Logo
ScienceStack
Table of Contents
Fetching ...
Sign In

Decentralized Federated Learning With Energy Harvesting Devices

Kai Zhang, Xuanyu Cao, Khaled B. Letaief

TL;DR

This work derives the convergence bound for wireless DFL with energy harvesting, and proposes a fully decentralized policy iteration algorithm that leverages only local state information from two-hop neighboring devices, thereby substantially reducing both communication overhead and computational complexity.

Abstract

Decentralized federated learning (DFL) enables edge devices to collaboratively train models through local training and fully decentralized device-to-device (D2D) model exchanges. However, these energy-intensive operations often rapidly deplete limited device batteries, reducing their operational lifetime and degrading the learning performance. To address this limitation, we apply energy harvesting technique to DFL systems, allowing edge devices to extract ambient energy and operate sustainably. We first derive the convergence bound for wireless DFL with energy harvesting, showing that the convergence is influenced by partial device participation and transmission packet drops, both of which further depend on the available energy supply. To accelerate convergence, we formulate a joint device scheduling and power control problem and model it as a multi-agent Markov decision process (MDP). Traditional MDP algorithms (e.g., value or policy iteration) require a centralized coordinator with access to all device states and exhibit exponential complexity in the number of devices, making them impractical for large-scale decentralized networks. To overcome these challenges, we propose a fully decentralized policy iteration algorithm that leverages only local state information from two-hop neighboring devices, thereby substantially reducing both communication overhead and computational complexity. We further provide a theoretical analysis showing that the proposed decentralized algorithm achieves asymptotic optimality. Finally, comprehensive numerical experiments on real-world datasets are conducted to validate the theoretical results and corroborate the effectiveness of the proposed algorithm.

Decentralized Federated Learning With Energy Harvesting Devices

TL;DR

This work derives the convergence bound for wireless DFL with energy harvesting, and proposes a fully decentralized policy iteration algorithm that leverages only local state information from two-hop neighboring devices, thereby substantially reducing both communication overhead and computational complexity.

Abstract

Decentralized federated learning (DFL) enables edge devices to collaboratively train models through local training and fully decentralized device-to-device (D2D) model exchanges. However, these energy-intensive operations often rapidly deplete limited device batteries, reducing their operational lifetime and degrading the learning performance. To address this limitation, we apply energy harvesting technique to DFL systems, allowing edge devices to extract ambient energy and operate sustainably. We first derive the convergence bound for wireless DFL with energy harvesting, showing that the convergence is influenced by partial device participation and transmission packet drops, both of which further depend on the available energy supply. To accelerate convergence, we formulate a joint device scheduling and power control problem and model it as a multi-agent Markov decision process (MDP). Traditional MDP algorithms (e.g., value or policy iteration) require a centralized coordinator with access to all device states and exhibit exponential complexity in the number of devices, making them impractical for large-scale decentralized networks. To overcome these challenges, we propose a fully decentralized policy iteration algorithm that leverages only local state information from two-hop neighboring devices, thereby substantially reducing both communication overhead and computational complexity. We further provide a theoretical analysis showing that the proposed decentralized algorithm achieves asymptotic optimality. Finally, comprehensive numerical experiments on real-world datasets are conducted to validate the theoretical results and corroborate the effectiveness of the proposed algorithm.
Paper Structure (28 sections, 3 theorems, 57 equations, 5 figures, 1 table, 1 algorithm)

This paper contains 28 sections, 3 theorems, 57 equations, 5 figures, 1 table, 1 algorithm.

Table of Contents

  1. Introduction
  2. Contributions
  3. Organization and Notations
  4. System Model
  5. Decentralized Federated Learning Model
  6. Local Computation Model
  7. Communication Model
  8. Energy Harvesting Model
  9. Convergence Analysis and Problem Formulation
  10. Convergence Analysis
  11. Problem Formulation
  12. MDP Formulation and Decentralized Policy Iteration Algorithm
  13. MDP Formulation
  14. Decentralized Policy Iteration Algorithm
  15. Localized Policy
  16. ...and 13 more sections

Key Result

Theorem 1

Under Assumptions ass_smoothness_assump, ass_bounded_var_assump, ass_bounded_G_assump, if we set $\eta=\tfrac{\sqrt{m}}{64LK\sqrt{T}}$, it follows that where $C_1 = \sigma_l^2 +4 K \sigma_g^2$.

Figures (5)

  • Figure 1: An illustration of the wireless DFL system with EH devices.
  • Figure 2: Test accuracy v.s. time slots for various transmission schemes on FMNIST.
  • Figure 3: Test accuracy v.s. time slots for various transmission schemes on CIFAR-10.
  • Figure 4: Test accuracy v.s. battery capacity for different transmission strategies on FMNIST.
  • Figure 5: Final test accuracy versus the number of policy-improvement iterations $R$.

Theorems & Definitions (8)

  • Theorem 1
  • proof
  • Remark 1
  • Remark 2
  • Theorem 2
  • proof
  • Remark 3
  • Lemma 1