Saved in:
| Main Authors: | , , , , , , |
|---|---|
| Format: | Preprint |
| Published: |
2026
|
| Subjects: | |
| Online Access: | https://arxiv.org/abs/2602.21343 |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| _version_ | 1866914349853966336 |
|---|---|
| author | Feng, Chao Grubl, Thomas von der Assen, Jan Hunkeler, Sandrin Raphael Spitz, Linn Anna Bovet, Gerome Stiller, Burkhard |
| author_facet | Feng, Chao Grubl, Thomas von der Assen, Jan Hunkeler, Sandrin Raphael Spitz, Linn Anna Bovet, Gerome Stiller, Burkhard |
| contents | 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. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2602_21343 |
| institution | arXiv |
| publishDate | 2026 |
| record_format | arxiv |
| spellingShingle | UnlinkableDFL: a Practical Mixnet Protocol for Churn-Tolerant Decentralized FL Model Sharing Feng, Chao Grubl, Thomas von der Assen, Jan Hunkeler, Sandrin Raphael Spitz, Linn Anna Bovet, Gerome Stiller, Burkhard Networking and Internet Architecture 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. |
| title | UnlinkableDFL: a Practical Mixnet Protocol for Churn-Tolerant Decentralized FL Model Sharing |
| topic | Networking and Internet Architecture |
| url | https://arxiv.org/abs/2602.21343 |