Understanding Data Reconstruction Leakage in Federated Learning from a Theoretical Perspective
This work addresses a critical privacy vulnerability in federated learning for researchers and practitioners, though it is incremental as it builds on existing attacks.
The authors tackled the lack of theoretical understanding of data reconstruction attacks in federated learning by proposing a framework to bound reconstruction error and compare attack effectiveness, showing that iDLG outperforms DLG on multiple datasets.
Federated learning (FL) is an emerging collaborative learning paradigm that aims to protect data privacy. Unfortunately, recent works show FL algorithms are vulnerable to the serious data reconstruction attacks. However, existing works lack a theoretical foundation on to what extent the devices' data can be reconstructed and the effectiveness of these attacks cannot be compared fairly due to their unstable performance. To address this deficiency, we propose a theoretical framework to understand data reconstruction attacks to FL. Our framework involves bounding the data reconstruction error and an attack's error bound reflects its inherent attack effectiveness. Under the framework, we can theoretically compare the effectiveness of existing attacks. For instance, our results on multiple datasets validate that the iDLG attack inherently outperforms the DLG attack.