Logo
ScienceStack
Table of Contents
Fetching ...
Sign In

Semi-Supervised Learning for Anomaly Detection in Blockchain-based Supply Chains

Do Hai Son, Bui Duc Manh, Tran Viet Khoa, Nguyen Linh Trung, Dinh Thai Hoang, Hoang Trong Minh, Yibeltal Alem, Le Quang Minh

TL;DR

The paper addresses anomaly detection in blockchain-based supply chains by focusing on network-layer traffic to detect attacks across the network, consensus, and beyond. It proposes a semi-supervised DAE-MLP that combines a Deep AutoEncoder for unsupervised normal-state modeling with a supervised MLP classifier, producing a joint anomaly score and an adaptive threshold for decision making. Experiments on a lab Ethereum-based BSC with five attack types show that DAE-MLP achieves about $96.5\%$ accuracy, outperforming unsupervised and traditional supervised detectors, and can improve detection of novel attacks by up to $33.1\%$ F1-score after updating the MLP. The approach provides a practical, adaptable defense for BSCs, enabling early attack detection and resilience as new threats emerge.

Abstract

Blockchain-based supply chain (BSC) systems have tremendously been developed recently and can play an important role in our society in the future. In this study, we develop an anomaly detection model for BSC systems. Our proposed model can detect cyber-attacks at various levels, including the network layer, consensus layer, and beyond, by analyzing only the traffic data at the network layer. To do this, we first build a BSC system at our laboratory to perform experiments and collect datasets. We then propose a novel semi-supervised DAE-MLP (Deep AutoEncoder-Multilayer Perceptron) that combines the advantages of supervised and unsupervised learning to detect anomalies in BSC systems. The experimental results demonstrate the effectiveness of our model for anomaly detection within BSCs, achieving a detection accuracy of 96.5%. Moreover, DAE-MLP can effectively detect new attacks by improving the F1-score up to 33.1% after updating the MLP component.

Semi-Supervised Learning for Anomaly Detection in Blockchain-based Supply Chains

TL;DR

The paper addresses anomaly detection in blockchain-based supply chains by focusing on network-layer traffic to detect attacks across the network, consensus, and beyond. It proposes a semi-supervised DAE-MLP that combines a Deep AutoEncoder for unsupervised normal-state modeling with a supervised MLP classifier, producing a joint anomaly score and an adaptive threshold for decision making. Experiments on a lab Ethereum-based BSC with five attack types show that DAE-MLP achieves about accuracy, outperforming unsupervised and traditional supervised detectors, and can improve detection of novel attacks by up to F1-score after updating the MLP. The approach provides a practical, adaptable defense for BSCs, enabling early attack detection and resilience as new threats emerge.

Abstract

Blockchain-based supply chain (BSC) systems have tremendously been developed recently and can play an important role in our society in the future. In this study, we develop an anomaly detection model for BSC systems. Our proposed model can detect cyber-attacks at various levels, including the network layer, consensus layer, and beyond, by analyzing only the traffic data at the network layer. To do this, we first build a BSC system at our laboratory to perform experiments and collect datasets. We then propose a novel semi-supervised DAE-MLP (Deep AutoEncoder-Multilayer Perceptron) that combines the advantages of supervised and unsupervised learning to detect anomalies in BSC systems. The experimental results demonstrate the effectiveness of our model for anomaly detection within BSCs, achieving a detection accuracy of 96.5%. Moreover, DAE-MLP can effectively detect new attacks by improving the F1-score up to 33.1% after updating the MLP component.
Paper Structure (12 sections, 8 equations, 3 figures, 4 tables, 1 algorithm)

This paper contains 12 sections, 8 equations, 3 figures, 4 tables, 1 algorithm.

Table of Contents

  1. Introduction
  2. BSC Systems and Our Proposed Anomaly Detection Model
  3. Blockchain-based Agriculture Supply Chains
  4. Proposed Anomaly Detection Model
  5. Proposed DAE-MLP Network
  6. Experimental Results
  7. Experiment Setup and Evaluation Methods
  8. Performance Evaluation
  9. Performance Comparison
  10. Threshold Values
  11. Adaptation
  12. Conclusion

Figures (3)

  • Figure 1: Proposed system model of anomaly detection on the Ethereum-based agriculture supply chain.
  • Figure 2: Proposed semi-supervised DAE-MLP network for anomaly detection in BSCs.
  • Figure 3: The performance of proposed DAE-MLP vs. other anomaly detectors and bi-classifiers.