[Retracted] Radio Frequency Link and User Selection Algorithm for 5G Mobile Communication System

Based on the millimetre‐wave time‐domain channel detector, this paper conducts channel measurement and modelling for different scenarios and antenna types in typical frequency bands and thus carries out a study on the propagation characteristics of millimetre‐wave channels in typical frequency bands and scenarios for 5G. The time‐varying characteristics and modelling methods of millimetre‐wave clusters are studied, and the time‐varying channel simulation strategy and linear regression fitting method are used to establish the time‐varying generation and extinction model of clusters by combining the measured data of 26 GHz large‐scale antenna channels. The algorithm is based on the correlation dictionary and achieves the efficient and optimal selection of large‐scale antenna arrays in the millimetre‐wave band. A low‐complexity multiuser hybrid precoding design scheme is proposed based on the wideband millimetre‐wave large‐scale Multiple‐Input Multiple‐Output (MIMO) system. The scheme aims to maximize the system spectral efficiency and introduces the net spectral efficiency as a bridge to decouple the analogy precoding design from the digital precoding design, thus significantly reducing the computational complexity. In the analogy precoding stage, a Hungarian algorithm‐based beam assignment method is proposed to avoid beam conflicts and maximize the net spectral efficiency of the system; in the digital precoding stage, the equivalent channel matrix of each subcarrier is diagonalized to eliminate multiuser interference. Finally, the effectiveness of the scheme is verified by simulation. In this paper, we focus on the radio frequency (RF) link and user selection algorithm of 5G mobile communication system. Through the study of these two aspects, the RF link of our communication system has high stability and reliability, and through the study of the algorithm used for selection, our results also have a strong practical value.