Logo
ScienceStack
Table of Contents
Fetching ...
Sign In

A Data-Driven Framework for Online Mitigation of False Data Injection Signals in Networked Control Systems

Mohammadamin Lari

TL;DR

This work tackles the threat of False Data Injection (FDI) in Networked Control Systems (NCSs) by introducing a two-stage, data-driven framework that combines meta-learning with online mitigation. Stage I converts time series data into scalogram-based time-frequency representations, uses an entropy-based complexity measure via Optimally Improved Permutation Entropy (OIPE), and classifies data complexity with a CNN to dynamically select a base forecasting model from a stacked ensemble. Stage II offline designs five diverse base models (GRU, LSTM, CNN-LSTM, ResNet-50-LSTM, Modified Xception-LSTM), assigns them to complexity levels, and performs online $q$-step forecasting to reconstruct and counteract FDI signals in real time. The framework is validated through simulations of a four-robot leader-follower formation, demonstrating improved prediction accuracy and robust trajectory tracking under FDI attacks, with favorable comparisons to existing data-driven mitigators. The approach advances online security for NCSs by integrating time-frequency analysis, entropy-based complexity classification, and adaptive model selection within a real-time mitigation loop, offering practical resiliency benefits for cyber-physical systems.

Abstract

This paper introduces a novel two-stage framework for online mitigation of False Data Injection (FDI) signals to improve the resiliency of Networked Control Systems (NCSs) and ensure their safe operation in the presence of malicious activities. The first stage involves meta learning to select a base time series forecasting model within a stacked ensemble learning architecture. This is achieved by converting time series data into scalograms using continuous wavelet transform, which are then split into image frames to generate a scalo-temporal representation of the data and to distinguish between different complexity levels of time series data based on an entropy metric using a convolutional neural network. In the second stage, the selected model mitigates false data injection signals in real-time. The proposed framework's effectiveness is demonstrated through rigorous simulations involving the formation control of differential drive mobile robots. By addressing the security challenges in NCSs, this framework offers a promising approach to maintaining system integrity and ensuring operational safety.

A Data-Driven Framework for Online Mitigation of False Data Injection Signals in Networked Control Systems

TL;DR

This work tackles the threat of False Data Injection (FDI) in Networked Control Systems (NCSs) by introducing a two-stage, data-driven framework that combines meta-learning with online mitigation. Stage I converts time series data into scalogram-based time-frequency representations, uses an entropy-based complexity measure via Optimally Improved Permutation Entropy (OIPE), and classifies data complexity with a CNN to dynamically select a base forecasting model from a stacked ensemble. Stage II offline designs five diverse base models (GRU, LSTM, CNN-LSTM, ResNet-50-LSTM, Modified Xception-LSTM), assigns them to complexity levels, and performs online -step forecasting to reconstruct and counteract FDI signals in real time. The framework is validated through simulations of a four-robot leader-follower formation, demonstrating improved prediction accuracy and robust trajectory tracking under FDI attacks, with favorable comparisons to existing data-driven mitigators. The approach advances online security for NCSs by integrating time-frequency analysis, entropy-based complexity classification, and adaptive model selection within a real-time mitigation loop, offering practical resiliency benefits for cyber-physical systems.

Abstract

This paper introduces a novel two-stage framework for online mitigation of False Data Injection (FDI) signals to improve the resiliency of Networked Control Systems (NCSs) and ensure their safe operation in the presence of malicious activities. The first stage involves meta learning to select a base time series forecasting model within a stacked ensemble learning architecture. This is achieved by converting time series data into scalograms using continuous wavelet transform, which are then split into image frames to generate a scalo-temporal representation of the data and to distinguish between different complexity levels of time series data based on an entropy metric using a convolutional neural network. In the second stage, the selected model mitigates false data injection signals in real-time. The proposed framework's effectiveness is demonstrated through rigorous simulations involving the formation control of differential drive mobile robots. By addressing the security challenges in NCSs, this framework offers a promising approach to maintaining system integrity and ensuring operational safety.
Paper Structure (33 sections, 52 equations, 9 figures, 10 tables, 3 algorithms)

This paper contains 33 sections, 52 equations, 9 figures, 10 tables, 3 algorithms.

Table of Contents

  1. Introduction
  2. Stage I - Meta Learning Based on Classification of Data in Time-Frequency Domain
  3. Problem Formulation
  4. Stacked Ensemble Learning Architecture
  5. Phase I: Scalo-Temporal Feature Extraction using Continuous Wavelet Transform
  6. Phase II: Complexity Analysis using Entropy Metric
  7. Phase III: FDI Complexity Classification using CNN
  8. Stage II - Mitigation of FDI Attacks
  9. Phase I: Design Procedure of Base Time Series Forecasting Models
  10. Gated Recurrent Unit (GRU) Network
  11. LSTM Network
  12. CNN-LSTM Network
  13. ResNet-50-LSTM Network
  14. Modified Xception-LSTM Network
  15. Phase II: Model Assignment to Distinct Complexity Levels
  16. ...and 18 more sections

Figures (9)

  • Figure 1: Diagram representation of the subsystems in a NCS with hierarchical structure and the security vulnerabilities (red circles)
  • Figure 2: Stage I of proposed framework for online mitigation of FDI attacks.
  • Figure 3: Stage II of proposed framework for online mitigation of FDI attacks.
  • Figure 4: The topology of the mobile robots.
  • Figure 5: Scalogram generated for: (a) Normal condition, (b) Under FDI attack.
  • ...and 4 more figures