With the rapid development of the Internet of Things and wireless communication technologies, human activity recognition based on WiFi Channel State Information (CSI) has become a core research topic in smart homes, health monitoring, and security surveillance, due to its non-invasiveness, privacy protection, and the need for no specialized equipment. CSI can accurately capture amplitude and phase perturbations of wireless channels caused by human movements, providing a reliable data basis for contactless action recognition. However, this technology faces critical challenges such as scarce labeled data, poor cross-scenario adaptability, and insufficient model generalization ability in practical applications. This paper analyzes the technical principles and research status of CSI-based action recognition, focusing on the application of diffusion models in data augmentation. By integrating generative data expansion, cross-domain feature alignment, and model robustness optimization, a complete CSI action recognition enhancement scheme is proposed. It systematically discusses how diffusion models alleviate the few-shot problem through high-fidelity sample generation and improve cross-scenario recognition performance via domain adaptive learning, providing theoretical support and technical references for the practical promotion of CSI action recognition technology.
References
[1]
Yadav, S.K., Sai, S., Gundewar, A., Rathore, H., Tiwari, K., Pandey, H.M., et al. (2022) CSITime: Privacy-Preserving Human Activity Recognition Using WiFi Channel State Information. Neural Networks, 146, 11-21. https://doi.org/10.1016/j.neunet.2021.11.011
[2]
Wang, D., Yang, J., Cui, W., Xie, L. and Sun, S. (2021) Multimodal CSI-Based Human Activity Recognition Using GANs. IEEE Internet of Things Journal, 8, 17345-17355. https://doi.org/10.1109/jiot.2021.3080401
[3]
Wang, Y., Yao, L., Wang, Y. and Zhang, Y. (2021) Robust CSI-Based Human Activity Recognition with Augment Few Shot Learning. IEEE Sensors Journal, 21, 24297-24308. https://doi.org/10.1109/jsen.2021.3111030
[4]
Sheng, B., Xiao, F., Sha, L. and Sun, L. (2020) Deep Spatial-Temporal Model Based Cross-Scene Action Recognition Using Commodity WiFi. IEEE Internet of Things Journal, 7, 3592-3601. https://doi.org/10.1109/jiot.2020.2973272
[5]
El Zein, H., Mourad-Chehade, F. and Amoud, H. (2024) CSI-Based Human Activity Recognition via Lightweight CNN Model and Data Augmentation. IEEE Sensors Journal, 24, 25060-25069. https://doi.org/10.1109/jsen.2024.3414168
[6]
Xiao, C., Han, Y., Yang, W., Hou, Y., Shi, F. and Chetty, K. (2024) Diffusion-Model-Based Contrastive Learning for Human Activity Recognition. IEEE Internet of Things Journal, 11, 33525-33536. https://doi.org/10.1109/jiot.2024.3429245
[7]
Lee, T., Park, J., Kim, H. and Andrews, J.G. (2025) Generating High Dimensional User-Specific Wireless Channels Using Diffusion Models. IEEE Transactions on Wireless Communications, 25, 2907-2921. https://doi.org/10.1109/TWC.2025.3600286
[8]
Zhang, Y., He, F., Wang, Y., Wu, D. and Yu, G. (2023) CSI-Based Cross-Scene Human Activity Recognition with Incremental Learning. Neural Computing and Applications, 35, 12415-12432. https://doi.org/10.1007/s00521-023-08389-0
[9]
Yang, X., Zhang, T., Zhou, M., Chen, B., Wang, Y. and Tong, R. (2026) CSI-Based Respiratory Pattern Recognition Using Diffusion Model and Deep Learning. IEEE Transactions on Network Science and Engineering, 13, 3894-3910. https://doi.org/10.1109/TNSE.2025.3638894
[10]
Chen, Z., Zhang, L., Jiang, C., Cao, Z. and Cui, W. (2019) WiFi CSI Based Passive Human Activity Recognition Using Attention Based BLSTM. IEEE Transactions on Mobile Computing, 18, 2714-2724. https://doi.org/10.1109/tmc.2018.2878233
[11]
Shalaby, E., ElShennawy, N. and Sarhan, A. (2022) Utilizing Deep Learning Models in CSI-Based Human Activity Recognition. Neural Computing and Applications, 34, 5993-6010. https://doi.org/10.1007/s00521-021-06787-w
[12]
Abdelnasser, H., Youssef, M. and Harras, K.A. (2015) WiGest: A ubiquitous WiFi-Based Gesture Recognition System. 2015 IEEE Conference on Computer Communications (INFOCOM), Hong Kong, 26 April-1 May 2015, 1472-1480. https://doi.org/10.1109/infocom.2015.7218525
[13]
Li, M., Meng, Y., Liu, J., Zhu, H., Liang, X., Liu, Y., et al. (2016) When CSI Meets Public WiFi: Inferring Your Mobile Phone Password via WiFi Signals. Proceedings of the 2016 ACM SIGSAC Conference on Computer and Communications Security, Vienna, 24-28 October 2016, 1068-1079. https://doi.org/10.1145/2976749.2978397
[14]
Hao, Y., Nallanathan, A. and Liu, Y. (2025) Latent Diffusion-Inspired Domain-Invariant Human Pose Estimation from WiFi Signals. 2025 IEEE/CIC International Conference on Communications in China (ICCC), Shanghai, 10-13 August 2025, 1-6. https://doi.org/10.1109/iccc65529.2025.11149317
[15]
Niu, J., He, X., Fang, B., Han, G., Wang, X. and He, J. (2024) CSI-F: A Human Motion Recognition Method Based on Channel-State-Information Signal Feature Fusion. Sensors, 24, Article No. 862. https://doi.org/10.3390/s24030862
[16]
Huang, S., Chen, P. and McCann, J. (2023) DiffAR: Adaptive Conditional Diffusion Model for Temporal-Augmented Human Activity Recognition. Proceedings of the 32nd International Joint Conference on Artificial Intelligence, Macao, 19-25 August 2023, 3812-3820. https://doi.org/10.24963/ijcai.2023/424
[17]
Xu, M., Li, Y., Han, K., Zhao, Y., Jiang, J. and Du, Y. (2024) A Conditional Diffusion Model Based WiFi Sensing Enhancement Method. 2024 IEEE 35th International Symposium on Personal, Indoor and Mobile Radio Communications (PIMRC), Valencia, 2-5 September 2024, 1-6. https://doi.org/10.1109/pimrc59610.2024.10817390
[18]
Chen, Z., Shin, H. and Nallanathan, A. (2025) Generative Diffusion Model-Based Variational Inference for MIMO Channel Estimation. IEEE Transactions on Communications, 73, 9254-9269. https://doi.org/10.1109/tcomm.2025.3556753
[19]
Kobir, M.I., Machado, P., Lotfi, A., Haider, D. and Ihianle, I.K. (2025) Enhancing Multi-User Activity Recognition in an Indoor Environment with Augmented Wi-Fi Channel State Information and Transformer Architectures. Sensors, 25, Article No. 3955. https://doi.org/10.3390/s25133955
[20]
Shi, Z., Cheng, Q., Zhang, J.A. and Yi Da Xu, R. (2022) Environment-Robust WiFi-Based Human Activity Recognition Using Enhanced CSI and Deep Learning. IEEE Internet of Things Journal, 9, 24643-24654. https://doi.org/10.1109/jiot.2022.3192973
[21]
Zhang, C. and Jiao, W. (2023) Imgfi: A High Accuracy and Lightweight Human Activity Recognition Framework Using CSI Image. IEEE Sensors Journal, 23, 21966-21977. https://doi.org/10.1109/jsen.2023.3296445
[22]
Deng, F., Jovanov, E., Song, H., Shi, W., Zhang, Y. and Xu, W. (2024) WiLDAR: WiFi Signal-Based Lightweight Deep Learning Model for Human Activity Recognition. IEEE Internet of Things Journal, 11, 2899-2908. https://doi.org/10.1109/jiot.2023.3294004
[23]
Ding, J. and Wang, Y. (2019) WiFi CSI-Based Human Activity Recognition Using Deep Recurrent Neural Network. IEEE Access, 7, 174257-174269. https://doi.org/10.1109/access.2019.2956952
[24]
Xiao, C., Tong, Y., Lu, J., Xue, Y., Tai, W., Cheng, Z., et al. (2025) Graph-Enhanced Multi-Scale Contrastive Learning for Graph Anomaly Detection with Adaptive Diffusion Models. IEEE Transactions on Big Data, 11, 2946-2957. https://doi.org/10.1109/tbdata.2025.3566619
[25]
Zhang, R., Jing, X., Wu, S., Jiang, C., Mu, J. and Yu, F.R. (2021) Device-Free Wireless Sensing for Human Detection: The Deep Learning Perspective. IEEE Internet of Things Journal, 8, 2517-2539. https://doi.org/10.1109/jiot.2020.3024234
[26]
Kim, H., Lee, T., Kim, H., De Veciana, G., Arfaoui, M.A., Koc, A., et al. (2025) Generative Diffusion Model-Based Compression of MIMO CSI. ICC 2025-IEEE International Conference on Communications, Montreal, 8-12 June 2025, 6323-6328. https://doi.org/10.1109/icc52391.2025.11161629
[27]
Lee, T., Park, J., Kim, H. and Andrews, J.G. (2024) High Dimensional User-Specific Channel Generation for CSI Compression Using Diffusion Models. 2024 58th Asilomar Conference on Signals, Systems, and Computers, Pacific Grove, 27-30 October 2024, 759-763. https://doi.org/10.1109/ieeeconf60004.2024.10942761
[28]
Khan, D. and Ho, I.W. (2021) Deep Learning of CSI for Efficient Device-Free Human Activity Recognition. 2021 IEEE 7th World Forum on Internet of Things (WF-IoT), New Orleans, 14 June-31 July 2021, 19-24. https://doi.org/10.1109/wf-iot51360.2021.9595661
[29]
Wang, Z. and Mao, S. (2024) AIGC for RF Sensing: The Case of RFID-Based Human Activity Recognition. 2024 International Conference on Computing, Networking and Communications (ICNC), Big Island, 19-22 February 2024, 1092-1097. https://doi.org/10.1109/icnc59896.2024.10556013
[30]
Zuo, S., Fortes, V., Suh, S., Sigg, S. and Lukowicz, P. (2023) Unsupervised Diffusion Model for Sensor-Based Human Activity Recognition. Adjunct Proceedings of the 2023 ACM International Joint Conference on Pervasive and Ubiquitous Computing & the 2023 ACM International Symposium on Wearable Computing, Cancun, 8-12 October 2023, 205. https://doi.org/10.1145/3594739.3610797
[31]
Wang, Z. and Mao, S. (2025) AIGC for Wireless Sensing: Diffusion-Empowered Human Activity Sensing. IEEE Transactions on Cognitive Communications and Networking, 11, 657-671. https://doi.org/10.1109/tccn.2025.3525588
[32]
Li, Z., Shi, Y., Gao, Y. and Xu, D. (2026) Diffusion-Based Data Augmentation for Image Recognition: A Systematic Analysis and Evaluation. International Journal of Computer Vision, 134, Article No. 126. https://doi.org/10.1007/s11263-026-02754-x
[33]
Custance, O., Khan, S., Parkinson, S. and Sheng, Q.Z. (2026) Parameter-Efficient Domain Adaption for CSI Crowd-Counting via Self-Supervised Learning with Adapter Modules. arXiv:2601.02203.
[34]
Gu, F., Chung, M., Chignell, M., Valaee, S., Zhou, B. and Liu, X. (2021) A Survey on Deep Learning for Human Activity Recognition. ACM Computing Surveys, 54, 1-34. https://doi.org/10.1145/3472290
[35]
Erak, O. and Abou-Zeid, H. (2023) Accelerating and Compressing Deep Neural Networks for Massive MIMO CSI Feedback. ICC 2023-IEEE International Conference on Communications, Rome, 28 May-1 June 2023, 1029-1035. https://doi.org/10.1109/icc45041.2023.10279462
[36]
Wang, J., Feng, C., Sun, G., Du, H., Niyato, D., Quek, T.Q.S., et al. (2025) Security Enhancement of CSI-Based Wireless Sensing via Generative AI. 2025 IEEE International Workshop on Radio Frequency and Antenna Technologies (iWRF & AT), Shenzhen, 23-26 May 2025, 83-88. https://doi.org/10.1109/iwrfat65352.2025.11103416
[37]
Xiao, C., Chen, S., Zhou, F. and Wu, J. (2022) Self-Supervised Few-Shot Time-Series Segmentation for Activity Recognition. IEEE Transactions on Mobile Computing, 22, 6770-6783. https://doi.org/10.1109/tmc.2022.3199015
[38]
Wang, Z., Yang, C. and Mao, S. (2025) AIGC for Rf-Based Human Activity Sensing. IEEE Internet of Things Journal, 12, 3991-4005. https://doi.org/10.1109/jiot.2024.3482256
[39]
Wang, Z. and Mao, S. (2024) AIGC for Wireless Data: The Case of RFID-Based Human Activity Recognition. ICC 2024-IEEE International Conference on Communications, Denver, 9-13 June 2024, 4060-4065. https://doi.org/10.1109/icc51166.2024.10622401
