Citation

Chowdhury, Md Tahmid Ashraf and Ullah, Fasee and Labonno, Shanjida Islam and Kamal, Shahid and Islam, Mohammad Ahsanul (2026) RollupFL: An Auditable Federated Learning Framework for Byzantine Client Accountability. International Journal of Advanced Computer Science and Applications, 17 (3). ISSN 2158107X

Text Paper_99-RollupFL_An_Auditable_Federated_Learning_Framework.pdf - Published Version Restricted to Repository staff only Download (2MB)

Abstract

Federated learning (FL) trains a shared model without sending raw data, but some clients can be Byzantine and send harmful updates. Robust aggregation methods like Median and Krum can reduce poisoning damage, but they do not clearly show which client attacked. In this study, we propose RollupFL, an audit layer for FL that improves accountability under Byzantine attacks. RollupFL keeps aggregation and auditing separate, so it can work with FedAvg, Median, or Krum without changing how aggregation is computed. We study two audit designs: simple logging, which is fast, but assumes a trusted server, and blockchain-based audit, which gives stronger integrity and attribution, but adds more latency. We evaluate MNIST training for 20 rounds with 10%–30% Byzantine clients under sign-flip and model-replacement attacks. Results show that auditing does not meaningfully change accuracy, but it improves accountability. At 30% Byzantine, blockchain audit achieves higher attribution (0.95) and tamper detection (0.92) than logging (0.65 and 0.58). Logging adds small per-round latency, while blockchain adds larger latency mainly due to ledger writing.

Item Type:	Article
Uncontrolled Keywords:	Federated learning, Byzantine attacks, audit layer, accountability, attacker attribution
Subjects:	Q Science > QA Mathematics > QA273-280 Probabilities. Mathematical statistics
Divisions:	Faculty of Computing and Informatics (FCI)
Depositing User:	Ms Rosnani Abd Wahab
Date Deposited:	05 Jun 2026 07:00
Last Modified:	05 Jun 2026 07:00
URII:	http://shdl.mmu.edu.my/id/eprint/16049

Downloads

Edit (login required)