Efficient Joint Transmit and Receive Beam Alignment via Sequential CNN LSTM Networks

This paper introduces a deep learning-assisted joint transmit and receive beam tracking approach for uplink multiple-input multiple-output (MIMO) communication over millimeter wave (mmWave) channels. In current wireless networks, beam alignment between transmitter and receiver is necessary to guaran...

Full description

Saved in:
Bibliographic Details
Main Authors: Takumi Yoshida, Koji Ishibashi, Hiroki Iimori, Paulo Valente Klaine, Szabolcs Malomsoky
Format: Article
Language:English
Published: IEEE 2025-01-01
Series:IEEE Access
Subjects:
Online Access:https://ieeexplore.ieee.org/document/11059919/
Tags: Add Tag
No Tags, Be the first to tag this record!
Description
Summary:This paper introduces a deep learning-assisted joint transmit and receive beam tracking approach for uplink multiple-input multiple-output (MIMO) communication over millimeter wave (mmWave) channels. In current wireless networks, beam alignment between transmitter and receiver is necessary to guarantee a strong and stable connection. However, to prevent misalignment due to user equipment (UE) mobility, frequent beam training is required, resulting in significant training overhead. To address this issue at the minimum performance degradation, a sequential convolutional neural network (CNN)–long short-term memory (LSTM) network is presented to capture the spatio-temporal correlation of the channel associated with UE movements. To further enhance the achievable rate, a deep learning model for codebook-free receive beam design is presented, which is capable of achieving the optimum maximum ratio combining (MRC) method without instantaneous channel state information (CSI) knowledge. Simulation results demonstrate that the proposed deep learning-based beam tracking methods are superior to traditional beam tracking algorithms in terms of beamforming gain and effective spectral efficiency.
ISSN:2169-3536