Segment-Wise Flow Matching for Vision-Aided mmWave V2I Beam Prediction

Fuente: arXiv
Saved in:
Bibliographic Details
Main Authors: Zheng, Can, He, Jiguang, Kang, Chung G., Cai, Guofa, Huang, Chongwen, Wymeersch, Henk
Format: Preprint
Published: 2025
Subjects:
Online Access:
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866910120136409088
author Zheng, Can
He, Jiguang
Kang, Chung G.
Cai, Guofa
Huang, Chongwen
Wymeersch, Henk
author_facet Zheng, Can
He, Jiguang
Kang, Chung G.
Cai, Guofa
Huang, Chongwen
Wymeersch, Henk
contents This paper proposes a vision-conditioned flow matching (FM) framework for beam prediction in millimeter-wave vehicle-to-infrastructure links. Instead of modeling discrete beam-index sequences, the proposed method learns the temporal evolution of normalized beam receive power vectors through a continuous vector field governed by an ordinary differential equation, enabling smooth dynamics and efficient sampling. By imposing FM over beam-state transitions and jointly optimizing beam prediction and flow consistency, the proposed framework provides a unified model for future beam prediction. Experimental results show that the proposed FM-based model significantly improves beam prediction performance over baselines, approaches the performance of large language model-based methods, and reduces predictor-side inference latency by about $6.9\times$ on GPU and $2.8\times10^3\times$ on CPU, respectively.
format Preprint
id arxiv_https___arxiv_org_abs_2511_20265
institution arXiv
publishDate 2025
record_format arxiv
spellingShingle Segment-Wise Flow Matching for Vision-Aided mmWave V2I Beam Prediction
Zheng, Can
He, Jiguang
Kang, Chung G.
Cai, Guofa
Huang, Chongwen
Wymeersch, Henk
Signal Processing
This paper proposes a vision-conditioned flow matching (FM) framework for beam prediction in millimeter-wave vehicle-to-infrastructure links. Instead of modeling discrete beam-index sequences, the proposed method learns the temporal evolution of normalized beam receive power vectors through a continuous vector field governed by an ordinary differential equation, enabling smooth dynamics and efficient sampling. By imposing FM over beam-state transitions and jointly optimizing beam prediction and flow consistency, the proposed framework provides a unified model for future beam prediction. Experimental results show that the proposed FM-based model significantly improves beam prediction performance over baselines, approaches the performance of large language model-based methods, and reduces predictor-side inference latency by about $6.9\times$ on GPU and $2.8\times10^3\times$ on CPU, respectively.
title Segment-Wise Flow Matching for Vision-Aided mmWave V2I Beam Prediction
topic Signal Processing
url https://arxiv.org/abs/2511.20265