Federated learning (FL) is a promising paradigm that enables collaboration among edge devices to train a neural network while preserving data privacy. This paper considers a previously unexamined scenario of over-the-air FL systems with model heterogeneity, where clients with varying computing capacities adopt local models comprising subsets of global parameters, and over-the-air computation is employed to accelerate parameter aggregation. We investigate three key design considerations, namely, subnet creation, client selection, and client composition, and assess their impact on signal distortion resulting from over-the-air computation and on the accuracy of model learning. Our results provide insights into the effects of each design aspect on the FL system’s performance and convergence.


    Access

    Check access

    Check availability in my library

    Order at Subito €


    Export, share and cite



    Title :

    Over-the-Air Federated Learning with Model Heterogeneity: A Comparative Study


    Contributors:


    Publication date :

    2024-10-07


    Size :

    5925293 byte





    Type of media :

    Conference paper


    Type of material :

    Electronic Resource


    Language :

    English



    Federated Representation Learning With Data Heterogeneity for Human Mobility Prediction

    Zhang, Xiao / Wang, Qilin / Ye, Ziming et al. | IEEE | 2023


    Federated Learning Framework Coping with Hierarchical Heterogeneity in Cooperative ITS

    Song, Rui / Zhou, Liguo / Lakshminarasimhan, Venkatnarayanan et al. | IEEE | 2022



    Cooperative Relay Assisted Federated Learning over Fading Channels

    Dong, Zhihao / Zhu, Xu / Cao, Jie et al. | IEEE | 2024


    Deep Learning Based Coded Over-the-Air Computation for Personalized Federated Learning

    Chen, Danni / Lei, Ming / Zhao, Ming-Min et al. | IEEE | 2023