Communication-Efficient Cluster Federated Learning in Large-scale Peer-to-Peer Networks. (arXiv:2204.03843v1 [cs.CR])

A traditional federated learning (FL) allows clients to collaboratively train
a global model under the coordination of a central server, which sparks great
interests in exploiting the private data distributed on clients. However, once
the central server suffers from a single point of failure, it will lead to
system crash. In addition, FL usually involves a large number of clients, which
requires expensive communication costs. These challenges inspire a
communication-efficient design of decentralized FL. In this paper, we propose
an …

cluster communication large networks peer-to-peer scale