Efficiency Optimization Techniques in Privacy-Preserving Federated Learning With Homomorphic Encryption: A Brief Survey¶
Authors: Qipeng Xie; Siyang Jiang; Linshan Jiang; Yongzhi Huang; Zhihe Zhao; Salabat Khan; Wangchen Dai; Zhe Liu; Kaishun Wu
Year: 2024
Venue: IEEE Internet of Things Journal
type: survey collection: ppml status: auto
Links: DOI
Summary¶
Auto-extracted summary (unreviewed)
This brief survey reviews techniques for reducing the efficiency overhead of homomorphic encryption when it is used to protect model updates in privacy-preserving federated learning. It covers general HE optimization strategies and their limits in the FL setting, then organizes algorithmic, hardware and hybrid optimizations into a taxonomy and suggests future research directions for HE-based PPFL.
Abstract¶
Federated learning (FL) offers distributed machine learning on edge devices. However, the FL model raises privacy concerns. Various techniques, such as homomorphic encryption (HE), differential privacy, and multiparty cooperation, are used to address the privacy issues of the FL model. Among them, HE ensures greater security and privacy since end-to-end encryption maintains data privacy throughout the computation process. Compared with other privacy-preserving techniques, HE does not require the establishment of a trusted environment or protocol among multiple parties and does not involve any artificial noise that can impair system performance. Unfortunately, it suffers from efficiency overhead when applied to privacy-preserving FL (PPFL). Some existing surveys on PPFL discuss the generic construction and organization of PPFL from the perspective of practical HE deployment in PPFL. However, none of them covers the efficiency optimization of HE when applied to PPFL. This article conducts a comprehensive review of the efficiency optimization of HE when applied to PPFL. First, we review general optimization strategies and discuss their limitations when Manuscript received 25 October 2023; revised 27 February 2024; accepted 20 March 2024. Date of current version 9 July 2024. This work was supported in part by China, NSFC under Grant U2001207; in part by the Guangdong Provincial Key Lab of Integrated Communication, Sensing and Computation for Ubiquitous Internet of Things under Grant 2023B1212010007; in part by the Project of DEGP under Grant 2023KCXTD042; in part by the Zhejiang Lab Open Research Project under Grant K2022PDOAB01; in part by the Foundation for Distinguished Young Talents in Higher Education of Guangdong Province, China, under Grant 2022KQNCX084; and in part by the Fund which aims to improve scientific research capability of key construction disciplines in Guangdong province “Light-Weight Federal Learning Paradigm and Its Application” under Grant 2022ZDJS058. (Corresponding authors: Salabat Khan; Kaishun Wu.) Qipeng Xie is with the Hong Kong University of Science and Technology (Guangzhou), Guangzhou, China, also with The Hong Kong University of Science and Technology, 511453 Hong Kong, and also with the Zhejiang Laboratory, Hangzhou 311121, China (e-mail: qxieaf@connect.ust.hk). Siyang Jiang is with the Department of Information Engineering, The Chinese University of Hong Kong, Hong Kong, SAR, and also with the Zhejiang Laboratory, Hangzhou 311121, China (e-mail: syjiang@ie.cuhk.edu.hk). Linshan Jiang is with the Institute of Data Science, National University Singapore, Singapore 119077 (e-mail: linshan@nus.edu.sg). Yongzhi Huang and Kaishun Wu are with the Hong Kong University of Science and Technology (Guangzhou), Guangzhou, China (e-mail: huangyongzhi@email.szu.edu.cn; wuks@hkust-gz.edu.cn). Zhihe Zhao is with the Department of Information Engineering, The Chinese University of Hong Kong, Hong Kong, SAR. Salabat Khan is with the School
Source: pdf
Extracted details¶
| Field | Value | Source | Confidence |
|---|---|---|---|
| contribution | review of HE efficiency optimization for privacy-preserving federated learning | llm | 0.70 |
| contribution | discussion of general HE optimization strategies and their limitations when applied to PPFL | llm | 0.70 |
| contribution | taxonomy of algorithmic, hardware and hybrid optimizations | llm | 0.70 |
| contribution | future research directions for HE-based PPFL | llm | 0.70 |
| schemes | HE | llm | 0.70 |
| tasks | federated learning | llm | 0.70 |
| llm relevance | low | llm | 0.70 |
Keyword mentions¶
Counts of vocabulary terms in the full text (first page of occurrence in brackets).
- scheme: Paillier 13 [p.2], MPC 11 [p.2], CKKS 6 [p.2], Differential privacy 6 [p.1], Partially homomorphic encryption 6 [p.3], Secret sharing 5 [p.2], TFHE 4 [p.4], BFV 3 [p.2], BGV 3 [p.2], FHEW 2 [p.4], Multi-key HE 2 [p.8], Leveled / somewhat HE 2 [p.8]
- model: Logistic regression 11 [p.8], Gradient boosting / XGBoost 8 [p.4], Decision tree 1 [p.8], Autoencoder 1 [p.7], GAN 1 [p.10]
- hardware: FPGA 16 [p.8], GPU 16 [p.8], AVX-512 2 [p.9], CUDA 1 [p.12]
- technique: Federated learning 52 [p.1], Quantization 14 [p.5], Semi-honest model 5 [p.5], SIMD packing 4 [p.4], Bootstrapping 3 [p.11], NTT 3 [p.9], Softmax 2 [p.8], Sigmoid 2 [p.8], Secure aggregation 2 [p.9], Malicious model 2 [p.1], Membership inference 2 [p.1]
- assumption: LWE 4 [p.2], RLWE 2 [p.4], Post-quantum 2 [p.2]
- task: IoT / edge 11 [p.1], Medical / healthcare 10 [p.1]
- topic: Homomorphic encryption 156 [p.1], Privacy-preserving machine learning 2 [p.2]
Files¶
| Path (under papers/) | Collection | Pages | Size | SHA-256 | Note |
|---|---|---|---|---|---|
references/ppml/federated-learning/xie-2024_efficiency-optimization-techniques-in-privacy-preserving-fl-with-he.pdf |
ppml/federated-learning | 12 | 1.7 MB | d6033b575d93 |
was Xie et al. - 2024 - Efficiency Optimization Techniques in Privacy-Preserving Federated Learning With Homomorphic Encrypt.pdf |
Database key: efficiency-optimization-techniques-in-privacy-preserving-federated · created 2026-11-05 · updated 2026-11-05 · edit with python3 tools/paperdb.py set efficiency-optimization-techniques-in-privacy-preserving-federated COLUMN VALUE