Buffer is All You Need: Defending Federated Learning against Backdoor Attacks under Non-iids via Buffering
Fuente:
arXiv
Salvato in:
| Autori principali: | , , , , , , |
|---|---|
| Natura: | Preprint |
| Pubblicazione: |
2025
|
| Soggetti: | |
| Accesso online: | |
| Tags: |
Aggiungi Tag
Nessun Tag, puoi essere il primo ad aggiungerne!!
|
| _version_ | 1866908290456223744 |
|---|---|
| author | Lyu, Xingyu Wang, Ning Xiao, Yang Li, Shixiong Li, Tao Chen, Danjue Chen, Yimin |
| author_facet | Lyu, Xingyu Wang, Ning Xiao, Yang Li, Shixiong Li, Tao Chen, Danjue Chen, Yimin |
| contents | Federated Learning (FL) is a popular paradigm enabling clients to jointly train a global model without sharing raw data. However, FL is known to be vulnerable towards backdoor attacks due to its distributed nature. As participants, attackers can upload model updates that effectively compromise FL. What's worse, existing defenses are mostly designed under independent-and-identically-distributed (iid) settings, hence neglecting the fundamental non-iid characteristic of FL. Here we propose FLBuff for tackling backdoor attacks even under non-iids. The main challenge for such defenses is that non-iids bring benign and malicious updates closer, hence harder to separate. FLBuff is inspired by our insight that non-iids can be modeled as omni-directional expansion in representation space while backdoor attacks as uni-directional. This leads to the key design of FLBuff, i.e., a supervised-contrastive-learning model extracting penultimate-layer representations to create a large in-between buffer layer. Comprehensive evaluations demonstrate that FLBuff consistently outperforms state-of-the-art defenses. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2503_23511 |
| institution | arXiv |
| publishDate | 2025 |
| record_format | arxiv |
| spellingShingle | Buffer is All You Need: Defending Federated Learning against Backdoor Attacks under Non-iids via Buffering Lyu, Xingyu Wang, Ning Xiao, Yang Li, Shixiong Li, Tao Chen, Danjue Chen, Yimin Cryptography and Security Artificial Intelligence Federated Learning (FL) is a popular paradigm enabling clients to jointly train a global model without sharing raw data. However, FL is known to be vulnerable towards backdoor attacks due to its distributed nature. As participants, attackers can upload model updates that effectively compromise FL. What's worse, existing defenses are mostly designed under independent-and-identically-distributed (iid) settings, hence neglecting the fundamental non-iid characteristic of FL. Here we propose FLBuff for tackling backdoor attacks even under non-iids. The main challenge for such defenses is that non-iids bring benign and malicious updates closer, hence harder to separate. FLBuff is inspired by our insight that non-iids can be modeled as omni-directional expansion in representation space while backdoor attacks as uni-directional. This leads to the key design of FLBuff, i.e., a supervised-contrastive-learning model extracting penultimate-layer representations to create a large in-between buffer layer. Comprehensive evaluations demonstrate that FLBuff consistently outperforms state-of-the-art defenses. |
| title | Buffer is All You Need: Defending Federated Learning against Backdoor Attacks under Non-iids via Buffering |
| topic | Cryptography and Security Artificial Intelligence |
| url | https://arxiv.org/abs/2503.23511 |