移动通信2024，Vol.48Issue(12)：39-45,75,8.DOI:10.3969/j.issn.1006-1010.20241023-0002
基于机器学习的车联网毫米波波束预测
A Machine Learning-Driven Approach for MmWave Beam Prediction in Vehicular Networks
摘要
Abstract
To address the issue of beam alignment in vehicular millimeter-wave communications,this paper proposes a machine learning-based beam prediction method incorporating panoramic traffic context coding.By simulating a dual-lane urban canyon environment,a large amount of environmental data are collected and encoded into panoramic traffic context information.Based on the coded information,a machine learning model is employed to train and predict the optimal beams.Simulation results demonstrate that,compared to traditional beam selection methods,the proposed beam alignment approach leverages environmental information more effectively,improving the accuracy and robustness of beam prediction of millimeter-wave communication in dynamic scenarios.关键词
车联网/毫米波通信/波束预测/全景交通上下文编码/机器学习Key words
vehicle networking/millimeter wave communication/beam prediction/panoramic traffic context coding/machine learning分类
电子信息工程引用本文复制引用
靳昊文,仲伟志,刘响,王文捷,林志鹏..基于机器学习的车联网毫米波波束预测[J].移动通信,2024,48(12):39-45,75,8.基金项目
江苏省重点研发计划(产业前瞻与关键核心技术)"6G普适无线信道建模理论方法与性能关键技术研发"(BE2022067,BE2022067-1,BE2022067-3) （产业前瞻与关键核心技术）
