Abstract
Federated Learning (FL) is a decentralized machine learning approach where client devices train models locally and send them to a server that performs aggregation to generate a global model. FL is vulnerable to model inversion attacks, where the server can infer sensitive client data from trained models. Google's Secure Aggregation (SecAgg) protocol addresses this data privacy issue by masking each client's trained model using shared secrets and individual elements generated locally on the client's device. Although SecAgg effectively preserves privacy, it imposes considerable communication and computation overhead, especially as network size increases. Building upon SecAgg, this poster introduces a Communication-Efficient Secure Aggregation (CESA) protocol that substantially reduces this overhead by using only two shared secrets per client to mask the model. We propose our method for stable networks with low delay variation and limited client dropouts. CESA is independent of the data distribution and network size (for higher than 6 nodes), preventing the honest-but-curious server from accessing unmasked models. Our initial evaluation reveals that CESA significantly reduces the communication cost compared to SecAgg.
| Original language | English |
|---|---|
| Title of host publication | 2024 54th Annual IEEE/IFIP International Conference on Dependable Systems and Networks - Supplemental Volume, DSN-S 2024 |
| Publisher | IEEE |
| Pages | 157-158 |
| Number of pages | 2 |
| ISBN (Electronic) | 9798350395709 |
| DOIs | |
| Publication status | Published - 29 Aug 2024 |
| Externally published | Yes |
| Event | 54th Annual IEEE/IFIP International Conference on Dependable Systems and Networks, DSN 2024 - Brisbane, Australia Duration: 24 Jun 2024 → 27 Jun 2024 |
Conference
| Conference | 54th Annual IEEE/IFIP International Conference on Dependable Systems and Networks, DSN 2024 |
|---|---|
| Abbreviated title | DSN 2024 |
| Country/Territory | Australia |
| City | Brisbane |
| Period | 24/06/24 → 27/06/24 |
Keywords
- n/a OA procedure
- Federated Learning (FL)
- Secure Aggregation (SecAgg)
- Communication Efficiency
Fingerprint
Dive into the research topics of 'Boosting Communication Efficiency of Federated Learning's Secure Aggregation'. Together they form a unique fingerprint.Research output
- 2 Citations
- 1 Preprint
-
Boosting Communication Efficiency of Federated Learning's Secure Aggregation
Nazemi, N., Tavallaie, O., Chen, S., Zomaya, A. Y. & Holz, R., 2 May 2024, ArXiv.org, 2 p.Research output: Working paper › Preprint › Academic
Open AccessFile1 Downloads (Pure)
Cite this
- APA
- Author
- BIBTEX
- Harvard
- Standard
- RIS
- Vancouver