Tsinghua University Collaborates with Huawei to Deploy the Industry's First Campus Network Agent

As universities' informatization construction enters a critical phase under the guidance of "new infrastructure for education", intelligent networks have become the intelligent foundation for academic breakthroughs and talent cultivation. Faced with exponential growth in the demand for AI applications, online teaching, and smart campus life from more than 70,000 teachers and students, Tsinghua University, adhering to the concept of leading the pace of informatization development in Chinese universities through technological innovation, has made early plans for the future development direction of network O&M, such as digital twins, foundation models, and agents. Tsinghua University has also worked with Huawei to build a future-oriented intelligent O&M platform, setting up a golden standard for education infrastructure in the digital era.

How to Avoid Heavy O&M Burden in a Campus with Tens of Thousands of Users and a Network with Tens of Thousands of Sites?

O&M personnel on the university's campus network often struggled to provide detailed reports on network quality or to identify root causes after users reported faults or frame-freezing issues. This significant challenge not only affected user experience but also created tremendous pressure for the university's network O&M team. They were in urgent need of intelligent O&M techniques.

• A must-answer question for universities: multiple vendors, multiple devices, and fewer personnel with more work

This is not only a question for Tsinghua University, but also a collective reflection of the digital transformation of Chinese universities. Universities across China are facing the challenges of multi-vendor devices, multiple management interfaces, and fragmented O&M data. With tens of thousands of campus network devices and hundreds of thousands of daily terminal loads, the demand for wireless services such as online teaching and smart classrooms continues to grow. The traditional "human-monitoring-device" mode can no longer meet the requirements. As the "New infrastructure for education" strategy accelerates, digital and intelligent network O&M has become a must for universities to achieve intelligent transformation.

• Complexity: deep water zone for technology and scenario adaptation

In addition to compatibility with Huawei devices and third-party devices on the live network, more than 30,000 APs, 3,000 switches, and 170,000 terminals in Tsinghua University need to be mapped to information nodes in the digital world (such as tens of thousands of sites). Furthermore, AI needs to understand the fault analysis and handling logic behind Wi-Fi freezing in dormitories and user disconnection. These are not only a profound test of new technologies, but also an ultimate challenge to the understanding of education scenarios.

• Necessity: intelligent O&M is critical to the future of education

As Tsinghua's online learning platform serves millions of students and smart classrooms reshape the traditional teaching model, campus networks have transcended their role as mere tools and have become the "digital neurons" of national strategic scientific strength. A network fault may interrupt cross-campus joint experiments, and a bandwidth bottleneck may significantly increase the interaction time in smart classrooms. Every outcry of "network exception" is a knock on the door of educational equity. Therefore, intelligent O&M is imperative.

Huawei Xinghe Intelligent Network, One Map for a Panoramic View and One Brain for Autonomous Driving

To achieve intelligent, future-proof O&M, the university collaborates with Huawei to deploy Huawei's iMaster NCE-CampusInsight. iMaster NCE-CampusInsight employs digital twin technology to mirror the network and create models based on the network status and user experience, and displays multidimensional information, including network, user, and terminal information, on a network digital map, offering a panoramic view for intelligent O&M. The university also incorporates the DeepSeek foundation model, transforming the interaction mode from complex menus toward natural languages. Additionally, an agent is deployed to automatically handle network faults in complex environments, substantially improving network O&M efficiency and the network satisfaction of teachers and students.

• Unified management of multiple vendors and protocols

To address challenges such as multi-vendor management, O&M data silos, and lack of a unified monitoring perspective, Huawei's iMaster NCE-CampusInsight integrates multi-vendor data through protocols including Telemetry, SNMP, and syslog. It replaces multiple O&M systems in the university and serves as the only entry for monitoring the status of all devices, analyzing network traffic, and parsing abnormal data. This system supports digital operations across all scenarios from routine inspection to major event assurance. This unified yet diverse device management capability paves a smooth, evolutionary path for universities to operate and maintain their existing multi-vendor devices.

• Panoramic view of the network digital map

The network digital map displays the network experience data of 30,000 devices, 70,000 teachers and students, and 170,000 terminals, offering a panoramic view for network O&M. iMaster NCE-CampusInsight employs digital twin and modeling technologies to build an experience evaluation system, detects users' real-time network experience, and displays the detection result from the time and space dimensions, including the access location, access time, access process (association, authentication, and DHCP phases), network experience, roaming status, and issues. This information enables network engineers to quickly resolve suboptimal user experience issues, achieving user experience-centric O&M while enhancing both service quality and user satisfaction.

Time-space analysis of user journey in the network digital map

• Autonomous O&M powered by the network agent

The network agent NetMaster closely collaborates with DeepSeek deployed on the campus network, providing two key capabilities: the network O&M copilot and the Wi-Fi optimization agent, enabling the evolution toward autonomous campus network O&M.

The network O&M copilot fundamentally transforms the university's system interaction to a natural-language mode. In this mode, users only need to enter their requirements in natural language, for example, "xx user's download speed is slow. Please locate the cause." The system understands user intent and works with iMaster NCE-CampusInsight to obtain data like users' roaming tracks, network experience, and device status. It then quickly locates AP channel conflicts on the network digital map and pushes a frequency band adjustment and power optimization solution. This "brain-map collaboration" not only makes the foundation model's answers more reliable but also significantly simplifies O&M.

Network O&M copilot of the network agent, natural language interaction,
and brain-map collaboration result display

The Wi-Fi optimization agent is the industry's first agent that handles network issues. It monitors the Wi-Fi network in every dormitory room at the university around the clock and accurately detects AP deployment locations, environmental interference, and terminal network performance. Upon detecting typical Wi-Fi network issues, including weak coverage and high interference, it infers from multiple dimensions—coverage, interference, frequency bandwidth, and load—to formulate and automatically execute the optimal solution. This process is equivalent to solving a multivariable nonlinear equation, whereas traditional network optimization is equivalent to solving a single-variable linear equation. This agent is a milestone that marks the evolution toward autonomous campus network O&M.

Wi-Fi optimization agent, automatically handling Wi-Fi network issues

Huawei Xinghe Intelligent Network as the Basis, Opening a New Chapter in University Network O&M

Guided by the Digital China strategy, intelligent transformation is becoming a core driver of the new education infrastructure. In this context, the intelligent O&M solution co-created by Tsinghua University and Huawei injects three innovative momentums into modern education and opens a new chapter in university network O&M.

• O&M innovation: AI unleashes manpower potential

Foundation models and digital twin technologies are deeply integrated, shifting the technical team from "passive firefighting" to network architecture optimization and cutting-edge technology breakthroughs, and empowering the technical team with new-quality productive forces.

• Campus quality improvement: reshaping smart campus experience

NetMaster provides 24/7 support for teachers' and students' network experience, significantly improving user satisfaction. It supports innovative scenarios such as 4K/8K online teaching, AI-based scientific research, and smart classrooms, ensuring smooth access to 170,000 terminals per day.

• Industry enablement: new paradigm for autonomous driving networks

The project has developed a replicable autonomous driving network (ADN) solution for campuses, setting an industry benchmark for autonomous driving network O&M.

At the 2025 China Cloud Network Conference, Tsinghua University, as the industry's first practice representative to implement the network agent for campus network optimization, joined China Academy of Information and Communications Technology (CAICT) and Huawei Technologies Co., Ltd. to launch the Autonomous Networks Navigator Program 2.0 – Network Operation and Maintenance Management Intelligent Agent Evaluation Pioneer Action. This action aims to build a future-oriented intelligent network O&M platform and lead the digital transformation of higher education institutions across China.

During HUAWEI CONNECT, Lu Haiou, President of Huawei's General Development Department, presented Tsinghua University with the "First Deployment of Huawei's Campus Network Agent NetMaster" commemorative medal.

During the China Cloud Network Conference, Wang Hui, President of the NCE Data Communication Domain of Huawei's Data Communication Product Line, presented a copy of the book Advancing Toward Autonomous Networks: IP ADN to Ma Yunlong, a senior engineer at Tsinghua University's Information Technology Center.

Outlook

Looking ahead, Tsinghua University will continue to work with Huawei to explore autonomous network solutions tailored to higher education campuses. NetMaster acts like a smart cockpit for campus networks, making network operations as simple and efficient as driving an autonomous vehicle.