Linqi Yin

Decentralized Federated Learning for Heterogeneous Multi-Task Semantic Communication

A personalized DSC framework that cuts off cross-task interference by limiting the system to absorbing complementary knowledge while actively blocking mismatched parameter updates, and which achieves a 4.77% global relative improvement over the no-aggregation baseline and outperforms decentralized FedAvg, FedAMP, and h...