ABU DHABI, 18th November, 2026 (WAM) -- Advancing research into the 6G networks, etisalat UAE and Khalifa University of Science and Technology jointly published a peer-reviewed blueprint for AI-native 6G networks, outlining how AI can be integrated into the foundations of future mobile networks, enabling them to anticipate demand, optimise performance, predict potential faults and operate with greater autonomy.
The study published in the Institute of Electrical and Electronics Engineers (IEEE) Open Journal of the Communications Society is built around a dedicated AI layer known as the “AI-plane” that sits alongside the layers that carry user traffic, control connections and manage operations. It gives operators, vendors and standards bodies a single reference for how 6G networks will learn, decide and act on their own.
Marwan Bin Shakar, Chief Technology Officer, etisalat UAE, and a co-author of the study, said: “Every generation of mobile technology has made networks faster. 6G is the first that can make them think. When intelligence is part of the architecture, the network can anticipate demand, predict faults and correct itself before customers notice, and our engineers move from operating the network to setting what it should achieve. With this study, our research with Khalifa University now covers how 6G will think, sense and be built, and we will keep taking that work to the forums where 6G is being defined.”
Prof. Bayan Sharif, Provost, Khalifa University, said: “The value of this collaboration lies in connecting rigorous academic research with the real-world challenges facing the telecommunications industry. By bringing together expertise in AI, computing and network engineering, our researchers are helping develop the knowledge and architectural foundations that can inform future 6G systems and international standardization discussions. This is how university–industry partnerships can translate research into long-term technological capability and strengthen the UAE’s contribution to emerging digital technologies.”
The study defines the end goal as autonomous connectivity where the networks can plan, deliver, assure and restore services by itself, with human operators setting intent, establishing policies and managing exceptions. Using the TM Forum's six-level autonomy scale, the research outlines a progression from today's largely assisted and partially automated networks towards higher levels of autonomy.The proposed architecture targets levels 3 to 4 in validated domains, with levels 4 to 5 for routine operations as the long-term objective.
The study also distinguishes AI-native from AI-assisted networks across six dimensions, including how AI is integrated into the architecture and role of human operators. It sets out a reference architecture with AI agents distributed across four tiers, from connected devices to the cloud, and full lifecycle management for AI models.
A key component is the network digital twin, a continuously synchronised replica of the live network, serving as a safe staging ground. This allows operators to train and test new AI models, including running them in shadow mode against the production network before they are allowed to act on live traffic.
The study maps AI-native design against the full standardization landscape: ITU IMT-2030, 3GPP Releases 18 to 21, the O-RAN Alliance, the AI-RAN Alliance and ETSI. It sets out a three-phase roadmap: research and prototyping (2026 to 2027), standardization and pre-commercial trials (2027 to 2029), and commercial deployment and evolution (2029 to 2032 and beyond).
The work was developed within the UAE’s TDRA-coordinated 6G framework. It reflects a national approach in which operators and universities contribute jointly to the requirements and architectures that global standards bodies are now writing into 6G.
The latest publication completes a three-part research series by etisalat UAE and Khalifa University in 2026. The first peer-reviewed study, published in July, explored integrated sensing and communication, examining how 6G networks could support applications such as intelligent transportation, industrial automation and smart cities.
The second, published in October in collaboration with NYU Abu Dhabi, presented an engineering framework for using centimetre-wave spectrum to increase network capacity while making efficient use of existing infrastructure. The latest study brings the intelligence dimension into focus, outlining how AI-native architecture could connect these capabilities