In the wave of global digital transformation, the optimization and intelligence of 5G networks have become an important area of technological competition. On January 14, 2025, China Telecom Research Institute, Guangdong Telecom, Jiangsu Telecom, and Huawei completed the industry's first live network trial of a new wireless AI dispatching technology. This test demonstrates China's strong capabilities in the field of 5G technology innovation and provides a reference for the intelligent upgrade of global 5G networks.
Technical background and core innovation
The highlight of this experiment is the application of AI time-frequency and space-time 3D scheduling technology to realize the intelligent management of wireless network resources. The test covered the T+T/T+F dual scenario, involving about 50 sites, and comprehensively tested the performance of AI technology in beam, spectral efficiency, and coverage optimization. Here's an overview of the technical details and key figures:
1. Intelligent Beam Prediction Technology:
In the experiment, a lightweight AI model was used to accurately predict the beam direction and user distribution, and realize the dynamic scheduling of resources.
- Performance improvement: In medium and high load scenarios, the average throughput of users is increased by more than 15%.
- Model stability: The performance gain fluctuates by less than 5%, ensuring continuous optimization of the user experience.
2. Collaborative scheduling of resources:
Through the joint optimization of time-frequency-space multi-dimensional resources, the performance of the live network is improved without the need for additional hardware costs.
- CPU and energy consumption: The CPU utilization rate and energy consumption of the base station increased by less than 1% during the AI model training stage, which remained unchanged after the AI model was launched, fully verifying its green and energy-saving characteristics.
3. Test Environment Data:
- Scenario coverage: 50 sites, covering high-density urban areas and suburban networks.
- Experimental load: Simulates user behavior under different peak and off-peak network loads.
This test not only demonstrates the adaptability of AI technology in 5G networks, but also opens up a new path for intelligent scheduling, providing practical cases for operators around the world.
Figure: China Telecom and Huawei jointly test a new wireless AI dispatching technology on the live network
Industry impact: intelligent and low-carbon win-win
1. Direct impact on operators
Through the introduction of wireless AI technology, operators can effectively reduce the cost of network optimization and significantly improve the user experience. Taking this experiment as an example, user throughput and network coverage have been significantly improved without additional hardware investment. This means that the intelligent upgrade of 5G networks can take into account both performance and economic benefits.
2. Low-carbon development contributes to sustainability
The experimental results show that the energy consumption performance of the AI model is extremely good. The change in CPU load and overall energy consumption is less than 1%, which not only reduces the operating cost of 5G networks, but also conforms to the current global development trend of carbon neutrality and green communications. This achievement provides a practical technical path for the green development of 5G.
3. Global market potential
Huawei's cooperation with China Telecom provides a strong endorsement for the promotion of 5G-A (5G Advanced) technology. With the acceleration of the commercialization of 5G-A technology, the application scenarios of intelligent dispatching technology will be expanded to unmanned driving, industrial Internet of Things, low-altitude economy and other fields in the future, providing a new growth point for the global communications industry.
Application Scenario Prospect: Towards a Smart Society
1. Smart city management
The deep integration of 5G and AI has laid the foundation for the construction of smart cities. For example:
- Transportation: Improve urban traffic efficiency through real-time traffic monitoring and intelligent scheduling technology.
- Energy management: Dynamically adjust the distribution of electricity to achieve precise regulation and control of the smart grid.
- Medical scenarios: It supports high-speed transmission of remote surgery and medical images to improve the utilization of medical resources.
2. Industrial intelligence
In the industrial Internet scenario, wireless AI scheduling technology will help the flexible deployment of production lines and the construction of smart factories to achieve a more efficient and low-cost production mode.
3. Agriculture and Ecological Management
The low energy consumption of wireless AI technology makes it have great potential for large-scale agricultural monitoring. For example, through the precise scheduling of sensor networks, real-time monitoring of soil and climate data provides technical support for smart agriculture.
Challenges and solutions: from technological innovation to large-scale application
Although there are many breakthroughs in this trial, there are still challenges in the actual scale-up:
1. Adaptability and complexity of AI models
Current AI models need to be further optimized in multi-scenario adaptation to meet the real-time requirements in highly dynamic environments. This requires strengthening industry-university-research cooperation and promoting the standardization of model algorithms.
2. Cyber security risks
With the deepening of 5G network intelligence, AI technology may also bring new cybersecurity problems, such as data leakage and model attacks. This requires the deployment of technology at the same time as the strengthening of network security protection mechanisms.
3. Balance between costs and benefits
Although AI technology has advantages in low-cost deployment, the initial investment is still a burden that cannot be ignored for small and medium-sized operators around the world. Therefore, it is necessary to lower the technical threshold and promote large-scale application through international cooperation and industry alliances.
Personal observations: from technological innovation to social value
The cooperation between China Telecom and Huawei is a testament to the global leadership of Chinese enterprises in the field of 5G technology. However, the development of technology ultimately serves the needs of society. Here are a few personal observations:
1. Technology empowers social development
The breakthrough of 5G AI scheduling technology is not only a victory for the communications industry, but also a milestone in social intelligence. It will promote progress in many fields such as smart city construction, transportation optimization, and medical inclusion.
2. Green communications lead the future
At a time when the global communications industry is facing huge carbon emission pressures, China Telecom and Huawei's trial shows how the technology can balance performance and energy saving goals, setting an example for green communications.
3. The need for global cooperation
In the future, the development of 5G technology will require wider international cooperation to jointly formulate standards and promote the global communications industry to move towards a smarter and more low-carbon direction.
Conclusion
China Telecom and Huawei jointly completed the wireless AI dispatching technology test, opening a new chapter in the development of 5G intelligence. In terms of improving network performance and energy saving performance, this technology has high application value and promotion potential. In the future, with the continuous evolution of 5G-A and 6G, China's communications industry is expected to continue to lead the global trend of technological innovation and inject strong impetus into the construction of a smart society.
