Edge Centric and Federated Deep Learning for Privacy Preserving Intelligent Systems
Download PDFAbstract
The proliferation of Internet of Things (IoT) devices and the increasing demand for intelligent applications have led to the rise of edge computing, a paradigm that brings computation and data storage closer to the sources of data. This chapter explores the integration of edge computing with federated learning (FL) to create privacy-preserving intelligent systems. Federated learning, a distributed machine learning approach, enables model training on decentralized data without compromising user privacy. We delve into the foundational concepts of edge computing and federated learning, highlighting the inherent privacy challenges in traditional centralized learning models. The chapter presents a comprehensive literature review of existing privacy-preserving techniques, such as differential privacy and secure aggregation, and their application in federated learning frameworks. We propose a novel methodology for implementing a privacy-centric federated learning system on the edge, detailing the system architecture, the federated learning process, and the integration of privacy-enhancing technologies. To validate our proposed methodology, we conduct extensive simulations using a synthetic dataset, demonstrating the effectiveness of our approach in balancing model accuracy and privacy. The results and discussions section provides a detailed analysis of the simulation outcomes, including the impact of different privacy settings on model performance. Finally, the chapter concludes with a summary of our key findings, contributions, and a discussion of future research directions in this rapidly evolving field.
References
- Coleman Bazelon and Paroma Sanyal. "How Much Licensed Spectrum is Needed to Meet Future Demands for Network Capacity?" In: White paper, The Brattle Group, April 17 (2023).
- Weisong Shi et al. "Edge computing: Vision and challenges". In: IEEE internet of things journal 3.5 (2016), pp. 637–646.
- Protection Regulation. "Regulation (EU) 2016/679 of the European Parliament and of the Council". In: Regulation (eu) 679.2016 (2016), pp. 10–3.
- Elizabeth Liz Harding et al. "Understanding the scope and impact of the california consumer privacy act of 2018". In: vol. 2. 3. Henry Stewart Publications, 2019, pp. 234–253.
- Jakub Konečnỳ et al. "Federated learning: Strategies for improving communication efficiency". In: arXiv preprint arXiv:1610.05492 (2016).
- Karthik Kumar and Yung-Hsiang Lu. "Cloud computing for mobile users: Can offloading computation save energy?" In: Computer 43.4 (2010), pp. 51–56.
- Blesson Varghese et al. "A survey on edge performance benchmarking". In: ACM Computing Surveys (CSUR) 54.3 (2021), pp. 1–33.
- Ji Wang et al. "Not just privacy: Improving performance of private deep learning in mobile cloud". In: Proceedings of the 24th ACM SIGKDD international conference on knowledge discovery & data mining. 2018, pp. 2407–2416.
- Brendan McMahan et al. "Communication-efficient learning of deep networks from decentralized data". In: Artificial intelligence and statistics. Pmlr. 2017, pp. 1273–1282.
- Tian Li et al. "Federated optimization in heterogeneous networks". In: Proceedings of Machine learning and systems 2 (2020), pp. 429–450.
- Keith Bonawitz et al. "Practical secure aggregation for privacy-preserving machine learning". In: proceedings of the 2017 ACM SIGSAC Conference on Computer and Communications Security. 2017, pp. 1175–1191.
- Yoshinori Aono et al. "Privacy-preserving deep learning via additively homomorphic encryption". In: IEEE transactions on information forensics and security 13.5 (2017), pp. 1333–1345.
- Cynthia Dwork. "Differential privacy: A survey of results". In: International conference on theory and applications of models of computation. Springer. 2008, pp. 1–19.
Deep Learning: Foundations, Advances, and Intelligent Applications
