Artificial Intelligence based Architecture and Implementation of Wireless Network

The explosion of 5G network data continues to advance the evolution of air interface technology and network architecture, which enables the network to support more frequency bands, antennas, more diversified frame structures and physical layer control procedures. Traditional operation and maintenance patterns of human intervention or threshold setting control are no longer reliable, thus the automated and intelligent management mechanisms are proposed to improve the network resource utilization and user experience. This article summarizes the research of implementations, algorithms, and optimizations of AI firstly. Then through the research of various international standards, different AI networked architecture design of the core network, RAN and O-RAN are introduced. The paper highlights UE trajectory prediction and energy-saving use cases of the wireless network, the details mainly focus on the comparative study of algorithms and signaling realization. Last but not the least, the intelligent network platform that has been applied in the industry is introduced, which enriches the ideas for future development of providing more reliable networks to vertical industries.