A Geometry-Based RIS-Assisted Multi-User Channel Model with Deep Reinforcement Learning

In this paper, a 3D geometry-based stochastic channel model (GBSM) is proposed for RIS-assisted multi-user communications. The proposed GBSM is divided into two sub-channels, that is, BS-RIS and RIS-Rx links, and propagation distances and angles of multipath components are derived to describe multi-...

Full description

Saved in:
Bibliographic Details
Published in2024 IEEE 99th Vehicular Technology Conference (VTC2024-Spring) pp. 1 - 5
Main Authors Yuan, Yuan, He, Ruisi, Ai, Bo, Wu, Tong, Chen, Ruifeng, Zhang, Zhengyu, Jin, Yunwei, Zhong, Zhangdui
Format Conference Proceeding
LanguageEnglish
Published IEEE 24.06.2024
Subjects
Channel capacity
channel model
Deep reinforcement learning
Delays
Multiuser
Reconfigurable intelligent surface (RIS)
Reconfigurable intelligent surfaces
Stochastic processes
Three-dimensional displays
Vehicular and wireless technologies
Online AccessGet full text

Cover

More Information
Summary:In this paper, a 3D geometry-based stochastic channel model (GBSM) is proposed for RIS-assisted multi-user communications. The proposed GBSM is divided into two sub-channels, that is, BS-RIS and RIS-Rx links, and propagation distances and angles of multipath components are derived to describe multi-user channels. In addition, optimization objective for multi-user channel is proposed, and deep reinforcement learning is introduced to solve high-dimensional RIS phase problem. Based on the proposed model and solved RIS phase, channel capacity and root mean square delay spread are derived. The simulation results show that RIS optimization parameters and channel parameters have major impact on channel characteristics. The conclusions can provide a reference for designing and developing of RIS-assisted multi-user systems.
ISSN:2577-2465
DOI:10.1109/VTC2024-Spring62846.2024.10683242