Skip to content

Latest commit

 

History

History
27 lines (19 loc) · 653 Bytes

README.md

File metadata and controls

27 lines (19 loc) · 653 Bytes

SecFedNIDS

Cite

Zhao Z, Yong Z, Da G, Lei Y, Zhao L,SecFedNIDS: Robust defense for poisoning attack against federated learning-based network intrusion detection system, Future Generation Computer Systems,Volume 134,2022,Pages 154-169,https://doi.org/10.1016/j.future.2022.04.010.

https://authors.elsevier.com/c/1ezXj,3q5xgP6x

Requirements

  • Pytorch
  • pyod
  • ...

Running Experiments

Attack model: Launch the label-flipping attack against FL-based NIDS on UNSW-NB15 dataset

Conduct the poisoned model detection:

python main_poisoned_model_det.py 

Conduct the Poisoned data detection:

python main_poisoned_data_det.py