1: \begin{abstract}

2: Federated learning enables distributed training of machine learning (ML) models across multiple devices in a privacy-preserving manner. Hierarchical federated learning (HFL) is further proposed to meet the requirements of both latency and coverage. 

3: % Some research has been done to deal with the issue of data heterogeneity in HFL, but few consider the impact of mobility, which is the nature of vehicles. Although the impact of mobility on communication is usually negative, there are chances to leverage mobility to reduce the impact of data heterogeneity in HFL. 

4: In this paper, we consider a data-heterogeneous HFL scenario with mobility, mainly targeting vehicular networks. We derive the convergence upper bound of HFL with respect to mobility and data heterogeneity, and analyze how mobility impacts the performance of HFL. While mobility is considered as a challenge from a communication point of view, our goal here is to exploit mobility to improve the learning performance by mitigating data heterogeneity. Simulation results verify the analysis and show that mobility can indeed improve the model accuracy by up to 15.1\% when training a convolutional neural network on the CIFAR-10 dataset using HFL. 

5: % As far as we know, this is the first paper to utilize mobility to enhance the performance of HFL.

6: \end{abstract}

7: