Intelligent beam training method and precoding system in millimeter wave communication system

Abstract

The invention discloses an intelligent beam training method and a precoding system in millimeter wave communication. The intelligent beam training method utilizes the time-space correlation of channels to determine the codebook interval where the optimal beam is located by perceiving changes in the environment. Specifically, the present invention uses the Markov decision process (MDP) to model the problem; solves it based on deep reinforcement learning, innovatively constructs a graphical state space to effectively utilize the structural information of the channel beam space, and proposes a difference Optimized action space to improve learning efficiency and convergence speed. The intelligent beam training method designed in the present invention can perceive the rate of change of the environment and adaptively adjust the beam training interval, which not only reduces the beam training overhead, but also effectively improves the system throughput.

Description

technical field [0001] The invention belongs to the field of wireless communication, and in particular relates to a precoding system of an intelligent beam training method in a millimeter wave communication system. Background technique [0002] The explosive growth in mobile data demand is posing unprecedented challenges for wireless service providers to overcome due to bandwidth shortages. As a key technology, millimeter wave has attracted widespread attention due to its large bandwidth characteristics that can meet the needs of high-speed data services. Compared with microwaves, mmWave signals have more serious path loss due to their short wavelength, but the path loss can be offset by installing a large-scale antenna array at the receiving / transmitting end. Considering the power consumption and cost of radio frequency (RF) links, mmWave communication systems usually adopt a mixed-analog-digital architecture, where the number of RF links is far less than the number of ant...

AI Technical Summary

Problems solved by technology

Using the above search method in a large-scale antenna array system will cause a huge beam training overhead

Especially in dynamic or time-varying scenarios, frequent switching of beams is required to obtain high-speed transmission rates, but the rapidly changing mmWave channel cannot achieve frequent and accurate beam training due to short coherence time, which also brings challenges to beam training

Images