MIGAN: A Privacy Leakage Evaluation Scheme for CIoT-Based Federated Learning Users

Federated Learning (FL) in Consumer Internet of Things (CIoT) encounters significant privacy security threats when collaborative training Machine Learning models using data distributed among numerous smart CIoT devices. This paper’s objective is to evaluate the potential privacy risks associated with FL within CIoT context, specifically through the lens of passive malicious users. We propose an evaluation scheme (MIGAN) for privacy leakage, which incorporates various privacy inference attacks. First, this paper analyzes the characteristics of privacy inference attacks in FL and summarizes privacy security threats to guide the design of the MIGAN evaluation scheme. Then, to fully utilize the privacy information in FL, this paper proposes faster model inversion and white-box membership inference. Integrating these two modules into the MIGAN framework facilitates fast and high-quality privacy data reconstruction. Finally, this paper constructs the evaluation function to measure the correlation between compromised privacy data and users’ training data. Experiments conducted on the MNIST and CelebA datasets reveal that MIGAN’s data privacy reconstruction outperforms comparable methods. Notably, MIGAN quantifies the likelihood of potential privacy leakage at 14.7% for MNIST and 8.3% for CelebA in FL with 30 clients. These results validates the practicality and relevance of MIGAN in evaluating privacy leakage risks.

I. Introduction

The Consumer Internet of Things (CIoT) and Machine Learning (ML) have advanced quickly [1]. This progress allows numerous smart CIoT devices to connect and create ML models that offer personalized AI services [2]. However, this deployment raises significant security and privacy concerns. These concerns are mainly from the fact that ML models typically use sensitive user data during training [3], [4]. To address these concerns, Federated Learning (FL) [5] emerges as a promising solution. FL facilitates collaborative training of ML models by multiple users in a decentralized manner, eliminating the need to share personal data. Through FL, the data generated by consumer electronics remains confined in users’ local environment [6]. Although FL provides basic privacy protection [7], the frequent transmission model weights by CIoT devices [8] amplifies the vulnerability of these models, potentially leading to the exploitation of privacy information. Xiong et al. [9] and Barroso et al. [10] indicated that FL is exposed to various privacy inference attacks for compromising user sensitive data. This revelation highlights the necessity for constant awareness and enhancement in privacy protection measures within CIoT.