On the Security of Verifiable and Oblivious Secure Aggregation for Privacy-Preserving Federated Learning

被引：2

作者：

Wu, Jiahui ^{[1
]}

Zhang, Weizhe ^{[1
,2
]}

机构：

[1] Peng Cheng Lab, New Network Dept, Shenzhen 518000, Peoples R China

[2] Harbin Inst Technol, Fac Comp, Sch Cyberspace Sci, Shenzhen 518055, Peoples R China

来源：

IEEE TRANSACTIONS ON DEPENDABLE AND SECURE COMPUTING | 2024年 / 21卷 / 05期

基金：

中国国家自然科学基金;

关键词：

Gold; Forgery; Aggregates; Resists; Protocols; Federated learning; Encryption; verifiability; privacy protection; secure aggregation;

D O I：

10.1109/TDSC.2024.3352170

中图分类号：

TP3 [计算技术、计算机技术];

学科分类号：

0812 ;

摘要：

Recently, to resist privacy leakage and aggregation result forgery in federated learning (FL), Wang et al. proposed a verifiable and oblivious secure aggregation protocol for FL, called VOSA. They claimed that VOSA was aggregate unforgeable and verifiable under a malicious aggregation server and gave detailed security proof. In this article, we show that VOSA is insecure, in which local gradients/aggregation results and their corresponding authentication tags/proofs can be tampered with without being detected by the verifiers. After presenting specific attacks, we analyze the reason for this security issue and give a suggestion to prevent it.

引用

页码：4324 / 4326

页数：3