detection, and zero‑touch lifecycle automation.
- Adopt a cloud‑native, MetaOS‑compatible architecture for interoperability, scalability, and sustainable integration into future deployments.
- Validate via metaverse pilots, applying rigorous KPI/KVI frameworks under diverse threat models.
- Maximise adoption and impact through active dissemination, standardisation input, capacity building, and engagement with industry, regulators, and open‑source communities.
Expected Scale and Significance
By delivering a blueprint and validated prototype for natively trustworthy, user‑centric 6G, SAFE‑6G is poised to:
- Reduce vulnerabilities in multi‑stakeholder edge–cloud deployments;
- Empower users and verticals to control trust policies and data governance;
- Lower barriers to innovation via open, secure API ecosystems;
- Provide measurable trustworthiness KPIs/KVIs that can inform EU policy and standardisation;
- Contribute to Europe’s leadership in secure, human‑centric network technology.
The first period of SAFE-6G marked the shift from conceptual design to initial technical implementation:
1)The consortium established the functional and architectural foundations of a user-centric 6G ecosystem with trustworthiness as a native feature.
2) A reference architecture was defined to support dynamic orchestration across the edge-cloud continuum, integrating domains such as user intent processing, AI-driven coordination, and trust assessment.
3) Two use cases illustrate its capabilities: an XR-enabled educational scenario and a factory digital twin.
4) Novel Key Value Indicators (KVIs) were introduced across five trust dimensions: resilience, privacy, safety, security, and dependability.
5) A GitLab repository was launched to share open-source components.
6) Key deployments included a meta-operating system across three sites, Cumucore Core at two sites, and the OpenCAPIF framework for safe API access. An initial MLOps framework was also provided to support AI-driven elements.
7) Progress was made on intelligent user-intent identification, with a chatbot prototype and a Cognitive Coordinator using BERT-based models for real-time orchestration of trust functions. All five trust dimensions achieved their first implementations.
8) A technical demonstration at EuCNC 2025 showcased these core components, marking a major milestone in the project.
1)The consortium established the functional and architectural foundations of a user-centric 6G ecosystem with trustworthiness as a native feature.
2) A reference architecture was defined to support dynamic orchestration across the edge-cloud continuum, integrating domains such as user intent processing, AI-driven coordination, and trust assessment.
3) Two use cases illustrate its capabilities: an XR-enabled educational scenario and a factory digital twin.
4) Novel Key Value Indicators (KVIs) were introduced across five trust dimensions: resilience, privacy, safety, security, and dependability.
5) A GitLab repository was launched to share open-source components.
6) Key deployments included a meta-operating system across three sites, Cumucore Core at two sites, and the OpenCAPIF framework for safe API access. An initial MLOps framework was also provided to support AI-driven elements.
7) Progress was made on intelligent user-intent identification, with a chatbot prototype and a Cognitive Coordinator using BERT-based models for real-time orchestration of trust functions. All five trust dimensions achieved their first implementations.
8) A technical demonstration at EuCNC 2025 showcased these core components, marking a major milestone in the project.
During the first reporting period, SAFE‑6G has delivered substantial advancements towards its vision of a natively trustworthy, user‑centric 6G architecture. A large language model–driven chatbot interface has been implemented to capture user intent in natural language and translate it into quantifiable trust requirements, ensuring that network behaviour and service delivery are directly aligned with user‑defined policies.
The project has operationalised a holistic trustworthiness framework that extends the traditional security paradigm to also integrate safety, privacy, resilience, and reliability as first‑class design objectives. Technically, key progress includes:
- Development of privacy‑preserving mechanisms embedded within the architecture to safeguard sensitive data in distributed edge–cloud environments.
- Definition of a comprehensive set of trustworthiness KPIs and KVIs, tailored to measure and optimise each of the five trust functions at system, service, and societal levels.
- Initial deployment of the Cognitive Coordinator, an AI/ML‑driven orchestration component capable of dynamically calibrating trust levels per user/system instance across the continuum.
- Integration of explainable AI (xAI) techniques to provide transparency in automated trust‑related decisions, enabling users to understand the rationale behind configuration and policy enforcement.
- Implementation of modular Trust Functions — deployable as microservices — that support adaptive trust control and enable real‑time security actions such as anomaly detection and threat mitigation.
Early metaverse‑based demonstrations have validated the SAFE‑6G architectural approach, showcasing its ability to adapt trust configurations on‑the‑fly under varying service, threat, and performance conditions. These initial results confirm the feasibility of SAFE‑6G’s concepts and highlight their potential for high‑impact applications.
In the next period, technical efforts will focus on broadening validation scenarios, deepening integration across trust functions, and aligning implementations with evolving standards to ensure interoperability, scalability, and sustainability beyond the project’s lifetime.
The project has operationalised a holistic trustworthiness framework that extends the traditional security paradigm to also integrate safety, privacy, resilience, and reliability as first‑class design objectives. Technically, key progress includes:
- Development of privacy‑preserving mechanisms embedded within the architecture to safeguard sensitive data in distributed edge–cloud environments.
- Definition of a comprehensive set of trustworthiness KPIs and KVIs, tailored to measure and optimise each of the five trust functions at system, service, and societal levels.
- Initial deployment of the Cognitive Coordinator, an AI/ML‑driven orchestration component capable of dynamically calibrating trust levels per user/system instance across the continuum.
- Integration of explainable AI (xAI) techniques to provide transparency in automated trust‑related decisions, enabling users to understand the rationale behind configuration and policy enforcement.
- Implementation of modular Trust Functions — deployable as microservices — that support adaptive trust control and enable real‑time security actions such as anomaly detection and threat mitigation.
Early metaverse‑based demonstrations have validated the SAFE‑6G architectural approach, showcasing its ability to adapt trust configurations on‑the‑fly under varying service, threat, and performance conditions. These initial results confirm the feasibility of SAFE‑6G’s concepts and highlight their potential for high‑impact applications.
In the next period, technical efforts will focus on broadening validation scenarios, deepening integration across trust functions, and aligning implementations with evolving standards to ensure interoperability, scalability, and sustainability beyond the project’s lifetime.