| 55 | 0 | 115 |
| 下载次数 | 被引频次 | 阅读次数 |
面向第六代移动通信(Sixth-Generation Mobile Communication, 6G)复杂非线性通信场景，传统物理层面临模型失配瓶颈，而纯数据驱动人工智能(Artificial Intelligence, AI)模型则存在泛化能力脆弱与可解释性缺失等落地痛点。针对上述问题，提出一种基于机理与数据双重驱动的智能物理层框架。该框架将经典通信数学算子，如星座映射与离散傅里叶变换(Discrete Fourier Transform, DFT),重构为算子级神经网络，实现了物理拓扑先验与深度学习(Deep Learning, DL)的有机融合。此外，设计了自适应回滚(rollback)机制以保障极端工况下的鲁棒性，并深入探讨了模块化迁移部署与全链路可微等关键使能技术。论证了该架构在保留系统可解释性的同时，能实现性能的跨越式提升，为未来6G“物理层智能”(AI-native PHY)的工程落地提供了可行范式。
Abstract：Facing complex and non-linear communication scenarios of Sixth-Generation Mobile Communication(6G) networks, traditional physical layer designs encounter the bottleneck of model mismatch, whereas pure data-driven Artificial Intelligence(AI) models suffer from practical deployment challenges, including fragile generalization and a lack of interpretability.To address these problems, this paper proposes an intelligent physical layer framework driven jointly by physical mechanisms and data.The framework reconstructs classical mathematical communication operators, such as constellation mapping and Discrete Fourier Transform(DFT),into operator-level neural networks, achieving an organic integration of physical topology priors and Deep Learning(DL).Furthermore, an adaptive rollback mechanism is designed to guarantee robustness under extreme channel conditions.Key enabling technologies, including modular transfer deployment and full-link differentiability, are also thoroughly investigated.This research demonstrates that the proposed architecture achieves a substantial performance leap while preserving system interpretability, offering a highly feasible paradigm for the engineering implementation of future 6G AI-native PHY systems.
[1] Al-Fuqaha A,Guizani M,Mohammadi M,et al.Internet of Things:a survey on enabling technologies,protocols,and applications[J].IEEE Communications Surveys & Tutorials,2015,17(4):2347-2376.
[2] Ge N,Jiang C X,Lu J H.A systematic perspective on communication innovations toward 6G[J].Engineering,2023,31(12):19-24.
[3] Mahmood N H,Alves H,López O A,et al.Six key features of machine type communication in 6G[C]//2020 2nd 6G Wireless Summit(6G SUMMIT).Levi:IEEE,2020:1-5.
[4] You X H,Wang C X,Huang J,et al.Towards 6G wireless communication networks:vision,enabling technologies,and new paradigm shifts[J].Science China Information Sciences,2021,64(1):5-78.
[5] Liu F,Cui Y H,Masouros C,et al.Integrated sensing and communications:toward dual-functional wireless networks for 6G and beyond[J].IEEE Journal on Selected Areas in Communications,2022,40(6):1728-1767.
[6] Pan C H,Ren H,Wang K Z,et al.Reconfigurable intelligent surfaces for 6G systems:principles,applications,and research directions[J].IEEE Communications Magazine,2021,59(6):14-20.
[7] Demir Ö T,Björnson E,Sanguinetti L.Foundations of user-centric cell-free massive MIMO[PP/OL].V3.arXiv(2023-11-27)[2026-02-26].https://arxiv.org/abs/2108.02541.
[8] Luo C,Hu J,Xiang L P,et al.Reconfigurable intelligent sensing surface aided wireless powered communication networks:a sensing-then-reflecting approach[J].IEEE Transactions on Communications,2024,72(3):1835-1848.
[9] Zhou S C,Xiang L P,Wang Y,et al.Extended target adaptive beamforming for ISAC:a perspective of predictive error ellipse[J].IEEE Transactions on Wireless Communications,2026,25:10604-10617.
[10] Zhang S L,Hu J,Zhao Y Z,et al.Low complexity pilot power control for long-term energy sustainability in cell-free networks[J/OL].IEEE Transactions on Vehicular Technology(Early Access),2025:1-13.https://ieeexplore.ieee.org/document/11229884.
[11] O'shea T,Hoydis J.An introduction to deep learning for the physical layer[J].IEEE Transactions on Cognitive Communications and Networking,2017,3(4):563-575.
[12] Hoydis J,Cammerer S,Aoudia F A,et al.Sionna:an open-source library for next-generation physical layer research[PP/OL].V2.arXiv(2023-03-20)[2026-02-26].https://arxiv.org/abs/2203.11854.
[13] He H T,Jin S,Wen C K,et al.Model-driven deep learning for physical layer communications[J].IEEE Wireless Communications,2019,26(5):77-83.
[14] Xiang L P,Cui J W,Hu J,et al.Polar coded integrated data and energy networking:a deep neural network assisted end-to-end design[J].IEEE Transactions on Vehicular Technology,2023,72(8):11047-11052.
[15] Hu J,Cui J W,Xiang L P,et al.End-to-end design of polar coded integrated data and energy networking[J].IEEE Transactions on Communications,2024,72(11):7017-7031.
[16] Chataut R,Akl R.Massive MIMO systems for 5G and beyond networks—overview,recent trends,challenges,and future research direction[J].Sensors,2020,20(10):2753.
[17] Wang T Q,Wen C K,Wang H Q,et al.Deep learning for wireless physical layer:opportunities and challenges[J].China Communications,2017,14(11):92-111.
[18] Nan G S,Li Z C,Zhai J L,et al.Physical-layer adversarial robustness for deep learning-based semantic communications[J].IEEE Journal on Selected Areas in Communications,2023,41(8):2592-2608.
[19] Jagannath A,Jagannath J,Melodia T.Redefining wireless communication for 6G:signal processing meets deep learning with deep unfolding[J].IEEE Transactions on Artificial Intelligence,2021,2(6):528-536.
[20] He H T,Wen C K,Jin S,et al.Deep learning-based channel estimation for beamspace mmWave massive MIMO systems[J].IEEE Wireless Communications Letters,2018,7(5):852-855.
[21] Luo C,Xiang L P,Hu J,et al.Bedrock models in communication and sensing:advancing generalization,transferability,and performance[PP/OL].arXiv(2025-03-11)[2026-02-26].https://arxiv.org/abs/2503.08220.
[22] Jiang T.Constructingair-interface links for mobile communications:from {0,1} to [0,1][J].Engineering,2025,46(3):17-23.
[23] Zappone A,Di Renzo M,Debbah M.Wireless networks design in the era of deep learning:model-based,AI-based,or both?[J].IEEE Transactions on Communications,2019,67(10):7331-7376.
[24] Hendrycks D,Gimpel K.A baseline for detecting misclassified and out-of-distribution examples in neural networks[PP/OL].V3.arXiv(2018-10-03)[2026-02-26].https://arxiv.org/abs/1610.02136.
[25] Hoydis J,Ait Aoudia F,Cammerer S,et al.Sionna RT:differentiable ray tracing for radio propagation modeling[C]//2023 IEEE Globecom Workshops(GC Wkshps).Kuala Lumpur:IEEE,2023:317-321.
[26] Bridle J S.Probabilistic interpretation of feedforward classification network outputs,with relationships to statistical pattern recognition[M]//Soulié F F,Hérault J.Neurocomputing:Algorithms,Architectures and Applications.Berlin:Springer,1990:227-236.
[27] Bengio Y,Léonard N,Courville A.Estimating or propagating gradients through stochastic neurons for conditional computation[PP/OL].arXiv(2013-08-15)[2026-02-26].https://arxiv.org/abs/1308.3432.
[28] Luo C,Xiang L P,Hu J,et al.Algorithm design and prototype validation for reconfigurable intelligent sensing surface:forward-only transmission[J].IEEE Transactions on Wireless Communications,2025,25:2860-2875.
基本信息:
中图分类号:TN929.5
引用信息:
[1]杨鲲,向路平,罗程.面向可落地的智能物理层框架与关键技术[J].无线电通信技术,2026,52(03):495-505.
基金信息:
国家自然科学基金(62301122); 江苏省基础研究重大项目(BK20243059); 姑苏创新项目(ZXL2024360); 苏州市高新区项目(RC2025001)~~
2026-05-09
2026-05-09
2026-05-09