ShanghaiTech University Collaborates with Huawei to Build "AI+Education" Autonomous Networks

Aligning with China's strategy for new educational infrastructure, ShanghaiTech University, a "Double-First Class" university in the country, is leading the deployment of smart campuses. It has deployed a wireless network across the entire campus, enabling rapid transmission of learning and research data and quick handling of online processes. On this network, students complete interactive learning activities such as group sharing, courseware sharing, assignment delivery, and in-class tests in classrooms. The network lays an efficient, stable digital foundation for the university's teaching, scientific research, and management services.

The university regards smart campus deployment as a crucial topic for digital transformation in higher education. Intelligent transformation, rather than bandwidth improvements alone, should be the focus of campus network upgrades, necessitating user experience assurance through advanced algorithms. To achieve intelligent transformation, the university partnered with Huawei to implement digital management, intelligent O&M, and network self-management using Huawei's network analysis solution, iMaster NCE-CampusInsight, and network agent NetMaster. The solutions enable the university to evolve toward AN Level 4, a process that is accelerated by the rapid advancement of AI technologies.

Transitioning from Service-Oriented Mode to a Production-Oriented Mode, the Campus Network Experiences an Increasing O&M Pressure

The emergence of AI-assisted teaching and scientific research has transformed campus networks from being service-oriented to production-oriented, creating the following new O&M challenges:

• The network scale and scenario complexity increase. Numerous devices (tens of thousands of APs and thousands of switches), along with diverse deployment scenarios, make AP experience optimization and routine O&M challenging.

• A wide range of terminals and services results in diverse requirements. Teachers and students use various wireless terminals and services, including flexible anytime, anywhere study, high-density interactive teaching, and the transmission of huge volumes of scientific research data. These activities pose stringent requirements on network performance.

• Network traffic demonstrates a tidal roaming pattern. Teachers and students move between dormitories, classrooms, and canteens, creating a prominent tidal pattern of Wi-Fi roaming. This further complicates network assurance.

The Network Digital Map Is Integrated with the Network Agent to Facilitate the Evolution Toward AN Level 4

The university's information center believes that teachers' and students' network experience is a significant yardstick for measuring network quality. With this in mind, it collaborated with Huawei to deploy an intelligent network center—"one map + one brain"—based on the network digital map and network agent, enabling an AI-powered campus network featuring autonomous O&M.

• The network digital map enables a transition from network monitoring to panoramic analysis.

The university deployed iMaster NCE-CampusInsight to transform its network O&M toward intelligent network operations. iMaster NCE-CampusInsight's network digital map, implemented using digital twin technology, integrates data from the network, user, and application dimensions, enabling advanced panoramic visualization and analysis. The map records students' network experience across time and space, and accurately traces key metrics, including the APs to which users are connected, signal strength, download speed, and network latency. The map can trace and locate fault causes anywhere, anytime, significantly improving issue-resolution efficiency and transforming network O&M from basic device monitoring toward intelligent, panoramic analysis.

• The network agent enables a transition from passive response to automatic optimization.

NetMaster redefines traditional O&M modes by enabling the Wi-Fi optimization agent, which transforms passive response to proactive optimization. The Wi-Fi optimization agent is the industry's first AI agent that can independently handle network issues. It accurately detects terminal locations, AP layouts, network experience, and other information based on over 200 experience metrics sensed by Huawei APs. This process is equivalent to building a high-precision map, thereby eliminating the need for partial or fragmented network optimization. The Wi-Fi optimization agent leverages a multi-objective, decision-making algorithm (developed based on more than 100 patents) to automatically optimize the network, just as a network expert would, working around the clock. It solves 80% of wireless issues without manual intervention.

Application example 1: In a multimedia classroom where multiple APs are deployed, some seats experience weak wireless signals. Although the teaching is not affected, the agent has already detected the potential risks and automatically calculated the optimal transmit power range for each AP. This improves the signal strength in the classroom and prevents signal interference to adjacent classrooms. Through precise control, the signal strength is restored to full, and the rate is increased by more than 20%, ensuring smooth smart teaching.

Application example 2: Hundreds of APs are deployed in a library, but severe co-channel interference occurs due to the complex building structure. The network agent automatically detects the issue, clearly constructs the relative positions of all APs based on the layout information of the network digital map, analyzes the optimal channel distribution solution, and automatically executes the solution when detecting that no one is using the network. After the optimization, the channel conflict issue is eliminated in the library, and user experience is significantly improved.

The university will accelerate its evolution toward AN Level 4 and enhance the efficiency of autonomous O&M, enabling the system to independently perform complex optimization analysis, make decisions, and even assist O&M engineers with hardware troubleshooting, thereby facilitating the development of smart campuses.