UnlinkableDFL: a Practical Mixnet Protocol for Churn-Tolerant Decentralized FL Model Sharing

TL;DR

UnlinkableDFL introduces a mixnet-based decentralized federated learning protocol that enhances participant unlinkability without significantly affecting learning convergence or resource efficiency.

Contribution

It presents a novel mixnet-based framework for fully decentralized FL that ensures unlinkability while maintaining comparable convergence and moderate resource use.

Findings

UnlinkableDFL achieves reliable convergence similar to FedAvg.

Communication latency is the main overhead introduced.

System maintains moderate memory and CPU usage.

Abstract

Decentralized Federated Learning (DFL) eliminates the need for a central aggregator, but it can expose communication patterns that reveal participant identities. This work presents UnlinkableDFL, a DFL framework that combines a peer-based mixnet with fragment-based model aggregation to ensure unlinkability in fully decentralized settings. Model updates are divided into encrypted fragments, sent over separate multi-hop paths, and aggregated without using any identity information. A theoretical analysis indicates that relay and end-to-end unlinkability improve with larger mixing sets and longer paths, while convergence remains similar to standard FedAvg. A prototype implementation evaluates learning performance, latency, unlinkability, and resource usage. The results show that UnlinkableDFL converges reliably and adapts to node churn. Communication latency emerges as the main overhead, while memory and CPU usage stay moderate. These findings illustrate the balance between anonymity and system efficiency, demonstrating that strong unlinkability can be maintained in decentralized learning workflows.