Game Theory Based Optimal Defensive Resources Allocation with Incomplete Information in Cyber-Physical Power Systems Against False Data Injection Attacks

Bingjing Yan, Zhenze Jiang, Pengchao Yao, Qiang Yang, Wei Li, Albert Y. ZomayaProtection 2024 · DOI

Abstract

Modern power grids are fast emerging as complex cyber-physical power systems (CPPS) integrating physical current-carrying components with cyber-embedded computing. This integration introduces significant cyberspace security threats, notably False Data Injection (FDI) attacks. This paper proposes a zero-sum Bayesian game-theoretical model to quantify attack consequences and optimally allocate limited defensive resources under incomplete information. By combining state estimation offsets with a comprehensive bus safety characterization, the model identifies critical nodes and optimal strategies. Evaluations on IEEE 33-bus and 14-bus networks demonstrate that the proposed approach significantly enhances system resilience compared to traditional vulnerability-based methods.

Introduction

The digital transformation of power grids into Cyber-Physical Power Systems (CPPS) has enabled more efficient management but also expanded the attack surface. In these systems, physical components are controlled by a digital layer of sensors, communication networks, and controllers. A primary concern is the False Data Injection (FDI) attack, where an adversary manipulates sensor measurements to deceive the control center's state estimation without triggering "Bad Data Detection" alarms.

CPPS architecture illustrating the interaction between attacker strategies and defender state estimation.

The challenge for grid operators lies in the scarcity of defensive resources—such as high-precision Phasor Measurement Units (PMUs) or advanced encryption. Furthermore, defenders often lack complete information regarding an attacker's specific resources or rewards, making traditional "complete information" models impractical for real-world deployment.

Methodology

The researchers developed a multi-layered framework to address the resource allocation problem under uncertainty. The process involves three main stages: impact quantification, game modeling, and optimization.

Flowchart of the proposed game-theoretical framework for optimal resource allocation.

1. Impact Quantification and Bus Safety

The framework first uses Weighted Least Squares (WLS) state estimation to calculate voltage offsets resulting from potential attacks. To prioritize nodes, a "bus weight" is calculated based on:

  • Topological Criticality: Measured via centrality and betweenness.
  • Network Cohesion: Analyzing how nodes cluster within the grid.
  • Equipment Value: Quantified using the Common Vulnerability Scoring System (CVSS) to reflect both physical and cyber importance.

2. Bayesian Game Model

Interaction is modeled as a zero-sum Bayesian game. Unlike standard games, this approach accounts for "incomplete information," where the defender estimates the attacker's capabilities through expected utilities. The Lemke-Howson Algorithm is employed to find the Nash Equilibrium—the point where neither player can improve their outcome by changing strategy alone.

3. Resource Optimization

Finally, the ORIGAMI algorithm is used to solve the resource allocation problem efficiently, ensuring that the limited budget is distributed to maximize the grid's overall safety.

Experimental Validation

The model was tested using two industry-standard benchmarks:

  • IEEE 33-bus test network: Conducted in MATLAB to verify the game-theoretical logic across various attack scenarios.
  • IEEE 14-bus network: Validated on a Real-Time Digital Simulator (RTDS). This hardware-in-the-loop testbed provides high-fidelity results by simulating the actual timing and physical interactions of a power system.

Key Results and Insights

The proposed game-theoretical approach demonstrated significant improvements over traditional strategies:

  • Enhanced Utility: The defender's expected utility increased by 19.81% compared to non-strategic allocation.
  • Superiority to Baselines: The method outperformed simple vulnerability-based allocation (relying solely on CVSS scores) by 7.87%.
  • Real-world Mitigation: RTDS experiments showed a 73.76% reduction in FDI attack impact, a 6.93% improvement over standard vulnerability analysis methods.

A critical insight from the study is the concept of the "Strategic Idle." The model identifies tipping points where the cost of an action (attack or defense) outweighs the potential gain, leading rational players to remain idle. Additionally, the holistic approach to node importance—combining software vulnerabilities with physical topology—identified critical nodes that would be missed by single-domain analysis.

Conclusion

This research provides a robust framework for defending CPPS against stealthy FDI attacks when information is incomplete and resources are limited. Future work will focus on extending the model to multi-player dynamics (multiple simultaneous attackers) and incorporating temporal factors like system recovery time to further enhance grid stability.

Summary updated

Abstract

Modern power grid is fast emerging as a complex cyber-physical power system (CPPS) integrating physical current-carrying components and processes with cyber-embedded computing, which faces increasing cyberspace security threats and risks. In this paper, the state (i.e., voltage) offsets resulting from false data injection (FDI) attacks and the bus safety characterization are applied to quantify the attack consequences. The state offsets are obtained by the state estimation method, and the bus safety characterization considers the power network topology as well as the vulnerability and connection relationship of buses. Considering the indeterminacy of attacker's resource consumption and reward, a zero-sum game-theoretical model from the defender's perspective with incomplete information is explored for the optimal allocation of limited defensive resources. The attacker aims to falsify measurements without triggering threshold alarms to break through the protection, leading to load shedding, over-voltage or under-voltage. The defender attempts to ensure the estimation results to be as close to the actual states as possible, and guarantee the system's safety and efficient defensive resource utilization. The proposed solution is extensively evaluated through simulations using the IEEE 33-bus test network and real-time digital simulator (RTDS) based testbed experiments of the IEEE 14-bus network. The results demonstrate the effectiveness of the proposed game-theoretical approach for optimal defensive resource allocation in CPPS when limited resources are available when under FDI attacks.
BibTeX
@inproceedings{yan2024game,
  author = {Bingjing Yan and Zhenze Jiang and Pengchao Yao and Qiang Yang and Wei Li and Albert Y. Zomaya},
  title = {{Game Theory Based Optimal Defensive Resources Allocation with Incomplete Information in Cyber-Physical Power Systems Against False Data Injection Attacks}},
  booktitle = {Protection},
  year = {2024},
  doi = {10.23919/pcmp.2023.000138},
}
Figure