YOU ARE AT:6GHow can AI improve the 5G and 6G air interface?

How can AI improve the 5G and 6G air interface?

AI is being explored in 5G-Advanced and expected to play a more important role in the 6G air interface

The building blocks for the use of artificial intelligence in the air interface are being developed and put together in 5G-Advanced, and AI is expected to play an even more important role in the 6G air interface.

Dr. Paul Harris, principal wireless architect in the CTO office at Viavi Solutions, gave a run-down of current research and development areas of focus for the 5G and 6G air interface in a session during 6G Forum (sessions available on-demand here).

“We’ve seen in the 5G timeframe already, some developments in how AI could be applied to communications systems and in particular, to the radio interface itself,” Harris said.

Researchers are considering the various components of both the transmit and receive portions of typical communications systems, selecting specific components for potential integration of AI and also considering the steps toward a fully AI-native air interface.

Viavi Solutions has an advanced research lab, its Marconi Lab in the U.K., where it explores and puts together proofs-of-concept to test out the use of AI in different blocks of the air interface and the results, as well as develop a full understanding of how to test and optimize these scenarios.

Harris said that use cases are already being worked on for AI to be used in the 5G-Advanced air interface. He cited several examples of work related to improving beam management in beamforming for massive MIMO systems, more precise positioning and better energy management, as well as better assessment of radio channel conditions.

Specifically, the use of AI in beam management has been evolving in 3GPP Release 18, in the areas of beam selection and prediction, Harris said—such as using AI to make better decisions and also, work with fewer beams. Viavi, he added, has developed its own centimeter-wave AI beam management proof-of-concept to evaluate how to train such a solution and also, how to test and optimize the associated requirements.

How can AI improve the 5G and 6G air interface?

Integrating AI into the 5G or future 6G air interface will require some fundamental changes, he explained, including the implementation of what he called a “neural network receiver” that would enable certain air interface blocks to be optimized by AI/ML processing.

Another area which is a focus for the application of AI to the 5G and 6G air interface is in the area of energy consumption, Harris said. Viavi has been researching this area as well. “Energy consumption has been a key aspect that has come up throughout 5G but also has been a high priority for 6G under many discussions now, and AI is one of the area where this is a concern,” Harris said, because of the large amounts of high-intensity data processing that have to be done to support AI, particularly during the AI model training phase.

However, as Viavi has been conducting research in this area and looking to he said that it sees cause for optimism. It has begun to explore what Harris described as a “neuromorphic receiver” approach that is based on a closer mimicking of the human brain and transmissions between neurons—particularly the fact that the human neural system doesn’t repeat every transmission fully every single cycle, which is typical in digital systems. Some activity is triggered only during a specific time window, so that many neurons can stay dormant until needed.

Such an approach to AI implementation has the potential for “dramatic” energy savings, Harris said. Viavi has moved toward the neuromorphic receiver approach and trained across a number of 3GPP channels (and specifically, replacing some channel estimation functions), Harris said.

“What’s quite incredible is that we found that for the scenarios that we’ve been evaluating, we can maintain similar performance … but a huge reduction in energy consumption by nine times,” Harris explained. “So this is quite exciting, because it shows this potential to leverage a lot of the benefits of what AI can bring for the air interface, but also not have to have huge amounts of energy costs in that process.”

All the sessions from 6G Forum, including discussions on the vision for 6G, security, use cases and more, are available on-demand here.

ABOUT AUTHOR

Kelly Hill
Kelly Hill
Kelly reports on network test and measurement, as well as the use of big data and analytics. She first covered the wireless industry for RCR Wireless News in 2005, focusing on carriers and mobile virtual network operators, then took a few years’ hiatus and returned to RCR Wireless News to write about heterogeneous networks and network infrastructure. Kelly is an Ohio native with a masters degree in journalism from the University of California, Berkeley, where she focused on science writing and multimedia. She has written for the San Francisco Chronicle, The Oregonian and The Canton Repository. Follow her on Twitter: @khillrcr