FedNIA: Noise-Induced Activation Analysis for Mitigating Data Poisoning in FL
Fuente:
arXiv
Saved in:
| Main Authors: | , , |
|---|---|
| Format: | Preprint |
| Published: |
2025
|
| Subjects: | |
| Online Access: | |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| _version_ | 1866913703325073408 |
|---|---|
| author | Hallaji, Ehsan Razavi-Far, Roozbeh Saif, Mehrdad |
| author_facet | Hallaji, Ehsan Razavi-Far, Roozbeh Saif, Mehrdad |
| contents | Federated learning systems are increasingly threatened by data poisoning attacks, where malicious clients compromise global models by contributing tampered updates. Existing defenses often rely on impractical assumptions, such as access to a central test dataset, or fail to generalize across diverse attack types, particularly those involving multiple malicious clients working collaboratively. To address this, we propose Federated Noise-Induced Activation Analysis (FedNIA), a novel defense framework to identify and exclude adversarial clients without relying on any central test dataset. FedNIA injects random noise inputs to analyze the layerwise activation patterns in client models leveraging an autoencoder that detects abnormal behaviors indicative of data poisoning. FedNIA can defend against diverse attack types, including sample poisoning, label flipping, and backdoors, even in scenarios with multiple attacking nodes. Experimental results on non-iid federated datasets demonstrate its effectiveness and robustness, underscoring its potential as a foundational approach for enhancing the security of federated learning systems. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2502_16396 |
| institution | arXiv |
| publishDate | 2025 |
| record_format | arxiv |
| spellingShingle | FedNIA: Noise-Induced Activation Analysis for Mitigating Data Poisoning in FL Hallaji, Ehsan Razavi-Far, Roozbeh Saif, Mehrdad Machine Learning Artificial Intelligence Cryptography and Security Federated learning systems are increasingly threatened by data poisoning attacks, where malicious clients compromise global models by contributing tampered updates. Existing defenses often rely on impractical assumptions, such as access to a central test dataset, or fail to generalize across diverse attack types, particularly those involving multiple malicious clients working collaboratively. To address this, we propose Federated Noise-Induced Activation Analysis (FedNIA), a novel defense framework to identify and exclude adversarial clients without relying on any central test dataset. FedNIA injects random noise inputs to analyze the layerwise activation patterns in client models leveraging an autoencoder that detects abnormal behaviors indicative of data poisoning. FedNIA can defend against diverse attack types, including sample poisoning, label flipping, and backdoors, even in scenarios with multiple attacking nodes. Experimental results on non-iid federated datasets demonstrate its effectiveness and robustness, underscoring its potential as a foundational approach for enhancing the security of federated learning systems. |
| title | FedNIA: Noise-Induced Activation Analysis for Mitigating Data Poisoning in FL |
| topic | Machine Learning Artificial Intelligence Cryptography and Security |
| url | https://arxiv.org/abs/2502.16396 |