Search papers, labs, and topics across Lattice.
To resolve discriminator collapse, semantic drift, and link-agnostic decoding in federated semantic communication, the authors design FedGenSC, an architecture that deploys generative adversarial priors across heterogeneous edge channels. The system stabilizes non-IID training by pairing a globally aggregated generator with client-confined discriminators, aligns distributed representations via a semantic prototype bank, and conditions generative decoding on real-time channel state information. Evaluated on the Europarl corpus over simulated Rayleigh fading channels, FedGenSC outperforms FedDeepSC baselines across all tested link conditions, yielding up to a 58.2% relative improvement in BLEU-1 at 18 dB SNR.
Federated GANs for semantic communication suffer catastrophic drift under non-IID data, but keeping discriminators strictly local while conditioning generation on physical SNR yields up to a 58.2% recovery in reconstructed text fidelity.
Integrating generative adversarial networks (GANs) into federated semantic communication (SemCom) is a natural progression, as generative priors can recover semantic fidelity under channel distortion that discriminative decoders cannot. However, naive GAN federation introduces three failure modes that prior work has, to the best of our knowledge, neither identified nor resolved: discriminator aggregation instability under non-independent and identically distributed (non-IID) data, semantic drift caused by divergent local embedding spaces, and channel-agnostic generation that cannot adapt to heterogeneous link conditions. We propose federated generative semantic communication (FedGenSC), which mitigates all three by employing a global generator with local-only discriminators, providing cross-client semantic information through a semantic prototype bank, and conditioning generation on the instantaneous signal-to-noise ratio (SNR). Experiments on the Europarl dataset over Rayleigh fading channels (K=10 clients, Dirichlet 伪=0.5) show that FedGenSC under non-IID data outperforms the FedDeepSC baseline across the tested SNR range, achieving up to a 58.2% relative improvement in bilingual evaluation understudy (BLEU)-1 at 18 dB. Ablation studies confirm the independent contribution of each component.
