面向OTFS的RepViT轻量化卷积神经网络信号检测方法

Lightweight Convolutional Neural Network Signal Detection Method Based on RepViT for OTFS Systems

摘要

正交时频空(OTFS,Orthogonal Time Frequency Space)调制技术能抑制多普勒频移,适配快时变信道,是6G核心候选技术之一.目前,基于深度学习的信号检测算法已在OTFS系统中得到广泛应用.RepViT(Revisiting Mobile CNN From ViT Perspective)融合ViT(Vision Transformer)和MobileNet网络,提升特征提取能力,并通过结构重参数化技术减少计算量.本文提出了一种基于RepViT的OTFS信号检测方法,其将时延多普勒域均衡后的信号输入RepViT网络以修正均衡残差,提高检测准确性.仿真表明,该方案在加性高斯白噪声(AWGN,Additive White Gaussian Noise)、扩展车辆A模型(EVA,Extended Vehicular A model)和扩展典型城市模型(ETU,Extended Typical Urban model)信道下,较基准方案降低复杂度并提升误比特率性能.

Abstract

Orthogonal Time Frequency Space (OTFS) modulation technology can suppress Doppler frequency shift and adapt to fast time-varying channels,making it one of the core candidate technologies for 6G.Currently,deep learning-based signal detection algorithms have been widely applied in OTFS systems.RepViT combines Vision Transformer (ViT) and MobileNet networks to enhance feature extraction capability,and reduces computational complexity through structural reparameterization technology.This paper proposes a lightweight convolutional neural network signal detection method based on RepVitT for OTFS systems,which inputs the signal after delay-Doppler domain equalization into the RepViT network to correct the equalization residual and improve detection accuracy.Simulations show that the proposed scheme reduces complexity and improves bit error rate performance compared with the benchmark schemes under AWGN,EVA,and ETU channels.

关键词

正交时频空OTFS / 轻量化卷积神经网络RepViT / 信号检测 / 结构重参数化

Key words

OTFS / lightweight convolutional neural network RepViT / signal detection / structural reparameterization

[1] Liu R,Hua M,Guan K,et al.6G enabled advanced transportation systems[J].IEEE Transactions on Intelligent Transportation Systems,2024,25(9):10564-10580.
[2] LIAO Y,YU C.A review of message passing based OTFS signal detection algorithms[J]Telecommunication Technology,2025,65(2):188-198.
[3] Hadani R,Rakib S,Tsatsanis M,et al.Orthogonal time frequency space modulation[C]//IEEE Wireless Communications and Networking Conference (WCNC),2017:1-6.
[4] Wei Z Q,Yuan W J,Li S G,et al.Orthogonal time-frequency space modulation:a promising next-generation waveform[J].IEEE Wireless Communications,2021,28(4):136-144.
[5] Wei Z Q,Yuan W J,Li S G,et al.Transmitter and receiver window designs for orthogonal time-frequency space modulation[J].IEEE Transactions on Communications,2021,69(4):2207-2223.
[6] Surabhi G,Chockalingam A.Low-complexity linear equalization for OTFS modulation[J].IEEE Communications Letters,2020,24(2):330-334.
[7] Raviteja P,Phan K T,Hong Y,et al.Interference cancellation and iterative detection for orthogonal time frequency space modulation[J].IEEE Transactions on Wireless Communications,2018,17(10):6501-6515.
[8] WANG Y C,ZHOU P,HUANG J.Research status of channel estimation and signal detection techniques for orthogonal time frequency space modulation[J].Journal on Communications,2024,45(9):229-243.
[9] Enku Y K,Bai B M,Wan F,et al.Two-dimensional convolutional neural network-based signal detection for OTFS systems[J].IEEE Wireless Communications Letters,2021,10(11):2514-2518.
[10] Gong Y,Li Q Y,Meng F K,et al.Data-driven deep learning for OTFS detection[J].China Communications,2023,20(1):88-101.
[11] Howard A G,Zhu M,Chen B,et al.Mobilenets:efficient convolutional neural networks for mobile vision applications[C]//IEEE Conference on Computer Vision and Pattern Recognition,2017,doi:10.48550/arXiv.1704.04861.
[12] Wang B,Yuan Z,Zheng S L,et al.Data-driven intelligent receiver for otfs communication in internet of vehicles[J].IEEE Transactions on Vehicular Technology,2024,73(5):6968-6979.
[13] Wang A,Chen H,Lin Z J,et al.RepViT:revisiting mobile cnn from ViT perspective[C]//IEEE/CVF Conference on Computer Vision and Pattern Recognition,2024:15909-15920.
[14] Murali K R,Chockalingam A.On OTFS modulation for high-doppler fading channels[C]//Information Theory and Applications Workshop,2018:1-10.
[15] Hong Y,Thaj T,Viterbo E.Delay-doppler communications:principles and applications[M].MA,USA,Academic Press,2022.
[16] Xiao T,Singh M,Mintun E,et al.Early convolutions help transformers see better[C]//Advances in Neural Information Processing Systems,2021:30392-30400.
[17] Ding X H,Zhang X Y,Ma N N,et al.RepVGG:making vgg-style convnets great again[C]//IEEE/CVF Conference on Computer Vision and Pattern Recognition,2021:13728-13737.
[18] Ding X H,Zhang X Y,Han J G,et al.Diverse branch block:building a convolution as an inception-like unit[C]//IEEE/CVF Conference on Computer Vision and Pattern Recognition,2021:10881-10890.
附中文参考文献:
[2] 廖 勇,于 晨.基于消息传递的OTFS信号检测算法综述[J].电讯技术,2025,65(2):188-198.
[8] 王永川,周 平,黄 局.正交时频空调制的信道估计与信号检测技术研究现状[J].通信学报,2024,45(9):229-243.

基金资助

国家自然科学基金重点项目(62131016)资助.