Case Study: China Unicom achieves faster, lower-cost network optimization with Nokia SWAT

China Unicom transforms operations with AI based automation

China Unicom, a major communications service provider, faced significant challenges in manually testing and optimizing its 4G and 5G networks. This process was extremely time-consuming, costly, and only covered 80% of the network, leading to slow problem resolution and reduced customer satisfaction. To address this, the company turned to vendor Nokia and its AI expertise to develop the Smart Wireless Analysis Tool (SWAT), based on the Nokia AVA AI and Analytics platform.

Nokia implemented SWAT, a cloud-native solution that uses over ten AI models to automate data collection and analysis from China Unicom's existing Hadoop platform. This allowed for 100% virtual network coverage and eliminated the need for extensive physical field testing. The results were substantial: Nokia's solution saved China Unicom €16 million annually, reduced annual working days spent on testing by 167,000, improved root cause analysis accuracy to 91%, and cut the time to resolve network problems by 67%.

China Unicom

LLiu XianSong

Director of Shanghai AI Center