Network intelligence refers to the use of intelligent information and communication technology (ICT) to achieve automation functions such as intent, perception, analysis, decision-making, and execution. It enables rapid processing and efficient management of information resources such as networks through technologies like data collection, analysis, and mining. In October 2022, TM Forum, an international organization for communication network management, collaborated with CCSA, China Mobile, ZTE, and 54 other industry partners to release the "Self-Intelligent Network White Paper 4.0." This initiative aims to promote the empowerment of industry digital transformation through self-intelligent networks, covering the entire process from strategy to implementation.

The construction of "self-intelligent networks" is designed to enhance the business agility and cost-effectiveness of operator networks through fully automated, intelligent networks and ICT infrastructure, agile operations, and all-scenario services. It aims to create telecommunication networks that are "self-service, self-provisioning, and self-assuring." Additionally, it aims to provide vertical industries with a customer experience of "zero waiting, zero contact, and zero faults," ensuring and enhancing the business experience of industry users in areas such as smart campuses, industrial Internet, research and development, and intelligent driving.

As the proportion of the digital economy in the national economy gradually increases, computational power will directly impact the speed of digital economic development and the height of social intelligence. According to the IDC Global Compute Index Assessment Report 2021-2022, computational power is closely related to the national economy. An average increase of one percentage point in computational power scale will drive a 3.5‰ growth in the national digital economy and a 1.8‰ growth in GDP. The "National Integrated Big Data Center Collaborative Innovation System Computational Hub Implementation Plan" has been released, including the "computational power network" in the scope of national new infrastructure development.

ZTE believes that in the process of convergence between computation and networking, network intelligence is a key technology for achieving autonomous perception, autonomous decision-making, autonomous learning, and autonomous optimization of the network.

Network intelligence helps address core requirements in efficient scenarios, such as effective connection of computational power, efficient scheduling of computing resources, effective allocation of storage resources, and green, low-carbon, and energy-efficient computing. Leveraging new technologies such as big data and AI, network intelligence focuses on applications, driving the integration of computation and networking to take root.

On the one hand, through the collaboration of computation and networking, it serves as a cohesive force to create a broader fusion of computation and networking. This includes cross-domain network collaboration, vertical cross-layer network-business collaboration, and cross-disciplinary collaboration. By reconstructing computational and network resources, a new type of intelligent digital foundation is created to support low-latency and highly reliable applications in industry scenarios. It builds a unified enterprise intelligent network platform, smooths infrastructure differences, achieves an improved user experience, and reduces costs.

On the other hand, demands such as intelligent operation and agile operation expand the boundaries of network intelligence from application and operation intelligence to the establishment of unified and distributed intelligent surfaces. ZTE's latest uSmartNet 2.0 autonomous evolution solution aims to establish a network-layered architecture with internal generation of network elements, single-domain autonomy, and cross-domain collaboration. It deploys intelligent layered, domain-based, and hierarchical evolutionary capabilities, introduces intent-driven and digital twinning, achieves business-driven adaptive closed-loop, facilitates rapid development and iterative deployment of intelligent network applications, and supports operators in building a win-win and symbiotic new ecosystem of computation and networking convergence.

Based on industry practices, ZTE has created multiple 5G private network cases in various fields. In the construction of smart campuses, a 5G private network can provide high-speed, low-latency, and highly reliable network communication services for the campus. However, the operation and maintenance of a 5G private network in smart campus scenarios face various challenges, such as complex network architecture, high-density user access, diverse application scenarios, and high reliability requirements.

In the operation and maintenance scenarios of 5G private networks in edge-sinking scenes, ZTE has built a solution for smart campus 5G private networks. It actively discovers faults and hidden dangers in the 5G private network of the campus, timely distributes problem tickets to various specialties, and ensures the quality of industry customer network use. Based on a layered architecture of data collection, capability construction, and application presentation, ZTE creates a deployable, cost-effective, and replicable characteristic solution. It provides a real-time monitoring business screen for the campus, a centralized operation and maintenance platform for campus business quality management, and develops OpenAPI capabilities to release capabilities such as end-to-end global data correlation, AI-driven delineation and traceability, and real-time monitoring and alarms for business quality to application sides. It builds 5G private network operation and maintenance modules and services. At the same time, based on the capabilities provided by the AI cloud-edge-side enterprise version intelligent AI platform, such as AI cloud-edge collaboration, Adlik computational efficiency improvement, and automatic training of network models, it empowers the construction of edge elastic networks, helps operators jointly expand enterprise AI ecological management, and drives enterprise AI ecological operation. Through practical landing, the real-time fault monitoring time delay of the campus 5G private network has been shortened from hours to minutes, the problem detection rate of the 5G private network in the campus has been increased from 20% to 100%, and the fault delineation time has been reduced from over 6 hours to 30 minutes.

Looking to the future, ZTE is willing to work with operators and industry partners to jointly establish a high-level autonomous evolving network. Through layered intelligent evolution, it achieves the overlay of computation on the network and injects intelligence into the network through computation. It is committed to promoting "the digital foundation growing upwards and network intelligence rooting downwards," inspiring the future of computation and network convergence.