To accommodate constantly changing road conditions, real-time model training is essential for autonomous driving (AD). Federated learning (FL) serves as a promising paradigm to enable autonomous vehicles to train models collaboratively with their onboard computing resources. However, existing vehicle selection schemes for FL all assume predetermined and location-independent vehicles' datasets, neglecting the fact that vehicles collect training data along their routes, thereby resulting in suboptimal vehicle selection. To improve the perception quality in AD for a region, we propose Sense4FL, a vehicular crowdsensing-enhanced FL framework featuring trajectory-dependent vehicular training data collection. To this end, we first derive the convergence bound of FL by considering the impact of both vehicles' uncertain trajectories and uploading probabilities, from which we discover that minimizing the training loss is equivalent to minimizing a weighted sum of local and global earth mover's distance (EMD) between vehicles' collected data distribution and global data distribution. Based on this observation, we formulate the trajectory-dependent vehicle selection and data collection problem for FL in AD. Given that the problem is NP-hard, we develop an efficient algorithm to find the solution with an approximation guarantee. Extensive simulation results have demonstrated the effectiveness of our approach in improving object detection performance compared with existing benchmarks.


    Access

    Download


    Export, share and cite



    Title :

    Sense4FL: Vehicular Crowdsensing Enhanced Federated Learning for Autonomous Driving


    Contributors:
    Ma, Yanan (author) / Hu, Senkang (author) / Fang, Zhengru (author) / Ji, Yun (author) / Deng, Yiqin (author) / Fang, Yuguang (author)

    Publication date :

    2025


    Remarks:

    16 pages, 5 figures


    Type of media :

    Preprint


    Type of material :

    Electronic Resource


    Language :

    English




    Large Scale Active Vehicular Crowdsensing

    Zhu, Xiru / Samadh, Shabir Abdul / Yu, Tzu-Yang | IEEE | 2018


    Environmental Monitoring via Vehicular Crowdsensing

    Morselli, Flavio / Zabini, Flavio / Conti, Andrea | IEEE | 2018


    Dynamic parking maps from vehicular crowdsensing

    Bock, Fabian / Gottfried Wilhelm Leibniz Universität Hannover | TIBKAT | 2018


    Dynamic parking maps from vehicular crowdsensing

    Bock, Fabian / Gottfried Wilhelm Leibniz Universität Hannover | GWLB - Gottfried Wilhelm Leibniz Bibliothek | 2018


    Reinforcement Learning Based Advertising Strategy Using Crowdsensing Vehicular Data

    Lou, Kaihao / Yang, Yongjian / Wang, En et al. | IEEE | 2021