Logo
ScienceStack
Table of Contents
Fetching ...
Sign In

Communication-Efficient Quantum Federated Learning over Large-Scale Wireless Networks

Shaba Shaon, Christopher G. Brinton, Dinh C. Nguyen

TL;DR

This study develops a novel sum-rate maximization problem within a muti-channel QFL framework, specifically designed for non-orthogonal multiple access (NOMA)-based large-scale wireless networks, and presents the first theoretical exploration of QFL convergence properties under full device participation.

Abstract

Quantum federated learning (QFL) combines the robust data processing of quantum computing with the privacy-preserving features of federated learning (FL). However, in large-scale wireless networks, optimizing sum-rate is crucial for unlocking the true potential of QFL, facilitating effective model sharing and aggregation as devices compete for limited bandwidth amid dynamic channel conditions and fluctuating power resources. This paper studies a novel sum-rate maximization problem within a muti-channel QFL framework, specifically designed for non-orthogonal multiple access (NOMA)-based large-scale wireless networks. We develop a sum-rate maximization problem by jointly considering quantum device's channel selection and transmit power. Our formulated problem is a non-convex, mixed-integer nonlinear programming (MINLP) challenge that remains non-deterministic polynomial time (NP)-hard even with specified channel selection parameters. The complexity of the problem motivates us to create an effective iterative optimization approach that utilizes the sophisticated quantum approximate optimization algorithm (QAOA) to derive high-quality approximate solutions. Additionally, our study presents the first theoretical exploration of QFL convergence properties under full device participation, rigorously analyzing real-world scenarios with nonconvex loss functions, diverse data distributions, and the effects of quantum shot noise. Extensive simulation results indicate that our multi-channel NOMA-based QFL framework enhances model training and convergence behavior, surpassing conventional algorithms in terms of accuracy and loss. Moreover, our quantum-centric joint optimization approach achieves more than a 100% increase in sum-rate while ensuring rapid convergence, significantly outperforming the state-of-the-arts.

Communication-Efficient Quantum Federated Learning over Large-Scale Wireless Networks

TL;DR

This study develops a novel sum-rate maximization problem within a muti-channel QFL framework, specifically designed for non-orthogonal multiple access (NOMA)-based large-scale wireless networks, and presents the first theoretical exploration of QFL convergence properties under full device participation.

Abstract

Quantum federated learning (QFL) combines the robust data processing of quantum computing with the privacy-preserving features of federated learning (FL). However, in large-scale wireless networks, optimizing sum-rate is crucial for unlocking the true potential of QFL, facilitating effective model sharing and aggregation as devices compete for limited bandwidth amid dynamic channel conditions and fluctuating power resources. This paper studies a novel sum-rate maximization problem within a muti-channel QFL framework, specifically designed for non-orthogonal multiple access (NOMA)-based large-scale wireless networks. We develop a sum-rate maximization problem by jointly considering quantum device's channel selection and transmit power. Our formulated problem is a non-convex, mixed-integer nonlinear programming (MINLP) challenge that remains non-deterministic polynomial time (NP)-hard even with specified channel selection parameters. The complexity of the problem motivates us to create an effective iterative optimization approach that utilizes the sophisticated quantum approximate optimization algorithm (QAOA) to derive high-quality approximate solutions. Additionally, our study presents the first theoretical exploration of QFL convergence properties under full device participation, rigorously analyzing real-world scenarios with nonconvex loss functions, diverse data distributions, and the effects of quantum shot noise. Extensive simulation results indicate that our multi-channel NOMA-based QFL framework enhances model training and convergence behavior, surpassing conventional algorithms in terms of accuracy and loss. Moreover, our quantum-centric joint optimization approach achieves more than a 100% increase in sum-rate while ensuring rapid convergence, significantly outperforming the state-of-the-arts.
Paper Structure (33 sections, 4 theorems, 88 equations, 8 figures, 2 tables, 3 algorithms)

This paper contains 33 sections, 4 theorems, 88 equations, 8 figures, 2 tables, 3 algorithms.

Table of Contents

  1. Introduction
  2. Related Works
  3. Key Contributions
  4. System Model
  5. QFL Learning Model
  6. Network Architecture
  7. Channel Model
  8. Frame Structure and Signal Model
  9. Analysis of Convergence Behavior of QFL
  10. Notation and Definition
  11. Assumptions
  12. Convergence Rates
  13. Complexity Analysis
  14. Sum-rate Maximization Problem Formulation
  15. Problem Statement
  16. ...and 18 more sections

Key Result

Lemma 1

Let Assumption Assumption1 hold, the expected inner product between stochastic gradient and full gradient is bounded by

Figures (8)

  • Figure 1: Proposed QFL framework where distributed quantum devices collaborate with a central server to train a shared ML model. Each device encodes practical data into quantum form using a state encoder, processes it through a PQC with trainable angle parameters, and then employs measurement outcomes to update local model parameters using the standard SGD method, subsequently sending the updated local models to the server for aggregation.
  • Figure 2: Training performance comparison of QFL for MNIST dataset (IID) for full device participation with varying number of quantum measurement shots.
  • Figure 3: Training performance comparison of QFL on Cifar10 and MNIST datasets (non-IID) for full device participation with varying number of quantum measurement shots.
  • Figure 4: Comparison of transmit power and channel selection optimization using QAOA and SCA for varying device numbers.
  • Figure 5: Comparative analysis of sum-rate performance across optimization schemes for varying number of devices.
  • ...and 3 more figures

Theorems & Definitions (11)

  • Lemma 1
  • proof
  • Lemma 2
  • proof
  • Lemma 3
  • proof
  • Lemma 4
  • proof
  • proof
  • Remark 1
  • ...and 1 more