Towards AI-Native 6G networks
The newly released deliverables “6GARROW Initial System Architecture”, “State of the art and challenges for AI/ML Enhanced Device Performance”, and “State of the art on AI/ML Solutions for RAN and Core Optimization”, together with “6GARROW Scenarios, Use Cases and related KPIs/KVIs,” lay the groundwork for the future of intelligent, adaptive, and sustainable mobile communication systems.
Building upon earlier deliverables that focused on project planning and methodology, these reports represent the first tangible results demonstrating how 6GARROW is advancing the state of the art in 6G network research.
Making 6G a success by introducing innovative functionalities and services
Deliverable 6GARROW Scenarios, Use Cases and related KPIs/KVIs introduces key use cases and related performance and value indicators meeting the requirements for European and Korean application of AI/ML in cellular networks. Building on the principle of an AI-native network design, the proposed use cases introduce novelty beyond the state of the art in fields of user experience and overall network efficiency. These are expected to be the main drivers for a future, commercially successful generation of mobile networks.
Establishing the Foundations of AI-Native 6G
Deliverable 6GARROW Initial System Architecture presents the initial 6GARROW system architecture, introducing a unified methodology for evolving and refining AI-native 6G architectures throughout the project. It builds on previous work on scenarios, use cases, and KPIs, performing a gap analysis to identify new functional requirements for semantic communication and joint radio access and core network optimization.
This first architectural blueprint sets a shared foundation for the consortium’s ongoing research and ensures that 6GARROW’s design approach remains technologically and economically relevant. Future iterations will refine this architecture as new findings emerge from work on device intelligence, network optimization, and system integration.
AI-Driven Device Intelligence: Towards Smarter and More Efficient Terminals
Deliverable State of the art and challenges for AI/ML Enhanced Device Performance establishes the baseline for enhancing device performance through AI and machine learning, supporting 6GARROW’s objective to embed intelligence directly into terminals and devices. The report reviews the current state of the art and identifies research opportunities across three key focus areas:
Terminal Traffic Automation and Optimization: leveraging AI for semantic communication and adaptive resource management.
Terminal Energy Efficiency Enhancement: developing energy-efficient AI architectures and distributed inference techniques.
Terminal Hardware Simplification: applying AI to reduce hardware complexity and improve systemlevel co-design.
These insights pave the way for the development of intelligent, energy-efficient, and adaptive devices that will be central to future AI-native networks.
Optimizing the Network Core and RAN with AI
Deliverable State of the art on AI/ML Solutions for RAN and Core Optimization focuses on the integration of AI/ML solutions into radio access (RAN) and core network domains, establishing a knowledge base for creating trustworthy, energy-aware, and resilient network infrastructures. It identifies key research priorities across three main themes:
AI-Native and Semantic 6G Architectures: exploring new designs that integrate semantic intelligence and comply with emerging AI regulations such as the EU AI Act.
AI-Driven Resource Usage and Energy Optimization: developing algorithms to improve energy efficiency and spectrum utilization.
Automation and Failure Recovery: advancing proactive resilience and cross-domain AI coordination for self-managing network systems.
These findings set the direction for 6GARROW’s future work on enabling networks that can selfoptimize, recover from failures autonomously, and deliver semantic, goal-oriented services.
A Strategic Step Toward Next-Generation Networks
Together, these deliverables represent a pivotal step in the 6GARROW project’s mission to redefine communication networks with AI at their core. They provide both a consolidated view of the current technological landscape and a strategic roadmap for future research and innovation.
“AI-native approaches represent a paradigm shift in the field of 6G wireless communication, offering unprecedented opportunities for innovation, optimization, and new business models,” says Nicolas Cassiau, the EU consortium deputy coordinator. “By embedding artificial intelligence at the heart of the network and seamlessly integrating devices into architectures, we can unlock new levels of automation, efficiency, flexibility, interoperability, security, and overall performance.”
6GARROW is a joint European–Korean research project supported by the Smart Networks and Services Joint Undertaking (SNS JU) and the Institute for Information & Communications Technology Promotion (IITP). Its mission is to develop AI-native, energy-aware, and resilient 6G network architectures that integrate intelligence across devices, radio access networks, and core systems.
Through close collaboration between leading academic, research, and industry partners, 6GARROW is advancing technologies that will shape the future of global communication infrastructure.
The 6GARROW consortium includes CEA-Leti, Fraunhofer HHI, Aalto University, University of Oulu, Intel, Hewlett Packard Enterprise, Grenoble Ecole de Management, Orange, Yonsei University, Korea University, ETRI and LG.