Byzantine Attack

Byzantine attacks target distributed systems, such as federated learning and decentralized consensus algorithms, by introducing malicious or faulty nodes that disrupt the computation or communication process. Current research focuses on developing robust aggregation rules and algorithms, often employing techniques like geometric median, trimmed mean, and median-of-means, to mitigate the impact of these attacks, even under data heterogeneity and non-IID settings. Understanding and addressing Byzantine robustness is crucial for ensuring the reliability and security of distributed machine learning and other decentralized applications, impacting the trustworthiness of models trained in these environments. The development of Byzantine-resilient algorithms is a significant area of ongoing research, with a focus on improving both theoretical guarantees and practical performance.

Papers

November 15, 2024

Embedding Byzantine Fault Tolerance into Federated Learning via Virtual Data-Driven Consistency Scoring Plugin
Youngjoon Lee, Jinu Gong, Joonhyuk Kang
Byzantine Attack Byzantine Resilience Byzantine Fault Tolerance Image Analysis Validation Consistency Metric

October 30, 2024

Byzantine-Robust Federated Learning: An Overview With Focus on Developing Sybil-based Attacks to Backdoor Augmented Secure Aggregation Protocols
Atharv Deshmukh
Native Robustness Secure Aggregation Byzantine Robust Byzantine Attack Population Protocol Byzantine Robust Federated Learning

October 17, 2024

Byzantine-Resilient Output Optimization of Multiagent via Self-Triggered Hybrid Detection Approach
Chenhang Yan, Liping Yan, Yuezu Lv, Bolei Dong, Yuanqing Xia
Detection Method Multiagent System Byzantine Attack Heterogeneous Multi Agent Event Triggered Control Resilient Algorithm

October 16, 2024

FedCAP: Robust Federated Learning via Customized Aggregation and Personalization
Youpeng Li, Xinda Wang, Fuxun Yu, Lichao Sun, Wenbin Zhang, Xuyu Wang
Byzantine Attack Consistent Aggregation

September 19, 2024

On the Hardness of Decentralized Multi-Agent Policy Evaluation under Byzantine Attacks
Hairi, Minghong Fang, Zifan Zhang, Alvaro Velasquez, Jia Liu
Hardness Result Byzantine Attack Multi Agent Policy Cooperative Multi Agent System Byzantine Agent

August 12, 2024

Understanding Byzantine Robustness in Federated Learning with A Black-box Server
Fangyuan Zhao, Yuexiang Xie, Xuebin Ren, Bolin Ding, Shusen Yang, Yaliang Li
Black Box Byzantine Robust Byzantine Attack

July 11, 2024

Generalization Error Matters in Decentralized Learning Under Byzantine Attacks
Haoxiang Ye, Qing Ling
Generalization Error Decentralized Learning Byzantine Attack Byzantine Resilience Byzantine Agent

May 6, 2024

Byzantine-Robust Gossip: Insights from a Dual Approach
Renaud Gaucher, Hadrien Hendrikx, Aymeric Dieuleveut
DCU Insight AQ Convergence Guarantee Decentralized FL Decentralized Optimization Gossip Algorithm Byzantine Attack Dual Approach

April 12, 2024

Anti-Byzantine Attacks Enabled Vehicle Selection for Asynchronous Federated Learning in Vehicular Edge Computing
Cui Zhang, Xiao Xu, Qiong Wu, Pingyi Fan, Qiang Fan, Huiling Zhu, Jiangzhou Wang
Deep Reinforcement Learning Vehicle Plan Asynchronous Federated Learning Byzantine Attack Vehicle Data Global Aggregation Vehicular Edge Computing Vehicle Edge Computing

April 9, 2024

Aggressive or Imperceptible, or Both: Network Pruning Assisted Hybrid Byzantines in Federated Learning
Emre Ozfatura, Kerem Ozfatura, Alptekin Kupcu, Deniz Gunduz
Network Pruning Byzantine Attack Imperceptible Pattern Imperceptible Attack Byzantine Failure

March 20, 2024

Byzantine-resilient Federated Learning With Adaptivity to Data Heterogeneity
Shiyuan Zuo, Xingrun Yan, Rongfei Fan, Han Hu, Hangguan Shan, Tony Q. S. Quek
Adaptive Importance Data Heterogeneity Non Convex Loss Function Byzantine Attack Byzantine Robust Federated Learning Convex Loss Function

March 19, 2024

Resilience in Online Federated Learning: Mitigating Model-Poisoning Attacks via Partial Sharing
Ehsan Lari, Reza Arablouei, Vinay Chakravarthi Gogineni, Stefan Werner
Global Impact Tiny Refinement Elicit Resilience Byzantine Attack Model Poisoning Attack Link Sharing Practice

February 25, 2024

Honeybee: Decentralized Peer Sampling with Verifiable Random Walks for Blockchain Data Sharding
Yunqi Zhang, Shaileshh Bojja Venkatakrishnan
Peer Agent Byzantine Attack Decentralized Algorithm Historical Data Verification Tool Table Sharding Blockchain Sharding

February 6, 2024

Towards Fair, Robust and Efficient Client Contribution Evaluation in Federated Learning
Meiying Zhang, Huan Zhao, Sheldon Ebron, Kan Yang
Robust Version Client Contribution Byzantine Attack Byzantine Robust Aggregation Client Centered Assessment

November 13, 2023

AGRAMPLIFIER: Defending Federated Learning Against Poisoning Attacks Through Local Update Amplification
Zirui Gong, Liyue Shen, Yanjun Zhang, Leo Yu Zhang, Jingwei Wang, Guangdong Bai, Yong Xiang
Poisoning Attack Byzantine Robust Byzantine Attack Byzantine Robust Aggregation Local Update

November 8, 2023

Byzantine-Tolerant Methods for Distributed Variational Inequalities
Nazarii Tupitsa, Abdulla Jasem Almansoori, Yanlin Wu, Martin Takáč, Karthik Nandakumar, Samuel Horváth, Eduard Gorbunov
Byzantine Robust Variational Inequality Byzantine Attack Byzantine Robust Method

September 11, 2023

SABLE: Secure And Byzantine robust LEarning
Antoine Choffrut, Rachid Guerraoui, Rafael Pinot, Renaud Sirdey, John Stephan, Martin Zuber
Homomorphic Encryption Byzantine Robust Byzantine Attack Efficient Homomorphic

September 7, 2023

Byzantine-Robust Federated Learning with Variance Reduction and Differential Privacy
Zikai Zhang, Rui Hu
Differential Privacy Privacy Guarantee Variance Reduction Privacy Analysis Byzantine Attack Byzantine Robust Federated Learning Client Level Differential Privacy

August 21, 2023

Federated Learning Robust to Byzantine Attacks: Achieving Zero Optimality Gap
Shiyuan Zuo, Rongfei Fan, Han Hu, Ning Zhang, Shimin Gong
Byzantine Attack Linear Convergence Robust Aggregation Optimality Gap Robust Federated Learning

July 25, 2023

High Dimensional Distributed Gradient Descent with Arbitrary Number of Byzantine Attackers
Puning Zhao, Zhiguo Wan
Gradient Descent Structured Output High Dimension Mean Estimation Byzantine Attack High Dimensional Problem Byzantine Failure Gradient Vector

Byzantine Attack

Papers

Embedding Byzantine Fault Tolerance into Federated Learning via Virtual Data-Driven Consistency Scoring Plugin

Byzantine-Robust Federated Learning: An Overview With Focus on Developing Sybil-based Attacks to Backdoor Augmented Secure Aggregation Protocols

Byzantine-Resilient Output Optimization of Multiagent via Self-Triggered Hybrid Detection Approach

FedCAP: Robust Federated Learning via Customized Aggregation and Personalization

On the Hardness of Decentralized Multi-Agent Policy Evaluation under Byzantine Attacks

Understanding Byzantine Robustness in Federated Learning with A Black-box Server

Generalization Error Matters in Decentralized Learning Under Byzantine Attacks

Byzantine-Robust Gossip: Insights from a Dual Approach

Anti-Byzantine Attacks Enabled Vehicle Selection for Asynchronous Federated Learning in Vehicular Edge Computing

Aggressive or Imperceptible, or Both: Network Pruning Assisted Hybrid Byzantines in Federated Learning

Byzantine-resilient Federated Learning With Adaptivity to Data Heterogeneity

Resilience in Online Federated Learning: Mitigating Model-Poisoning Attacks via Partial Sharing

Honeybee: Decentralized Peer Sampling with Verifiable Random Walks for Blockchain Data Sharding

Towards Fair, Robust and Efficient Client Contribution Evaluation in Federated Learning

AGRAMPLIFIER: Defending Federated Learning Against Poisoning Attacks Through Local Update Amplification

Byzantine-Tolerant Methods for Distributed Variational Inequalities

SABLE: Secure And Byzantine robust LEarning

Byzantine-Robust Federated Learning with Variance Reduction and Differential Privacy

Federated Learning Robust to Byzantine Attacks: Achieving Zero Optimality Gap

High Dimensional Distributed Gradient Descent with Arbitrary Number of Byzantine Attackers