Logo
ScienceStack
Table of Contents
Fetching ...
Sign In

Energy Efficient Federated Learning with Hyperdimensional Computing (HDC)

Yahao Ding, Yinchao Yang, Jiaxiang Wang, Zhonghao Liu, Zhaohui Yang, Mingzhe Chen, Mohammad Shikh-Bahaei

TL;DR

This paper investigates the problem of minimizing total energy consumption for secure federated learning in wireless edge networks, a key paradigm for decentralized big data analytics and proposes an FL with hyperdimensional computing and differential privacy (FL-HDC-DP) framework.

Abstract

This paper investigates the problem of minimizing total energy consumption for secure federated learning (FL) in wireless edge networks, a key paradigm for decentralized big data analytics. To tackle the high computational cost and privacy challenges of processing large-scale distributed data with conventional neural networks, we propose an FL with hyperdimensional computing and differential privacy (FL-HDC-DP) framework. Each edge device employs hyperdimensional computing (HDC) for lightweight local training and applies differential privacy (DP) noise to protect transmitted model updates. The total energy consumption is minimized through a joint optimization of the HDC dimension, transmit power, and CPU frequency. An efficient hybrid algorithm is developed, combining an outer enumeration search for HDC dimensions with an inner one-dimensional search for resource allocation. Simulation results show that the proposed framework achieves up to 83.3% energy reduction compared with baseline schemes, while maintaining high accuracy and faster convergence.

Energy Efficient Federated Learning with Hyperdimensional Computing (HDC)

TL;DR

This paper investigates the problem of minimizing total energy consumption for secure federated learning in wireless edge networks, a key paradigm for decentralized big data analytics and proposes an FL with hyperdimensional computing and differential privacy (FL-HDC-DP) framework.

Abstract

This paper investigates the problem of minimizing total energy consumption for secure federated learning (FL) in wireless edge networks, a key paradigm for decentralized big data analytics. To tackle the high computational cost and privacy challenges of processing large-scale distributed data with conventional neural networks, we propose an FL with hyperdimensional computing and differential privacy (FL-HDC-DP) framework. Each edge device employs hyperdimensional computing (HDC) for lightweight local training and applies differential privacy (DP) noise to protect transmitted model updates. The total energy consumption is minimized through a joint optimization of the HDC dimension, transmit power, and CPU frequency. An efficient hybrid algorithm is developed, combining an outer enumeration search for HDC dimensions with an inner one-dimensional search for resource allocation. Simulation results show that the proposed framework achieves up to 83.3% energy reduction compared with baseline schemes, while maintaining high accuracy and faster convergence.
Paper Structure (20 sections, 20 equations, 3 figures, 2 tables, 1 algorithm)

This paper contains 20 sections, 20 equations, 3 figures, 2 tables, 1 algorithm.

Table of Contents

  1. Introduction
  2. System Model
  3. HDC Model
  4. Encoding
  5. Training
  6. Inference
  7. Retraining
  8. The Framework of FL-HDC-DP
  9. Transmission Model
  10. Energy Consumption Model
  11. Energy Consumption of Computation Model
  12. Energy Consumption of Transmission Model
  13. Total Energy Consumption
  14. Problem Formulation
  15. Algorithm Design
  16. ...and 5 more sections

Figures (3)

  • Figure 1: The FL-HDC-DP model over wireless communication networks.
  • Figure 2: Total energy versus HDC dimension with $T=30s$.
  • Figure 3: Total energy consumption versus total time.