Logo
ScienceStack
Table of Contents
Fetching ...
Sign In

Privacy-Preserving Hierarchical Model-Distributed Inference

Fatemeh Jafarian Dehkordi, Yasaman Keshtkarjahromi, Hulya Seferoglu

TL;DR

PrivateMDI significantly reduces the ML inference time as compared to the baselines and uses model-distributed inference at the edge servers and reduces the amount of communication to/from the cloud server.

Abstract

This paper focuses on designing a privacy-preserving Machine Learning (ML) inference protocol for a hierarchical setup, where clients own/generate data, model owners (cloud servers) have a pre-trained ML model, and edge servers perform ML inference on clients' data using the cloud server's ML model. Our goal is to speed up ML inference while providing privacy to both data and the ML model. Our approach (i) uses model-distributed inference (model parallelization) at the edge servers and (ii) reduces the amount of communication to/from the cloud server. Our privacy-preserving hierarchical model-distributed inference, privateMDI design uses additive secret sharing and linearly homomorphic encryption to handle linear calculations in the ML inference, and garbled circuit and a novel three-party oblivious transfer are used to handle non-linear functions. privateMDI consists of offline and online phases. We designed these phases in a way that most of the data exchange is done in the offline phase while the communication overhead of the online phase is reduced. In particular, there is no communication to/from the cloud server in the online phase, and the amount of communication between the client and edge servers is minimized. The experimental results demonstrate that privateMDI significantly reduces the ML inference time as compared to the baselines.

Privacy-Preserving Hierarchical Model-Distributed Inference

TL;DR

PrivateMDI significantly reduces the ML inference time as compared to the baselines and uses model-distributed inference at the edge servers and reduces the amount of communication to/from the cloud server.

Abstract

This paper focuses on designing a privacy-preserving Machine Learning (ML) inference protocol for a hierarchical setup, where clients own/generate data, model owners (cloud servers) have a pre-trained ML model, and edge servers perform ML inference on clients' data using the cloud server's ML model. Our goal is to speed up ML inference while providing privacy to both data and the ML model. Our approach (i) uses model-distributed inference (model parallelization) at the edge servers and (ii) reduces the amount of communication to/from the cloud server. Our privacy-preserving hierarchical model-distributed inference, privateMDI design uses additive secret sharing and linearly homomorphic encryption to handle linear calculations in the ML inference, and garbled circuit and a novel three-party oblivious transfer are used to handle non-linear functions. privateMDI consists of offline and online phases. We designed these phases in a way that most of the data exchange is done in the offline phase while the communication overhead of the online phase is reduced. In particular, there is no communication to/from the cloud server in the online phase, and the amount of communication between the client and edge servers is minimized. The experimental results demonstrate that privateMDI significantly reduces the ML inference time as compared to the baselines.
Paper Structure (11 sections, 1 theorem, 1 equation, 5 figures, 3 tables, 4 algorithms)

This paper contains 11 sections, 1 theorem, 1 equation, 5 figures, 3 tables, 4 algorithms.

Table of Contents

  1. Introduction
  2. Related Work
  3. System Model & Preliminaries
  4. Privacy-Preserving Model-Distributed Inference (privateMDI) Design
  5. Offline privateMDI
  6. Online privateMDI
  7. Privacy and Communication Analysis
  8. Privacy
  9. Communication Overhead
  10. Evaluation
  11. Conclusion

Key Result

Theorem 1

privateMDI is secure according to Definition def:privacy_protocol assuming the use of secure garbled circuits, linearly homomorphic encryption, and three-party OT.

Figures (5)

  • Figure 1: Hierarchical ML inference.
  • Figure 2: Model-distributed inference.
  • Figure 3: Our novel three-party OT protocol.
  • Figure 4: ML inference time of privateMDI in online and offline phases with increasing number of clusters.
  • Figure 5: ML inference time of privateMDI in online and offline phases with increasing number of clusters.

Theorems & Definitions (3)

  • Definition 1
  • Theorem 1
  • proof