Tuanfa Qin

Tuanfa Qin

Auto-generated profile, based on publicly available information

Tuanfa Qin

School of Computer and Electronic Information, Guangxi University, China

Research Focus & Metrics

Representative Publications

View All Publications
6 December 2023
Verified
Conference Paper
Open access
ZCzhu chaoyang
zhu chaoyang
XZXiao Zhu
Xiao Zhu
Zhenrong ZhangZhenrong Zhang
Zhenrong Zhang
TQTuanfa Qin
Tuanfa Qin
The UAV (unmanned aerial vehicles) network is a promising solution to support real-time applications and services. However, local UAV nodes are prone to privacy disclosure, and base station edge nodes are difficult to effectively identify trusted local UAV nodes and resist malicious attacks. To address the privacy concern, we propose the adaptive nonlinear privacy protection to preserve model gradient in the federated learning training process, overcoming the drawbacks associated with empirically selecting DP(Differential privacy) parameters. Through the model accuracy analysis, it can be seen that the proposed algorithm achieves higher accuracy in privacy protection using adaptive layered contribution factor compared with the privacy protection method based on DP.Experiments show that the proposed algorithm can effectively protect the privacy and security of UAV gradient, and achieve higher accuracy. Under the conditions of 10%, 20% and 30% pollution ratio of data sets, the experiment shows that the proposed algorithm still has strong anti-data poisoning attack ability.