Logo
ScienceStack
Table of Contents
Fetching ...
Sign In

Balancing Security and Accuracy: A Novel Federated Learning Approach for Cyberattack Detection in Blockchain Networks

Tran Viet Khoa, Mohammad Abu Alsheikh, Yibeltal Alem, Dinh Thai Hoang

Abstract

This paper presents a novel Collaborative Cyberattack Detection (CCD) system aimed at enhancing the security of blockchain-based data-sharing networks by addressing the complex challenges associated with noise addition in federated learning models. Leveraging the theoretical principles of differential privacy, our approach strategically integrates noise into trained sub-models before reconstructing the global model through transmission. We systematically explore the effects of various noise types, i.e., Gaussian, Laplace, and Moment Accountant, on key performance metrics, including attack detection accuracy, deep learning model convergence time, and the overall runtime of global model generation. Our findings reveal the intricate trade-offs between ensuring data privacy and maintaining system performance, offering valuable insights into optimizing these parameters for diverse CCD environments. Through extensive simulations, we provide actionable recommendations for achieving an optimal balance between data protection and system efficiency, contributing to the advancement of secure and reliable blockchain networks.

Balancing Security and Accuracy: A Novel Federated Learning Approach for Cyberattack Detection in Blockchain Networks

Abstract

This paper presents a novel Collaborative Cyberattack Detection (CCD) system aimed at enhancing the security of blockchain-based data-sharing networks by addressing the complex challenges associated with noise addition in federated learning models. Leveraging the theoretical principles of differential privacy, our approach strategically integrates noise into trained sub-models before reconstructing the global model through transmission. We systematically explore the effects of various noise types, i.e., Gaussian, Laplace, and Moment Accountant, on key performance metrics, including attack detection accuracy, deep learning model convergence time, and the overall runtime of global model generation. Our findings reveal the intricate trade-offs between ensuring data privacy and maintaining system performance, offering valuable insights into optimizing these parameters for diverse CCD environments. Through extensive simulations, we provide actionable recommendations for achieving an optimal balance between data protection and system efficiency, contributing to the advancement of secure and reliable blockchain networks.
Paper Structure (22 sections, 1 theorem, 20 equations, 9 figures, 2 tables, 1 algorithm)

This paper contains 22 sections, 1 theorem, 20 equations, 9 figures, 2 tables, 1 algorithm.

Table of Contents

  1. Introduction
  2. Related work
  3. Machine Learning-Based Detection Methods for Blockchain Networks
  4. Collaborative Detection Methods for Blockchain Networks
  5. Proposed CCD in Blockchain-Based Data Sharing
  6. The Fundamental of Blockchain Networks
  7. The Proposed System Model
  8. Privacy-preserving CCD
  9. Gaussian Mechanism
  10. Laplace Mechanism
  11. Moment Accountant Mechanism
  12. Aggegation Process
  13. Experiment Setup and Evaluation Method
  14. Blockchain Cyberattack Dataset
  15. CCD System Setup
  16. ...and 7 more sections

Key Result

Proposition 1

The aggregation of a collaborative learning in a blockchain network is ($\bar{\epsilon},\bar{\delta}$)-differentially private, where $\bar{\epsilon}$ and $\bar{\delta}$ are bounded as follows: where $\grave{\delta}\geq0$ helps bound the cumulative privacy loss over multiple aggregations (multiple iterations and devices), $T$ is the number of training iterations, $N$ is the number of clusters, and

Figures (9)

  • Figure 1: The proposed CCD systems deployed in a blockchain-based data-sharing network for collaborative training-process and real-time cyberattack detection.
  • Figure 2: The collaborative learning process among clusters.
  • Figure 3: The visualization of the dataset.
  • Figure 4: Securing CCD system with different DL models.
  • Figure 5: Securing CCD system with different noises: (a) The Gaussian noises, (b) The Laplace noises, and (c) The Gaussian MA noises.
  • ...and 4 more figures

Theorems & Definitions (3)

  • Definition 1
  • Proposition 1
  • proof