29 January 2025 in IEEE Internet of Things Journal
In the emerging Industrial Internet of Things (IIoT) applications, federated learning (FL) enables model training without the need to transmit raw data directly. Nevertheless, transmitting model parameters could still reveal private information. To further protect local model parameters, differential privacy combined with FL (DPFL) has been introduced. Nonetheless, adding noise in DPFL can severely impact model performance, especially in non-independent and identically distributed (non-iid) data scenarios typical of IIoT environments. It is necessary to carefully balance privacy preservation and utility. In this article, we propose a robust DPFL scheme leveraging local model partition and sparsification (namely, FedMPS) for heterogeneous IIoT scenarios. The local model is divided into a shared part, which is sparsified before adding noise to mitigate its impact, and a private part that remains on the client. We provide a theoretical analysis of the privacy guarantees. Extensive experiments on common datasets, including Fashion-MNIST, CIFAR-10, and CIFAR-100, demonstrate that the proposed approach achieves a better privacy-utility tradeoff, with a 10%–20% improvement compared to baseline methods, and performs well especially in non-iid scenarios.
12 September 2025 in IEEE Transactions on Vehicular Technology
With the rapid development of the Internet of Vehicles (IoV), mobile vehicle clients can employ Federated Learning (FL) to collaboratively train an AI model in a distributed manner, thereby unlocking the full potential of IoV. However, due to the high mobility of vehicles, they may move out the coverage of edge server before completing local model training, making it difficult to upload their local models for aggregation on time. Furthermore, lazy vehicles with low-quality data or non-action contribute minimally to the global model while consuming system resources. These factors severely degrade the robustness and accuracy of FL. To mitigate the adverse effects, it's essential for the success of FL systems to select suitable vehicle clients based on the data quality and mobility. In this paper, we propose a mobility-aware reputation-based vehicle client selection scheme (MoReCS), where edge servers in a hierarchical federated learning (HFL) architecture dynamically select suitable vehicle clients based on real-time mobility information and reputation. Experimental results in mobile scenarios demonstrate that MoReCS effectively reduces the impact of vehicle mobility on model training and further enhances accuracy through the integration of a reputation mechanism.
24 August 2025 in Expert Systems with Applications