Beyond 5G: The dawn of intelligent networks 

Smart homes, self-driving cars and immersive games are just a few examples of applications powered by 5G technology, which provides greater data rates and increased bandwidths than previous generations of wireless technology.  

As the internet of things (IoT) continues to evolve, 5G’s capabilities could soon fall short of supporting the scale, complexity, and innovation of future applications. Additionally, the growing need for sustainable and energy-efficient networks due to climate change is set to drive innovation in eco-friendly practices. These challenges have prompted the information and communication technology community to begin planning for the next generation.  

Enhancing connectivity and efficiency with AI 

Led by its founding director Mérouane Debbah, Khalifa University’s 6G Research Center (6GRC) is preparing to meet these challenges by devising next-generation protocols, infrastructure and AI models that can handle these transformations.  

“Unlike 5G, which is primarily about speed and connectivity, 6G will embody an ecosystem of intelligent networks capable of self-optimization and real-time responsiveness,” says Debbah, who received the IEEE Communications Society Industrial Innovation Award for the Europe, Middle East, and Africa Region in 2024.  

Debbah’s team focuses on the intersections of advanced wireless technologies and AI. “Leading the 6G initiative was a natural progression,” says Debbah, who joined Khalifa University in 2023 as a professor and established the 6GRC to advance these fields.  

In line with the United Arab Emirates’ knowledge-based economy, 6GRC supports technological innovation and sets ambitious goals for the UAE’s telecoms future. Alongside Abu Dhabi’s Technology Innovation Institute (TII), 6GRC also organizes the annual 6G Abu Dhabi Summit, a key conference that brings together the main stakeholders in the field.  

Collective expertise

“Collaboration is central to our research philosophy. We actively engage with industry leaders, governmental bodies, and academic peers to ensure our research addresses real-world challenges and leverages collective expertise,” Debbah says, adding that 6GRC works closely with the UAE’s governmental agency Telecommunications and Digital Government Regulatory Authority (TDRA), global telecom operators and AI-centered research institutions to align its priorities with emerging industry needs and regulatory standards.

“Unlike 5G, which is primarily about speed and connectivity, 6G will embody an ecosystem of intelligent networks capable of self-optimization and real-time responsiveness.”

Mérouane Debbah 

These collaborations provide valuable insights that guide the center’s research and accelerate the transition from concept to real-world application, allowing the team to deliver impactful, practical solutions. “Our partnership with TDRA on the UAE’s strategic vision for 6G reflects this approach and aligns our work with national and international telecom objectives,” Debbah says.  

One key project is TelecomGPT, a large language model (LLM) tailored for the telecom sector.¹ Trained on telecom data including standards, patents and technical documents, TelecomGPT performs tasks ranging from mathematical modeling to generating protocol workflows. “This has led to breakthroughs in intelligent network management and real-time communication insights,” says Debbah. Building on TelecomGPT, 6GRC has worked with international collaborators to launch Hermes², an LLM framework that uses structured guides or ‘blueprints’ to address common telecom modeling challenges, ensuring reliable and efficient operation and bringing fully autonomous mobile networks closer to reality. 

Advancing wireless technology 

In 2023, 6GRC established the 6G KU–TII Chair program to promote advancements in wireless technology and related communication activities through research and innovation. “LLMs are likely to considerably impact 6G. They can significantly improve user-mobile phone interactions, engineer-network interactions, and algorithms, such as radio resource allocation algorithms,” says the current 6G Chair on Native AI Samson Lasaulce. Users will be able to make various requests over the phone and rely on AI to complete the corresponding tasks in a flexible, automated, and efficient manner, he explains. This sharply contrasts with 5G networks and devices, which are quite rigid and involve tedious interfaces, he adds. 

An example of this is an LLM-based system developed by Lasaulce and coworkers that allows users to smartly recharge their electric vehicle using user-driven resource scheduling³. “A mobile user can say ‘Charge my vehicle by 5 a.m. while managing its battery lifetime.’ The model will automatically simulate the corresponding physical problem, solve it, and produce the appropriate control for the vehicle,” Lasaulce says. 

The researchers are also exploring the ability of an LLM to solve an optimization problem, which involves finding the best solution from a set of possible choices. “Most people are non-experts and do not know how to formulate and solve an optimization problem. We want to democratize optimization using LLMs,” Lasaulce says.  

Deputy director Sami Muhaidat spearheads research at 6GRC, investigating projects ranging from low-power IoT networks to machine learning for future wireless communications. Researchers are exploring the integration of active machine learning in 6G networks to address challenges, such as high data demands, low latency and efficient network management. Unlike conventional systems, active models can selectively ask questions about the training data, reducing training data needs and improving learning efficiency. Based on this concept, Muhaidat and coworkers have recently designed a framework that simultaneously labels and acquires data⁴. By leveraging AI, digital twins and active learning, the framework is set to create intelligent, adaptive and resource-efficient 6G networks.  

Meanwhile, leadership team member Arafat Al-Dweik oversees the Localization and Sensing research program, which is dedicated to enhancing localization and environmental mapping —capabilities crucial for deploying autonomous systems. In addition, they are developing innovative AI-driven algorithms to address various security challenges, such as jamming detection and secure key sharing. In a joint research project with TII, they have recently created a high-update-rate key sharing scheme to facilitate secured communications, multicast, and broadcast transmission in unmanned aerial vehicle swarm networks ⁵. An essential element of the scheme is its frequent updating of secret keys, enhancing security by making it more challenging for attackers to compromise the system.  

“We envision a future where 6G networks provide connectivity and serve as intelligent frameworks capable of autonomous operation and environmental adaptability. In the next decade the 6G landscape will likely present widespread adoption of AI-powered telecom infrastructure, extensive use of edge computing, and a significant shift toward data-driven resource management, all of which 6GRC is actively exploring,” Debbah says. 

References

1. Maatouk, A., Piovesan, N., Ayed, F., De Domenico, A. & Debbah, M. Large language models for telecom: forthcoming impact on the industry. IEEE Communications Magazine, 1–7 (2024) | Article
2. Ayed, F., Maatouk, A., Piovesan, N., De Domenico, A., Debbah, M. & Luo, Z.-Q. Hermes: a large language model framework on the journey to autonomous networks. arXiv preprint (2024) | Article
3. Mongaillard, T., Lasaulce, S., Hicheur, O., Zhang, C., Bariah, L., Varma, V. S., Zou, H., Zhao, Q. & Debbah, M. Large language models for power scheduling: a user-centric approach. arXiv preprint (19 July 2024) | Article
4. Alhussein, O. & Zhuang, W. Active ML for 6G: towards efficient data generation, acquisition, and annotation. arXiv preprint (5 June 2024)| Article
5. Jangsher, S., Al-Dweik, A., Iraqi, Y., Pandey, A. & Giacalone, J.-P. Group secret key generation using physical layer security for UAV swarm communications. IEEE Transactions on Aerospace and Electronic Systems59, 8550–8564 (2023) | Article