In this paper, we propose a reliable federated learning (FL) scheme for vehicular communication networks. The scheme is named intelligent vehicle selection and resource optimization (IVSRO), which aims to improve the federate learning reliability by reducing the probability of incorrect packet transmission in mobility scenario, and determining the most suitable vehicle for learning based on the incorrect packet probability. Specifically, we introduce a FL model for the vehicular communication network and analyze the probability of incorrect packet transmission caused by dynamic channel changes under this network. In consideration of FL convergence accuracy, an optimization problem is formulated to minimize the incorrect packet transmission rate, which is achieved through selecting the optimal connected vehicles from the training set, allocating transmission power and wireless spectrum resources to the selected vehicles. By employing convergence analysis and determining the optimal power for each selected vehicle, the proposed optimization problem can be handled using a bipartite matching algorithm. Simulation results show that the identification accuracy of the proposed IVSRO scheme is higher than that of existing baseline schemes. The results of this study demonstrate how the proposed IVSRO scheme improve the reliability of the FL scheme in vehicular communication networks while considering the varying channel conditions and proper vehicle selection, making it valuable for FL implementations in the domains of intelligent transportation and road safety management.


    Zugriff

    Zugriff prüfen

    Verfügbarkeit in meiner Bibliothek prüfen

    Bestellung bei Subito €


    Exportieren, teilen und zitieren



    Titel :

    Reliable Federated Learning in Vehicular Communication Networks: An Intelligent Vehicle Selection and Resource Optimization Scheme


    Beteiligte:
    Yang, Tongzhou (Autor:in) / Li, Qihao (Autor:in) / Zhang, Ning (Autor:in) / Zhao, Linlin (Autor:in) / Hu, Fengye (Autor:in)


    Erscheinungsdatum :

    24.06.2024


    Format / Umfang :

    698425 byte





    Medientyp :

    Aufsatz (Konferenz)


    Format :

    Elektronische Ressource


    Sprache :

    Englisch



    Client selection and resource scheduling in reliable federated learning for UAV-assisted vehicular networks

    ZHAO, Hongbo / GENG, Liwei / FENG, Wenquan et al. | Elsevier | 2024

    Freier Zugriff


    Federated Learning in Vehicular Networks

    Elbir, Ahmet M. / Soner, Burak / Coleri, Sinem et al. | IEEE | 2022


    Sensing Vehicle Selection Scheme Optimization in Vehicular Crowdsensing

    Yu, Haiyang / Liu, Chenyang / Yang, Yang et al. | ASCE | 2020


    Clustered Vehicular Federated Learning: Process and Optimization

    Taik, Afaf / Mlika, Zoubeir / Cherkaoui, Soumaya | IEEE | 2022