In the heterogeneous vehicular networks (HetVNets), the base stations (BSs) can exploit the massive amounts of valuable data collected by vehicles to complete federated learning tasks. However, most of the existing studies consider the scenario of one task requester (TR) and ignore the fact that multiple TRs may concurrently generate their model training requests in the HetVNets. In this paper, we consider the scenario of multi-TR and multi-BS and propose a digital twin enabled scheme for multitask federated learning to address the two-way selection problem between the TRs and the BSs. We first analyze the diversified requirements of the TRs in the HetVNets. Then, we develop a novel model that jointly considers the available training data, the declared price, and the training experience to evaluate the differentiated training capabilities of the BSs. After that, based on the requirements of the TRs and the training capabilities of the BSs, the two-way selection problem between the TRs and the BSs is formulated as a matching game in the digital twin networks, where a matching algorithm is designed to obtain their optimal strategies. The simulation results demonstrate that the proposed scheme can obtain the highest model accuracy and bring the highest utility to the TRs compared with the conventional schemes.


    Access

    Check access

    Check availability in my library

    Order at Subito €


    Export, share and cite



    Title :

    Digital Twin Enabled Multi-task Federated Learning in Heterogeneous Vehicular Networks


    Contributors:
    Hui, Yilong (author) / Zhao, Gaosheng (author) / Yin, Zhisheng (author) / Cheng, Nan (author) / Luan, Tom H. (author)


    Publication date :

    2022-06-01


    Size :

    473222 byte





    Type of media :

    Conference paper


    Type of material :

    Electronic Resource


    Language :

    English




    Federated Learning for Anomaly Detection in Vehicular Networks

    Tham, Chen-Khong / Yang, Lu / Khanna, Akshit et al. | IEEE | 2023



    Federated Learning Enabled Credit Priority Task Processing for Transportation Big Data

    Wu, Guangjun / Li, Jun / Ning, Zhaolong et al. | IEEE | 2024


    Blockchain-Enabled Federated Learning for Enhanced Collaborative Intrusion Detection in Vehicular Edge Computing

    Abou El Houda, Zakaria / Moudoud, Hajar / Brik, Bouziane et al. | IEEE | 2024