China Mobile partnered with Nokia to conduct two trials using AI-powered RANs to forecast bandwidth requirements and detect anomalies, with the operator planning to automate its networks to reduce operating costs.

In a trial in Shanghai, a RAN intelligent controller (RIC) was deployed in the edge cloud using Nokia’s AirFrame Open Edge server platform, which the vendor said enables “increased network optimisation capabilities through policy-guided, closed-loop automation”.

Real-time traffic prediction accuracy exceeded 90 per cent on the operator’s live 5G network. Nokia said its RIC acts as a gateway to AI and machine learning (ML) technologies, enabling a range of real-time, data-driven network enhancements.

The second trial in the city of Taiyuan, Shanxi province, examined operation automation on China Mobile’s 4G and 5G networks across more than 10,000 cells. Nokia stated the AI-powered system detected network problems more accurately and can significantly reduce “the time needed to analyse issues and find the root cause”, Nokia said.

Deputy director of China Mobile Research Institute Huang Yuhong stated RIC can play a key role in enabling AI and ML capability in the RAN.

Pasi Toivanen, Nokia’s head of Edge Cloud Platforms, said the company believes advance RAN intelligence “will be a key asset in improving wireless network efficiency” and user experience.