Deep integration of intelligence

AI internalization: AI technology will shift from external applications to being embedded within the network architecture, providing a full workflow lifecycle environment for the network and achieving a deep integration of communication, sensing, computing and intelligence. For instance, China Mobile's 6G smart and simplified wireless network integrates AI technology within the network architecture to achieve intelligent scheduling and optimization of network resources.

Intent-driven enhancement: The concept of intent-driven will further develop. Through technologies such as natural language processing, the network can more accurately understand users' business needs and intentions, and automatically configure and optimize the network. For instance, Huawei's IDN architecture, driven by intent, can reduce network configuration time from hours to minutes, and achieve a policy execution accuracy rate of 98%.

The network architecture is extremely simple

Platform-based sharing: The hardware architecture will shift from the traditional "siloed stacking" to platform-based sharing, integrating heterogeneous hardware resources to achieve computing power pooling and unified management. For instance, the 6G smart and simplified wireless network, through technologies such as Chiplet and heterogeneous accelerators, reconstructs the hardware foundation of the wireless network, and is expected to increase the utilization rate of computing power to over 75%.

Functional modularization: Network functions will be atomically decomposed and molecular-reconstructed. Based on service-oriented RAN, protocol functions will be decomposed into atomic services, and then dynamically combined through AI-driven to achieve "on-demand networking". For instance, in a certain intelligent manufacturing scenario, this molecular reconstruction has increased the air interface efficiency of industrial AR services by 55%, while reducing computing resource consumption by 35%.