This review paper takes a comprehensive look at malicious attacks against FL, categorizing them from new perspectives on attack origins and targets, and providing insights into their methodology and impact. In this survey, we focus on threat models targeting the learning process of FL systems. Based on the source and target of the attack, we categorize existing threat models into four types, Data to Model (D2M), Model to Data (M2D), Model to Model (M2M) and composite attacks. For each attack type, we discuss the defense strategies proposed, highlighting their effectiveness, assumptions and potential areas for improvement. Defense strategies have evolved from using a singular metric to excluding malicious clients, to employing a multifaceted approach examining client models at various phases. In this survey paper, our research indicates that the to-learn data, the learning gradients, and the learned model at different stages all can be manipulated to initiate malicious attacks that range from undermining model performance, reconstructing private local data, and to inserting backdoors. We have also seen these threat are becoming more insidious. While earlier studies typically amplified malicious gradients, recent endeavors subtly alter the least significant weights in local models to bypass defense measures. This literature review provides a holistic understanding of the current FL threat landscape and highlights the importance of developing robust, efficient, and privacy-preserving defenses to ensure the safe and trusted adoption of FL in real-world applications.
Federated Learning (FL) is a widely adopted privacy-preserving machine learning approach where private data remains local, enabling secure computations and the exchange of local model gradients between local clients and third-party parameter servers. However, recent findings reveal that privacy may be compromised and sensitive information potentially recovered from shared gradients. In this study, we offer detailed analysis and a novel perspective on understanding the gradient leakage problem. These theoretical works lead to a new gradient leakage defense technique that secures arbitrary model architectures using a private key-lock module. Only the locked gradient is transmitted to the parameter server for global model aggregation. Our proposed learning method is resistant to gradient leakage attacks, and the key-lock module is designed and trained to ensure that, without the private information of the key-lock module: a) reconstructing private training data from the shared gradient is infeasible; and b) the global model's inference performance is significantly compromised. We discuss the theoretical underpinnings of why gradients can leak private information and provide theoretical proof of our method's effectiveness. We conducted extensive empirical evaluations with a total of forty-four models on several popular benchmarks, demonstrating the robustness of our proposed approach in both maintaining model performance and defending against gradient leakage attacks.
Data privacy has become an increasingly important issue in machine learning. Many approaches have been developed to tackle this issue, e.g., cryptography (Homomorphic Encryption, Differential Privacy, etc.) and collaborative training (Secure Multi-Party Computation, Distributed Learning and Federated Learning). These techniques have a particular focus on data encryption or secure local computation. They transfer the intermediate information to the third-party to compute the final result. Gradient exchanging is commonly considered to be a secure way of training a robust model collaboratively in deep learning. However, recent researches have demonstrated that sensitive information can be recovered from the shared gradient. Generative Adversarial Networks (GAN), in particular, have shown to be effective in recovering those information. However, GAN based techniques require additional information, such as class labels which are generally unavailable for privacy persevered learning. In this paper, we show that, in Federated Learning (FL) system, image-based privacy data can be easily recovered in full from the shared gradient only via our proposed Generative Regression Neural Network (GRNN). We formulate the attack to be a regression problem and optimise two branches of the generative model by minimising the distance between gradients. We evaluate our method on several image classification tasks. The results illustrate that our proposed GRNN outperforms state-of-the-art methods with better stability, stronger robustness, and higher accuracy. It also has no convergence requirement to the global FL model. Moreover, we demonstrate information leakage using face re-identification. Some defense strategies are also discussed in this work.
Typical machine learning approaches require centralized data for model training, which may not be possible where restrictions on data sharing are in place due to, for instance, privacy protection. The recently proposed Federated Learning (FL) frame-work allows learning a shared model collaboratively without data being centralized or data sharing among data owners. However, we show in this paper that the generalization ability of the joint model is poor on Non-Independent and Non-Identically Dis-tributed (Non-IID) data, particularly when the Federated Averaging (FedAvg) strategy is used in this collaborative learning framework thanks to the weight divergence phenomenon. We propose a novel boosting algorithm for FL to address this generalisation issue, as well as achieving much faster convergence in gradient based optimization. We demonstrate our Federated Boosting (FedBoost) method on privacy-preserved text recognition, which shows significant improvements in both performance and efficiency. The text images are based on publicly available datasets for fair comparison and we intend to make our implementation public to ensure reproducibility.