Secure Decentralized Federated Learning via Gossip and Virtual Voting

2026-07-09

AI Summary

This work addresses the lack of finality guarantees for model provenance and vulnerability to Byzantine or lazy nodes in existing decentralized federated learning systems, as well as the global coordination overhead introduced by ledger-based approaches. To overcome these limitations, the authors propose gspDAG-FL, a framework that constructs a gossip-based topological directed acyclic graph (DAG) and establishes consensus over gossip histories. Finality is defined through model-provenance tuples, eliminating the need for global synchronization. The approach integrates multiple verification layers—including virtual voting, compact event certificates, receiver endorsements, and private semantic auditing—to enhance security and robustness. Experiments on MNIST and Penn Treebank demonstrate that gspDAG-FL maintains high detection rates for invalid provenance under mixed adversarial conditions, achieves performance comparable to verified ledger schemes, and significantly reduces coordination overhead while improving throughput.

Abstract

Decentralized federated learning (DFL) removes the central server by letting nodes exchange model updates through peer-to-peer gossip, but existing gossip-based methods often lack provenance finality and resilience to Byzantine or lazy participants. Ledger-assisted federated learning (FL) improves auditability, yet blockchains, shards, or settlement committees can reintroduce global coordination costs that conflict with DFL locality. This paper proposes \emph{gspDAG-FL}, a secure DFL framework that derives consensus from the same gossip history used to disseminate models. Nodes exchange model payloads only with neighbors, while full nodes collect event certificates and receiver-endorsed accepted gossip proofs, reconstruct a compact Topology directed acyclic graph (DAG), and run Hashgraph-style virtual voting followed by compact full-node certificates. Finality is over unique model-origin tuples, not identical local parameter states. To improve resilience, gspDAG-FL combines payload validation, accepted-proof validation, and private semantic audit before aggregation. We formalize the adversarial setting, prove safety and conditional liveness of the control plane, and give a convergence guarantee for certified perturbed gossip under time-varying effective mixing. Experiments on MNIST classification and Penn Treebank language modeling, using fair held-out validation/audit data and networks up to \(N=100\), show that gspDAG-FL achieves learning quality close to validation-based ledger FL while reducing coordination bottlenecks, improving throughput, and maintaining high invalid-origin detection under mixed Byzantine and lazy participation.

Problem

• Decentralized Federated Learning
• Byzantine resilience
• Gossip protocol
• Finality
• Auditability

Innovation

• gossip-based consensus
• virtual voting
• Topology DAG
• Byzantine resilience
• decentralized federated learning