Nokia and NTT Docomo have achieved two technological milestones on the path to 6G.

The first is the implementation of artificial intelligence (AI) and machine learning (ML) into the radio air interface, effectively giving 6G radios the ability to learn. The second is the use of new sub-terahertz (sub-THz) spectrum to boost network capacity.

The AI-native air interface and sub-THz spectrum are both critical research topics that Nokia and NTT Docomo are exploring for future 6G networks. These technologies could pave the way for immersive metaverse and extended reality (XR) experiences and a new generation of mobile applications.

The companies have implemented both technologies as proofs of concept at Nokia Bell Labs in Stuttgart, Germany, and Nokia and Docomo will demonstrate both technologies at Mobile World Congress in Barcelona later this month.

By pairing an AI-based learned waveform in a transmitter with a deep-learning receiver, Nokia Bell Labs, Docomo and NTT researchers were able to design and implement a learning air interface that transmits data efficiently under many different scenarios. This AI and ML-based implementation reduces signalling overhead, producing up to a 30% improvement in throughput.

In addition, the AI-native air interface will grant 6G networks the flexibility to adapt to the type of connection demanded by an application, device or user. For instance, a network in a factory can be optimised for industrial sensors at one moment and then reconfigured for robotic systems or video surveillance. In the public network, an AI-enhanced network can provide an optimised connection for a pedestrian in an XR session as well as an emergency vehicle traveling at high speed.

The sub-THz bands (100GHz and above) have never been designated for cellular use because of their propagation characteristics, but techniques such as beamforming could open up those frequencies to future 6G networks. These higher frequencies are well suited for high-accuracy radio sensing, which will likely be another key feature of 6G.

In their proof-of-concept, Nokia and NTT Docomo were able to demonstrate a 25Gbit/s connection on a single 256QAM stream over a carrier frequency of 144GHz using beamforming. Accessing the sub-THz bands would inject enormous capacity into 6G networks. The sub-THz bands won’t just improve overall capacity, they will allow 6G networks to support the most bandwidth intensive future use cases requiring multi-gigabit average connections.

Nokia and NTT Docomo first launched their 6G collaboration last year, with the aim of jointly defining and developing technologies for networking. The collaboration continues a long history of pioneering between Docomo and Nokia. Starting with 3G in the 1990s, to 4G, and to today’s 5G including collaboration in 5G O-RAN, the companies have turned ideas into implementations and pushed boundaries to create optimised experiences for end users.

“We are delighted to announce the steady progress of two key technologies towards 6G as achievements of the collaboration with NTT and Nokia started in June 2022,” said Takehiro Nakamura, chief technology architect at Docomo. “Through the collaboration, we continue to establish innovative 6G technologies and contribute to the global standardisation and commercialisation of 6G.”

Peter Vetter, president of Bell Labs Core Research at Nokia, added: “For the 6G era, we are using communication as a starting point. Networks will think, sense and act, and they will become the nexus point that bridges our digital and physical realities. Docomo and NTT share our 6G vision, and together we’re doing the fundamental research that will breathe life into that future.”

Nokia believes 6G will not just build on existing technologies and systems but expand and transform what a network can do. It will fuse the human, physical and digital worlds to liberate innate human potential. To achieve this, Nokia has envisioned six key technologies that will be vital components of future 6G networks. These are: spectrum technologies; AI native air interface; network as a sensor; extreme connectivity; cognitive, automated and specialised architectures; and security and trust.