Skip to main content

China Telecom lets AI plan 5G networks, efficiency doubles

China Telecom lets AI plan 5G networks, efficiency doubles

Remember the days when planning a 5G network meant sending teams out for on-site inspections, manually collecting data, and relying on experts to pick each base station location? It was slow, labor-intensive, and often far from optimal. Well, China Telecom has just handed that whole process over to a large AI model—and the results are impressive.

On July 21, China Telecommunications Daily reported that the company's 5G wireless network planning large model has completed field trials. The model boosts planning efficiency by 50% and achieves an accuracy rate of over 75% for the planning schemes. It can automatically output design plans and accurately predict construction outcomes.

From human expertise to machine intelligence

Traditional wireless network planning has long depended on human experience and physical effort. In recent years, the industry has turned to big data analysis and ray tracing model simulations to speed things up. China Telecom has taken it a step further by integrating AI into the entire process—from network insight and site planning to twin simulation and scheme generation.

The company built a set of digital employee capabilities that cover full-scenario perception, analysis, decision-making, and execution. This focuses on three key areas: multi-modal data integration, atomic capability intelligent orchestration, and precise simulation of dynamic business changes.

Building a solid data foundation

The first step was creating a high-quality network planning dataset with multiple features. China Telecom integrated multi-source heterogeneous data—performance, configuration, geography, population, and service experience—using knowledge graphs, rule engines, NLP information extraction, and LLM prompt engineering. This systematically translated the experience of planning experts into structured data assets. Smart data parsing tools then ensured data quality, laying a solid foundation for intelligent planning.

A pioneering double-twin architecture

What really sets this apart is the double-twin large model collaborative planning architecture. The channel twin model handles high-precision reconstruction of the network space and forward-looking site planning. Meanwhile, the traffic twin model accurately predicts traffic fluctuations in complex scenarios. These two models work together to optimize and automatically generate the best plan that balances network coverage, capacity, value, and structure. This is the first time network planning has truly shifted from experience-based to model-driven.

Real-world testing in Shanghai

But theory is one thing; real-world results are another. China Telecom partnered with ZTE Communications to test the model in complex scenarios like residential areas and underground parking lots in Pudong, Shanghai. In residential areas, the model integrated multi-dimensional data such as population density, building distribution, and grid MR, achieving a station placement accuracy of over 80%. In underground parking lots, the team innovatively introduced the correlation between mobile speed and signal attenuation characteristics, achieving a coverage issue identification accuracy of over 75%.

A strategic shift for telecom operators

This isn't just a technical upgrade—it's a strategic shift. At the 2026 China Internet Conference, China Telecom's vice president Luan Xiaowei stated that the company will actively embrace and fully integrate intelligence, continuously improving the integrated system of computing power, platform, data, large models, and intelligent entities. The goal is to promote the transformation of enterprise operations from traditional traffic operations to Token operations. When base station site selection starts being decided by large models, the operating logic of telecom operators is being quietly rewritten.

Key Points

  • China Telecom's 5G wireless network planning large model improves planning efficiency by 50% and achieves over 75% accuracy in solutions.
  • The model uses a double-twin architecture (channel twin and traffic twin) for collaborative planning.
  • Field trials in Shanghai residential areas and underground parking lots showed station placement accuracy over 80% and coverage issue identification accuracy over 75%.
  • The initiative aligns with the Ministry of Industry and Information Technology's 'AI plus information and communication' push.
  • China Telecom aims to shift from traffic-based operations to Token-based operations, signaling a broader industry transformation.