Privacy-preserving and Communication-efficient Convolutional Neural Network Prediction Framework in Mobile Cloud Computing = Privacy-preserving and Communication-efficient Convolutional Neural Network Prediction Framework in Mobile Cloud Computing

다국어 초록 (Multilingual Abstract)

Deep Learning as a Service (DLaaS), utilizing the cloud-based deep neural network models to provide customer prediction services, has been widely deployed on mobile cloud computing (MCC). Such services raise privacy concerns since customers need to send private data to untrusted service providers. In this paper, we devote ourselves to building an efficient protocol to classify users’ images using the convolutional neural network (CNN) model trained and held by the server, while keeping both parties’ data secure. Most previous solutions commonly employ homomorphic encryption schemes based on Ring Learning with Errors (RLWE) hardness or two-party secure computation protocols to achieve it. However, they have limitations on large communication overheads and costs in MCC. To address this issue, we present LeHE4SCNN, a scalable privacy-preserving and communication-efficient framework for CNN-based DLaaS. Firstly, we design a novel low-expansion rate homomorphic encryption scheme with packing and unpacking methods (LeHE). It supports fast homomorphic operations such as vector-matrix multiplication and addition. Then we propose a secure prediction framework for CNN. It employs the LeHE scheme to compute linear layers while exploiting the data shuffling technique to perform non-linear operations. Finally, we implement and evaluate LeHE4SCNN with various CNN models on a real-world dataset. Experimental results demonstrate the effectiveness and superiority of the LeHE4SCNN framework in terms of response time, usage cost, and communication overhead compared to the state-of-the-art methods in the mobile cloud computing environment.

참고문헌 (Reference)

1 M. S. Riazi, "XONN: Xnor-based oblivious deep neural network inference" 1501-1518, 2019

2 J. Fan, "Somewhat practical fully homomorphic encryption" 2012 : 144-, 2012

3 X. Jiang, "Secure outsourced matrix computation and application to neural networks" 1209-1222, 2018

4 E. Hesamifard, "Privacy-preserving machine learning as a service" 2018 (2018): 123-142, 2018

5 J. Liu, "Oblivious neural network predictions via minionn transformations" 619-631, 2017

6 A. Paverd, "Modelling and automatically analysing privacy properties for honest-but-curious adversaries" Univ. Oxford 2014

7 T. H. Noor, "Mobile cloud computing : Challenges and future research directions" 115 : 70-85, 2018

8 N. Fernando, "Mobile cloud computing : A survey" 29 (29): 84-106, 2013

9 M. Ribeiro, "Mlaas: Machine learning as a service" 896-902, 2015

10 C. Ke, "Low expansion rate encryption algorithm based on mlwe" 46 (46): 144-150, 2019

11 J. H. Cheon, "Homomorphic encryption for arithmetic of approximate numbers" 409-437, 2017

12 C. Juvekar, "GAZELLE: A low latency framework for secure neural network inference" 1651-1669, 2018

13 C. Gentry, "Fully homomorphic encryption using ideal lattices" 169-178, 2009

14 N. P. Smart, "Fully homomorphic SIMD operations" 71 (71): 57-81, 2014

15 W. Zheng, "Fighting fire with fire : A spatial–frequency ensemble relation network with generative adversarial learning for adversarial image classification" 36 (36): 2081-2121, 2021

16 E. Chou, "Faster cryptonets : Leveraging sparsity for real-world encrypted inference"

17 S. Li, "Falcon: A fourier transform based approach for fast and secure convolutional neural network predictions" 8705-8714, 2020

18 P. Mishra, "Delphi: A cryptographic inference service for neural networks" 2505-2522, 2020

19 B. D. Rouhani, "Deepsecure: Scalable provably-secure deep learning" 1-6, 2018

20 M. S. Riazi, "Deep learning on private data" 17 (17): 54-63, 2019

21 D. Shen, "Deep learning in medical image analysis" 19 : 221-248, 2017

22 R. Gilad-Bachrach, "Cryptonets: Applying neural networks to encrypted data with high throughput and accuracy" 201-210, 2016

23 M. S. Riazi, "Chameleon: A hybrid secure computation framework for machine learning applications" 707-721, 2018

24 B. Zhou, "Augmentation techniques for mobile cloud computing : A taxonomy, survey, and future directions" 51 (51): 1-38, 2019

25 M. Tavana, "An artificial neural network and bayesian network model for liquidity risk assessment in banking" 275 : 2525-2554, 2018

26 Z. Brakerski, "(Leveled)fully homomorphic encryption without bootstrapping" 6 (6): 1-36, 2014