Logo
ScienceStack
Table of Contents
Fetching ...
Sign In

Federated Learning with Quantum Computing and Fully Homomorphic Encryption: A Novel Computing Paradigm Shift in Privacy-Preserving ML

Siddhant Dutta, Pavana P Karanth, Pedro Maciel Xavier, Iago Leal de Freitas, Nouhaila Innan, Sadok Ben Yahia, Muhammad Shafique, David E. Bernal Neira

TL;DR

This work instantiates the FHE scheme by applying the FHE scheme to a Federated Learning Neural Network architecture that integrates both classical and quantum layers.

Abstract

The widespread deployment of products powered by machine learning models is raising concerns around data privacy and information security worldwide. To address this issue, Federated Learning was first proposed as a privacy-preserving alternative to conventional methods that allow multiple learning clients to share model knowledge without disclosing private data. A complementary approach known as Fully Homomorphic Encryption (FHE) is a quantum-safe cryptographic system that enables operations to be performed on encrypted weights. However, implementing mechanisms such as these in practice often comes with significant computational overhead and can expose potential security threats. Novel computing paradigms, such as analog, quantum, and specialized digital hardware, present opportunities for implementing privacy-preserving machine learning systems while enhancing security and mitigating performance loss. This work instantiates these ideas by applying the FHE scheme to a Federated Learning Neural Network architecture that integrates both classical and quantum layers.

Federated Learning with Quantum Computing and Fully Homomorphic Encryption: A Novel Computing Paradigm Shift in Privacy-Preserving ML

TL;DR

This work instantiates the FHE scheme by applying the FHE scheme to a Federated Learning Neural Network architecture that integrates both classical and quantum layers.

Abstract

The widespread deployment of products powered by machine learning models is raising concerns around data privacy and information security worldwide. To address this issue, Federated Learning was first proposed as a privacy-preserving alternative to conventional methods that allow multiple learning clients to share model knowledge without disclosing private data. A complementary approach known as Fully Homomorphic Encryption (FHE) is a quantum-safe cryptographic system that enables operations to be performed on encrypted weights. However, implementing mechanisms such as these in practice often comes with significant computational overhead and can expose potential security threats. Novel computing paradigms, such as analog, quantum, and specialized digital hardware, present opportunities for implementing privacy-preserving machine learning systems while enhancing security and mitigating performance loss. This work instantiates these ideas by applying the FHE scheme to a Federated Learning Neural Network architecture that integrates both classical and quantum layers.
Paper Structure (8 sections, 2 figures, 1 table, 1 algorithm)

This paper contains 8 sections, 2 figures, 1 table, 1 algorithm.

Table of Contents

  1. Introduction
  2. Related Work
  3. Problem Description and Methodology
  4. Quantum Neural Network (QNN) Initialization and Client-Side Training
  5. Server-Side Aggregation and Global Model Update
  6. Computational Results
  7. Discussion
  8. Perspectives

Figures (2)

  • Figure 1: The client nodes utilize local data to train a model that incorporates both classical and quantum layers. After training, these models are encrypted and transmitted to a central server for aggregation into the average of all encrypted models. This new global model is then distributed to all clients. This global model is decrypted on each client, initiating another round of training. Intruders to the client-server communication would only intercept of quantum-safe encrypted models.
  • Figure 2: Example of quantum circuit used at the client's level for a QFL approach. Input data is encoded into a quantum state using angle embedding via parameterized rotation gates $R_X(\theta_i)$. The encoded quantum states are then processed by a PQC, where the weights and parameters of the PQC are encrypted using FHE. The encrypted quantum states undergo operations involving parameterized rotation gates and Controlled-NOT gates, facilitating entanglement and complex quantum state manipulation in the encrypted domain. After measurement, classical outputs are obtained, and the Cross-Entropy loss is computed. The encrypted parameters $\theta_i$ are updated during training using a classical optimizer.