On February 25, at the Artificial Intelligence Summit, MWC Shanghai 2021, Hu Junjie, global chief marketing officer of ZTE transport network products, delivered a speech entitled Time Synchronization Network Powered by AI, which analyzed how to address new challenges in the time synchronization network in the 5G era through the application of Artificial Intelligence (AI) technology. He also shared views on how to unleash AI values, allowing the time network to achieve further use in high-quality 5G industry applications.

At present, new 5G services are booming such as public emergency services, children and seniors tracking, autonomous driving, the Internet of vehicle (IoV) and 5G industrial Internet for manufacturing require that new time synchronization requirements are put forward for high-precision positioning. However, the traditional satellite timing signals cannot provide effective services due to physical barriers and electromagnetic interference. So the ground time network is particularly important, and only cooperation between the ground and satellite networks assures full coverage. 

Faced with complicated deployment and difficult maintenance of the ground time network, Hu Junjie said: “The AI-based intelligent time network can upgrade the time synchronization network from the traditional manually deploying physical devices to the software-defining mode, and adopt time separation, error compensation, fault location and precision evaluation algorithms to automatically plan, deploy, operate and maintain the time synchronization network on the transport network.” 

He also added: “The above four AI algorithms can integrate the ultra-high-precision time synchronization network and positioning network into the communication network. We can take transport network as a synchronization service, and communication network as a positioning service.”

Currently, ZTE is in cooperation with a mainstream operator in China to carry out a series of verification and live tests on its 5G mobile backhaul network, with workflow engines and rule engines introduced to enable automatic deployment and fault location of the time synchronization network. The test results prove that the intelligent time network solution can automatically calculate and plan the primary and secondary clock synchronization of network element (NE) topologies, and automatically adjust the planning and configuration according to the changes. Now a single person can deploy 500 NEs a day, increasing the efficiency by about 7 times compared with 75 NEs originally. Additionally, the knowledge graph can be used for path planning to shorten fault location and resolution time from more than 24 hours to less than 15 minutes, which improves diagnosis efficiency nearly 100 times.

The AI empowered time network will show more value in the future. ZTE will continue working with industry partners to explore and deepen the integration of AI and 5G, serve thousands of industries, and move high-quality development forward.