Local Differential Privacy-Based Federated Learning for Internet of Things

Zhao, Yang; Zhao, Jun; Yang, Mengmeng; Wang, Teng; Wang, Ning; Lyu, Lingjuan; Niyato, Dusit; Lam, Kwok-Yan

doi:10.1109/jiot.2020.3037194

Cited by 204 publications

(90 citation statements)

References 36 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…The learned model contains valid information to be further used in tasks such as resource allocation without revealing their raw data, which further enhances privacy protection. Similarly, to avoid the privacy threat to vehicular IoT networks, a new approach is suggested in [102] using FL combined with local differential privacy which aims for perturbing gradients generated by vehicles while not compromising the utility of gradients. This would prevent attackers from deducing original data even though they obtain perturbed gradients.…”

Section: F Fl-based Techniques For Privacy and Security In Iot Services And Networkmentioning

confidence: 99%

Federated Learning for Internet of Things: A Comprehensive Survey

Nguyen

Ding

Pathirana

et al. 2021

IEEE Commun. Surv. Tutorials

447

107

View full text Add to dashboard Cite

Section: F Fl-based Techniques For Privacy and Security In Iot Services And Networkmentioning

confidence: 99%

Federated Learning for Internet of Things: A Comprehensive Survey

Nguyen

Ding

Pathirana

et al. 2021

IEEE Commun. Surv. Tutorials

447

107

View full text Add to dashboard Cite

“…Among them, multiple works focus on the integration of DP and federated learning. However, most of them (e.g., [8], [9], [10], [11], [12], [13], [14], [15], [16], [17]) demonstrate the performance of proposed approaches merely by experiments, with no theoretical analysis on the convergence. However, experimental observations are not always reliable since the performance of machine learning algorithms heavily rely on hyper-parameter tuning, and an algorithm that is observed to perform better than other algorithms may just be the algorithm that is better tuned.…”

Section: Related Workmentioning

confidence: 99%

“…In order to motivate and retain edge devices in federated learning, it is desirable to provide rigorous differential privacy (DP) guarantee for devices. While there have been multiple works focusing on the integration of DP and federated learning [8], [9], [10], [11], [12], [13], [14], [15], [16], [17], most of the work demonstrate the performance of proposed approaches by experiments, whose results heavily rely on hyper-parameter tuning. The main focus of this paper is to provide a differentially-private federated learning approach with convergence performance bound.…”

Section: Introductionmentioning

confidence: 99%

Concentrated Differentially Private Federated Learning With Performance Analysis

Guo

2021

IEEE Open J. Comput. Soc.

View full text Add to dashboard Cite

Federated learning engages a set of edge devices to collaboratively train a common model without sharing their local data and has advantage in user privacy over traditional cloud-based learning approaches. However, recent model inversion attacks and membership inference attacks have demonstrated that shared model updates during the interactive training process could still leak sensitive user information. Thus, it is desirable to provide rigorous differential privacy (DP) guarantee in federated learning. The main challenge to providing DP is to maintain high utility of federated learning model with repeatedly introduced randomness of DP mechanisms, especially when the server is not fully trusted. In this paper, we investigate how to provide DP to the most widely adopted federated learning scheme, federated averaging. Our approach combines local gradient perturbation, secure aggregation, and zero-concentrated differential privacy (zCDP) for better utility and privacy protection without a trusted server. We jointly consider the performance impacts of randomness introduced by DP mechanisms, client sampling and data subsampling in our approach, and theoretically analyze the convergence rates and end-to-end DP guarantee with non-convex loss functions. We also demonstrate that our proposed method has good utility-privacy trade-off through extensive numerical experiments on a real-world dataset.

show abstract

“…As a promising branch of IoT, the Internet of Vehicles (IoV) has stimulated the development of vehicular crowdsourcing applications, which also results in unexpected privacy threats on vehicle users. Zhao et al [ 176 ] adopted both LDP and FL models to avoid sensitive information leakage in IoV applications. Besides, the work in [ 39 ] has provided a detailed summary of the applications of LDP in the Internet of connected vehicles.…”

Section: Applicationsmentioning

confidence: 99%

A Comprehensive Survey on Local Differential Privacy toward Data Statistics and Analysis

Wang

Zhang

Feng

et al. 2020

Sensors

Self Cite

View full text Add to dashboard Cite

Collecting and analyzing massive data generated from smart devices have become increasingly pervasive in crowdsensing, which are the building blocks for data-driven decision-making. However, extensive statistics and analysis of such data will seriously threaten the privacy of participating users. Local differential privacy (LDP) was proposed as an excellent and prevalent privacy model with distributed architecture, which can provide strong privacy guarantees for each user while collecting and analyzing data. LDP ensures that each user’s data is locally perturbed first in the client-side and then sent to the server-side, thereby protecting data from privacy leaks on both the client-side and server-side. This survey presents a comprehensive and systematic overview of LDP with respect to privacy models, research tasks, enabling mechanisms, and various applications. Specifically, we first provide a theoretical summarization of LDP, including the LDP model, the variants of LDP, and the basic framework of LDP algorithms. Then, we investigate and compare the diverse LDP mechanisms for various data statistics and analysis tasks from the perspectives of frequency estimation, mean estimation, and machine learning. Furthermore, we also summarize practical LDP-based application scenarios. Finally, we outline several future research directions under LDP.

show abstract

Local Differential Privacy-Based Federated Learning for Internet of Things

Cited by 204 publications

References 36 publications

Federated Learning for Internet of Things: A Comprehensive Survey

Federated Learning for Internet of Things: A Comprehensive Survey

Concentrated Differentially Private Federated Learning With Performance Analysis

A Comprehensive Survey on Local Differential Privacy toward Data Statistics and Analysis

Contact Info

Product

Resources

About