The in-depth convergence of 5G and vertical industries is accelerating drive-in, and 5G B2B services are experiencing explosive growth. However, service development faces problems such as slow service rollout, low O&M efficiency, and difficult reliability assurance. It is urgent to introduce new technical means to break the situation. Based on the Mixture of model architecture, multi-agent technologies ensure the full lifecycle of 5G B2B services. Before service provisioning: Requirements from the government and enterprise department to the office management, resource allocation, and work order generation. Service provisioning: Based on the large language model (LLM), configuration agents automatically generate natural language instructions to standard configurations and pre-review compliance. The intelligent simulation algorithm simulates and verifies the generated configuration file to ensure reliable solutions and apply them to networks, ensuring agile service rollout and improving mobile network operation security. After service provisioning: Complaint handling agents and alarm handling agents ensure efficient service O&M. Complex signaling messages can be automatically parsed based on signaling model. Complaint/alarm handling agents can automatically identify intentions, automatically diagnose based on the chain of thoughts, and intelligently fill in TTs based on analysis conclusions. One-stop automatic closure of complaints/alarms, greatly improving efficiency.
By 2025, the number of IoT connections in China has exceeded 2.8 billion, with an annual growth rate of about 22% in the past three years. The industry is developing rapidly. According to data from the Ministry of Industry and Information Technology (MIIT), more than 20,000 5G private network have been deployed, covering key industries such as internet of vehicles(IoV), energy, transportation, and healthcare. 5G has become the core engine of industry digital transformation. Challenge 1: Service configuration depends on manual operations, resulting in low efficiency and reliability. In the face of complex networking, massive rule parameters, frequent service provisioning operations, configuration preparation, review, and impact evaluation rely heavily on expert experience, resulting in low efficiency and reliability. A large number of network incidents are caused by incorrect configuration. Therefore, AI is urgently required to implement automatic configuration. Challenge 2: The alarm and complaint handling efficiency is insufficient, and the SLA cannot be ensured. Take a single province as an example. O&M personnel handle tens of thousands of alarm tickets and thousands of complaint tickets every year, causing great pressure. The traditional manual mode is inefficient and cannot meet the strict SLA requirements of the IoT. AI needs to be introduced to improve efficiency and security.
