Skip to main content

China Mobile and Nokia deliver 70% cost saving with AI-powered 5G RAN

China Mobile 5G image.
(Image credit: China Mobile.)

By using AI to remove the human resource required to operate a radio access network (RAN), Nokia and China Mobile have demonstrated 90% efficiency and cost savings of more than 70%.

Nokia and China Mobile (CMCC) announced today that they have successfully completed live trials of an AI-powered radio access network (RAN) over CMCC’s network, using AI to forecast bandwidth and detect anomalies with 90 percent accuracy. 

“We believe it will be a key asset in improving the wireless network efficiency and the experience of its subscribers."

Pasi Toivanen, Nokia.

“We are excited to have worked with China Mobile on this project to advance RAN network intelligence,” said Pasi Toivanen, head of edge cloud platforms at Nokia. “We believe it will be a key asset in improving the wireless network efficiency and the experience of its subscribers. This is an example of Nokia’s commitment to supporting our customers in the delivery of world-class network performance.” 

Enabling AI and ML

Utilizing CMCC’s 4G and 5G networks, the companies completed an AI-based real-time user equipment (UE) traffic bandwidth forecast trial in Shanghai, whilst also detecting network anomalies in Taiyuan, the capital of China’s Shanxi province.

China Mobile used its ‘i-wireless-intelligent and simplicity 5G network’ concept during the trial, which is a series of technologies designed to create a greener and smarter 5G network. The near-real time RAN Intelligent Controller (RIC) is one of these technologies, enabling near real-time control and optimization of RAN elements using AI and ML applications.

“RIC plays a key role in enabling AI/ML capability in the RAN, which is of great significance to realize the concept of the ‘i-wireless-intelligent and simplicity 5G network’,” explained Huang Yuhong, deputy director of China Mobile Research Institute (CMRI). “Nokia and CMCC’s trials are very meaningful for RIC commericalization.”

In Shanghai, the trial confirmed that AI-based real-time user equipment (UE) traffic prediction accuracy exceeded 90 percent, which was achieved by estimating the UE radio quality and related throughput for 100 milliseconds. Nokia’s 5G AirScale basestation was then able to send UE radio quality information to the RIC in real-time.

“We are pleased to complete these trials using AI to forecast UE transmission bandwidth and detect anomalies on CMCC’s live network."

Huang Yuhong, CMRI.

“China Mobile has put effort into the AI-assisting RAN network technology,” Yuhong said. “We are pleased to complete these trials using AI to forecast UE transmission bandwidth and detect anomalies on CMCC’s live network with our partner Nokia. The field trial proved the availability of RIC enabling network enhancements through customized real-time BTS data analysis and control.”

The trial employed network automation using China Mobile’s 4G and 5G sites across more than 10,000 cell locations, and the companies found that AI/ML powered anomaly detection could deliver a cost saving of more than 70 percent on human resources.

Dan Oliver

Dan is a British journalist with 20 years of experience in the design and tech sectors, producing content for the likes of Microsoft, Adobe, Dell and The Sunday Times. In 2012 he helped launch the world's number one design blog, Creative Bloq. Dan is now editor-in-chief at 5Gradar, where he oversees news, insight and reviews, providing an invaluable resource for anyone looking to stay up-to-date with the key issues facing 5G.