|
|
无网络环境下基于联邦学习的移动群智感知框架
|
Abstract:
为解决传统的MCS任务无法在一些偏远或网络不稳定的地区执行，本文提出了一种新型的模型架构，专门针对那些无网络环境下的MCS任务执行。基于链路QoS感知建立节点之间数据传输机制，用于传输联邦学习参数，同时针对节点间性能以及任务执行时间的差异，对联邦聚合算法进行优化。实验结果表明，本文提出的框架与对应机制的组合，不仅提高了任务执行的效率，还保障了数据的安全性和模型的性能。
To tackle the issue that traditional MCS tasks cannot be executed in remote areas with unstable networks, this paper proposes a new framework specifically for executing MCS tasks in environments without network connectivity. A data transmission mechanism based on link QoS awareness is established between nodes for transmitting federated learning parameters, and the federated aggregation algorithm is optimized in consideration of the performance differences between nodes and the time taken to execute tasks. Experimental results show that the framework proposed in this paper, combined with the corresponding mechanisms, not only improves the efficiency of task execution but also ensures the security of data and the performance of the model.
| [1] | Capponi, A., Fiandrino, C., Kantarci, B., et al. (2019) A Survey on Mobile Crowdsensing Systems: Challenges, Solutions, and Opportunities. IEEE Communications Surveys & Tutorials, 21, 2419-2165. https://doi.org/10.1109/COMST.2019.2914030 |
| [2] | Yang, Q. (2019) AI and Data Privacy Protection: The Way to Federated Learning. Journal of Information Security Reserach, 5, 961-965. |
| [3] | Mcmahan, H.B., Moore, E., Ramage, D., et al. (2016) Communication-Efficient Learning of Deep Networks from Decentralized Data. https://ui.adsabs.harvard.edu/abs/2016arXiv160205629M |
| [4] | Wang, H. (2022) A Survey of Application and Key Techniques for Mobile Crowdsensing. Wireless Communications and Mobile Computing, 2022, Article 3693537. https://doi.org/10.1155/2022/3693537 |
| [5] | 万涛, 李婉琦, 葛晶晶. 基于区块链的边缘移动群智感知声誉更新方案[J]. 计算机应用研究, 2023, 40(6): 1636-1640. |
| [6] | Mehanna, S. (2023) Data Quality Issues in Mobile Crowdsensing Environments. Signal and Image Processing. Master’s Thesis, Université Paris-Saclay, Paris-Saclay. |
| [7] | Yang, H., Zhang, X., Khanduri, P., et al. (2022) Anarchic Federated Learning. Proceedings of the 39th International Conference on Machine Learning, Baltimore, 17-23 July 2022, 25331-25363. |
| [8] | Wu, X., Huang, F., Hu, Z., et al. (2023) Faster Adaptive Federated Learning. Proceedings of the AAAI Conference on Artificial Intelligence, 37, 10379-10387. https://doi.org/10.1609/aaai.v37i9.26235 |
| [9] | Zeng, D., Liang, S., Hu, X., et al. (2023) Fedlab: A Flexible Federated Learning Framework. Journal of Machine Learning Research, 24, 1-7. |
| [10] | Banerjee, M., Borges, C., Choo, K.K.R., et al. (2022) A Hardware-Assisted Heartbeat Mechanism for Fault Identification in Large-Scale IoT Systems. IEEE Transactions on Dependable and Secure Computing, 19, 1254-1265. |
| [11] | Li, Q., Diao, Y., Chen, Q., et al. (2021) Federated Learning on Non-IID Data Silos: An Experimental Study. https://ui.adsabs.harvard.edu/abs/2021arXiv210202079L.10.48550/arXiv.2102.02079 |
| [12] | Wang, J., Liu, Q., Liang, H., et al. (2020) Tackling the Objective Inconsistency Problem in Heterogeneous Federated Optimization. https://ui.adsabs.harvard.edu/abs/2020arXiv200707481W.10.48550/arXiv.2007.07481 |
