How AI will be used to Boost 5G Capacity

News provided by

Jun 03, 2024, 08:38 ET

30% higher capacity in the radio network from AI optimization

CAMPBELL, Calif., June 3, 2024 /PRNewswire/ -- Mobile Experts released a groundbreaking techno-market report today, illustrating how AI/ML techniques will be used by mobile operators in their 5G networks to add capacity, without buying additional equipment.

Cost savings associated with AI enhancement of 5G radio networks

American mobile operators face a difficult future. Their capacity in dense urban markets is adequate today, and they even use some of this capacity for Fixed Wireless Access and other applications. But in 2027-2029, Mobile Experts is anticipating a shortfall of capacity in dense urban hotspots, due to ongoing growth in video traffic over the 5G network.

Artificial Intelligence is one reason for higher 5G data demand in the future. Today, AI plays only a very minor role in driving network traffic, but as photos and videos become increasingly modified by cloud-based AI engines, we expect to see ongoing data growth that will outstrip the capacity of the network.

"This new report examines how AI/ML will boost capacity of existing 5G infrastructure," commented Joe Madden, Principal Analyst at Mobile Experts. "We've been able to review field trials by all of the major network vendors, and we've put together a complete view of how this feature will be used—as well as how it will be sold into the market."

This new report, AI for the RAN 2024, includes a five-year outlook for software upgrades in the network, as well as the impact on baseband chipsets and GPU/SoC modules to support AI/ML in the network.

This report also looks closely at the difficulties for AI/ML use in Open RAN networks. Despite the good work of Open RAN Alliance and other organizations, Mobile Experts points out some aspects of the raw data used by AI and the training models that could give incumbent vendors an edge over Open RAN competitors.

Subscribers to Mobile Experts research will receive: