Received: July 21, 2026
Accepted: August 14, 2026
Publication Date: August 22, 2026
Privacy-Preserving Federated Training End-to-End Flow
Copyright The Author(s). This is an open access article distributed under the terms of the Creative Commons Attribution License (CC BY 4.0), which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are cited.
Download Citation: BibTeX | http://dx.doi.org/10.6180/jase.202611_34.064
Wearable and edge-aware systems for sports motion recognition increasingly collect multi-source data (posture, acceleration, angular velocity) plus smart-equipment sensor signals (vibration, impact, temperature, wear). Centralized training exposes exercise habits and usage patterns while imposing high communication burden. This study proposes a scalable, privacy-preserving federated learning framework integrating motion recognition with equipment monitoring under non-IID distributions: edge devices preprocess and segment data via sliding window, train a lightweight CNN-BiGRU-Attention model locally, then prune, DP-perturb, and compress gradients before the server securely aggregates and redistributes the model to selected clients. Under non-IID conditions, FedAvg reaches 89.7% accuracy, FedProx 91.1%, and the proposed method 94.2%. Communication cost falls from 5.60 GB to 2.66 GB after 140 rounds (52.4% below FedAvg), and member-inference/gradient inversion attack success rates fall to 15%/12%.
Keywords: sports motion recognition, federated learning, smart sports equipment monitoring, wearable sensors, differential privacy, secure aggregation, communication compression
- [1] S. Zhang, Y. Li, S. Zhang, F. Shahabi, S. Xia, Y. Deng, and N. Alshurafa, (2022) “Deep learning in human activity recognition with wearable sensors: A review on advances” Sensors 22(4): 1476. DOI: 10.3390/s22041476.
- [2] D. Navakauskas and M. Dumpis, (2025) “Wearable sensor-based human activity recognition: Performance and interpretability of dynamic neural networks” Sensors 25(14): 4420. DOI: 10.3390/s25144420.
- [3] Q. Jia, J. Guo, P. Yang, and Y. Yang, (2024) “A holistic multi-source transfer learning approach using wearable sensors for personalized daily activity recognition” Complex & Intelligent Systems 10: 1459–1471. DOI: 10.1007/s40747-023-01218-w.
- [4] J. Liu, W. Zhu, D. Li, X. Hu, and L. Song, (2025) “Domain generalization with semi-supervised learning for people-centric activity recognition” Science China Information Sciences 68: 112103. DOI: 10.1007/s11432-022-3860-y.
- [5] Q. Shen, H. Feng, R. Song, D. Song, and H. Xu, (2023) “Federated meta-learning with attention for diversity-aware human activity recognition” Sensors 23(3): 1083. DOI: 10.3390/s23031083.
- [6] I. Iwan, B. N. Yahya, and S.-L. Lee, (2025) “Federated model with contrastive learning and adaptive control variates for human activity recognition” Frontiers of Information Technology & Electronic Engineering 26(6): 896–911. DOI: 10.1631/FITEE.2400797.
- [7] W. Guo, F. Zhuang, X. Zhang, Y. Tong, and J. Dong, (2024) “A comprehensive survey of federated transfer learning: Challenges, methods and applications” Frontiers of Computer Science 18: 186356. DOI: 10.1007/s11704-024-40065-x.
- [8] V. Mothukuri, R. M. Parizi, S. Pouriyeh, Y. Huang, A. Dehghantanha, and G. Srivastava, (2021) “A survey on security and privacy of federated learning” Future Generation Computer Systems 115: 619–640. DOI: 10.1016/j.future.2020.10.007.
- [9] Z. Chen, H. Zheng, and G. Liu, (2024) “AWDP-FL: An adaptive differential privacy federated learning framework” Electronics 13(19): 3959. DOI: 10.3390/electronics13193959.
- [10] H. K. Tayyeh and A. S. A. Al-Jumaili, (2024) “Balancing privacy and performance: A differential privacy approach in federated learning” Computers 13(11): 277. DOI: 10.3390/computers13110277.
- [11] N. Bouacida and P. Mohapatra, (2021) “Vulnerabilities in federated learning” IEEE Access 9: 63229–63247. DOI: 10.1109/ACCESS.2021.3075203.
- [12] K. Pillutla, S. M. Kakade, and Z. Harchaoui, (2022) “Robust aggregation for federated learning” IEEE Transactions on Signal Processing 70: 1142–1154. DOI: 10.1109/TSP.2022.3153135.
- [13] S. Zhou, L. Wang, L. Chen, Y. Wang, and K. Yuan, (2025) “Group verifiable secure aggregate federated learning based on secret sharing” Scientific Reports 15: 9712. DOI: 10.1038/s41598-025-94478-0.
- [14] T.-M. H. Hsu, H. Qi, and M. Brown, (2019) “Measuring the effects of non-identical data distribution for federated visual classification” arXiv: DOI: 10.48550/arXiv.1909.06335.
- [15] C. Yang and Y. Ma, (2026) “Secure aggregation for heterogeneous enterprise data based on federated meta-learning” Scientific Reports 16: 4484. DOI: 10.1038/s41598-025-34735-4.