MANDERA: Malicious Node Detection in Federated Learning via Ranking
This addresses a critical security issue in federated learning for applications like healthcare or finance, though it appears incremental as it builds on known gradient distribution differences.
The paper tackles the problem of detecting malicious nodes in federated learning under Byzantine attacks by proposing MANDERA, a method that transforms gradients into a ranking matrix to separate benign and malicious gradients, achieving efficient detection with theoretical guarantees and outperforming state-of-the-art defenses in experiments on IID and Non-IID datasets.
Byzantine attacks hinder the deployment of federated learning algorithms. Although we know that the benign gradients and Byzantine attacked gradients are distributed differently, to detect the malicious gradients is challenging due to (1) the gradient is high-dimensional and each dimension has its unique distribution and (2) the benign gradients and the attacked gradients are always mixed (two-sample test methods cannot apply directly). To address the above, for the first time, we propose MANDERA which is theoretically guaranteed to efficiently detect all malicious gradients under Byzantine attacks with no prior knowledge or history about the number of attacked nodes. More specifically, we transfer the original updating gradient space into a ranking matrix. By such an operation, the scales of different dimensions of the gradients in the ranking space become identical. The high-dimensional benign gradients and the malicious gradients can be easily separated. The effectiveness of MANDERA is further confirmed by experimentation on four Byzantine attack implementations (Gaussian, Zero Gradient, Sign Flipping, Shifted Mean), comparing with state-of-the-art defenses. The experiments cover both IID and Non-IID datasets.