通信学报2025，Vol.46Issue(6)：89-100,12.DOI:10.11959/j.issn.1000-436x.2025084
基于变分自编码器的太赫兹信道多径分簇算法
Variational autoencoder-based multipath clustering algorithm for terahertz channels
摘要
Abstract
To address the shortcomings of multipath clustering algorithms in terahertz channel modeling,particularly in terms of multidimensional parameter adaptability and unsupervised feature separation,a variational autoencoder-based latent space multipath clustering(VAE-LMC)model was proposed.Firstly,the variational autoencoder(VAE)was uti-lized to learn latent representations of multipath delays and arrival angles,enhancing feature separability.Secondly,K-Means clustering was embedded into the VAE framework,with joint optimization of reconstruction loss,KL divergence,and clustering loss functions to resolve the challenges of feature separation in unsupervised learning.Finally,multipath clustering was performed in the latent space,and the results were mapped back to the real data space.Terahertz channel measurements at 129.5～135 GHz were conducted in a small factory scenario to construct training datasets and testing da-tasets.Experimental results demonstrate that the VAE-LMC model exhibits significant advantages in intra-cluster and inter-cluster characteristics,environmental consistency,and computational complexity,providing an efficient solution for terahertz channel multipath clustering in complex scenarios.关键词
太赫兹信道/信道测量/多径分簇/变分自编码器/无监督学习Key words
terahertz channel/channel measurement/multipath clustering/variational autoencoder/unsupervised learning分类
信息技术与安全科学引用本文复制引用
郝昕宇,廖希,郑相全,王洋,林峰,陈前斌,张杰..基于变分自编码器的太赫兹信道多径分簇算法[J].通信学报,2025,46(6):89-100,12.基金项目
国家自然科学基金资助项目(No.62271095,No.62171071) （No.62271095,No.62171071）
重庆市自然科学基金资助项目(No.cstc2021jcyj-msxmX0634,No.CSTB2022NSCQ-MSX1125) （No.cstc2021jcyj-msxmX0634,No.CSTB2022NSCQ-MSX1125）
重庆市教委科学技术研究资金资助项目(No.KJZD-K202300607) （No.KJZD-K202300607）
重庆市自然科学基金创新发展联合基金资助项目(No.CSTB2022NSCQ-LZX0037) The National Natural Science Foundation of China(No.62271095,No.62171071),The Natural Science Founda-tion of Chongqing(No.cstc2021jcyj-msxmX0634,No.CSTB2022NSCQ-MSX1125),The Science and Technology Research Program of Chongqing Municipal Education Commission(No.KJZD-K202300607),The Natural Science Foundation Innovation and Develop-ment Joint Fund Project of Chongqing(No.CSTB2022NSCQ-LZX0037) （No.CSTB2022NSCQ-LZX0037）
