Cyber Risk Analysis of Combined Data Attacks Against Power System State Estimation

Kaikai Pan, André Teixeira, Milos Cvetkovic, Peter Palensky

Abstract

Understanding smart grid cyber attacks is key for developing appropriate protection and recovery measures. Advanced attacks pursue maximized impact at minimized costs and detectability. This paper conducts risk analysis of combined data integrity and availability attacks against the power system state estimation. We compare the combined attacks with pure integrity attacks - false data injection (FDI) attacks. A security index for vulnerability assessment to these two kinds of attacks is proposed and formulated as a mixed integer linear programming problem. We show that such combined attacks can succeed with fewer resources than FDI attacks. The combined attacks with limited knowledge of the system model also expose advantages in keeping stealth against the bad data detection. Finally, the risk of combined attacks to reliable system operation is evaluated using the results from vulnerability assessment and attack impact analysis. The findings in this paper are validated and supported by a detailed case study.

Introduction

Modern power grids rely on State Estimation (SE) to transform raw sensor data from Remote Terminal Units (RTUs) into a real-time representation of the system's operational health. To protect against sensor errors or malicious interference, grids employ Bad Data Detection (BDD) mechanisms.

Historically, research has focused on False Data Injection (FDI) attacks, where an adversary modifies sensor readings to bypass BDD. However, pure FDI is resource-intensive, requiring perfect knowledge of the grid's physics and control over numerous sensors. This paper investigates combined attacks—hybrid strategies that mix data tampering (integrity) with data blocking (availability/DoS). By blocking specific data streams, an attacker can "blind" the BDD, making the remaining injected false data significantly harder to detect.

Vulnerability Assessment Framework

The authors develop a mathematical framework to quantify the risk of these hybrid threats using a new Security Index. This index measures the minimum resources (a combination of FDI and DoS) needed to compromise a specific measurement while remaining stealthy. The problem is formulated as a Mixed Integer Linear Programming (MILP) problem using the "big M" method.

Risk is defined as the product of Likelihood (the probability an attack is launched and remains undetected) and Impact (the resulting error in the grid's load estimation). Unlike many studies that assume an omniscient attacker, this model incorporates "structured uncertainty," simulating an adversary with up to a 20% error in their knowledge of line parameters.

Figure 1: Security index mapping on the IEEE 14-bus system, highlighting vulnerable components by color-coding their security levels.

Key Insights and Results

Resource Efficiency

The study demonstrates that combined attacks are the "economical" choice for adversaries. In a case study of the IEEE 14-bus system, compromising a specific measurement via pure FDI required 11 sensors. However, a combined attack achieved the same goal with only one integrity compromise and ten availability compromises. If blocking a sensor is cheaper than hacking it to change its value, the barrier to entry for damaging attacks is significantly lowered.

Stealth Under Uncertainty

One of the most critical findings is that combined attacks are more robust to attacker ignorance. Even with an imperfect model of the grid, a combined attack is less likely to trigger a BDD alarm than a pure FDI attack.

Figure 2: Detection probability of various attack types against attack magnitude, demonstrating that combined attacks (e.g., (2,9)-tuple) maintain higher stealth than pure FDI.

As shown in Figure 2, empirical Monte Carlo simulations confirm that combined attacks (such as the (2,9)-tuple) have a lower probability of detection than pure 11-tuple FDI attacks across various attack magnitudes. This increased stealth directly translates to a higher overall risk metric for the system.

Conclusion and Future Work

The analysis proves that combined data integrity and availability attacks pose a more severe threat to power system stability than traditional FDI attacks. They require fewer resources and are harder to detect under realistic conditions of limited attacker knowledge.

While this study utilizes the linear DC power flow model, the authors acknowledge that real-world grids operate on non-linear AC power. Future research will focus on extending this framework to AC models and investigating direct physical consequences, such as line overflows or cascading blackouts. Additionally, the findings suggest that future defense strategies must integrate IT network traffic analysis with power system state estimation to detect the signature of availability-based attacks.

Summary updated

Abstract

Understanding smart grid cyber attacks is key for developing appropriate protection and recovery measures. Advanced attacks pursue maximized impact at minimized costs and detectability. This paper conducts risk analysis of combined data integrity and availability attacks against the power system state estimation. We compare the combined attacks with pure integrity attacks - false data injection (FDI) attacks. A security index for vulnerability assessment to these two kinds of attacks is proposed and formulated as a mixed integer linear programming problem. We show that such combined attacks can succeed with fewer resources than FDI attacks. The combined attacks with limited knowledge of the system model also expose advantages in keeping stealth against the bad data detection. Finally, the risk of combined attacks to reliable system operation is evaluated using the results from vulnerability assessment and attack impact analysis. The findings in this paper are validated and supported by a detailed case study.
BibTeX
@article{pan2019cyber,
  author = {Kaikai Pan and André Teixeira and Milos Cvetkovic and Peter Palensky},
  title = {{Cyber Risk Analysis of Combined Data Attacks Against Power System State Estimation}},
  journal = {IEEE Trans. Smart Grid},
  year = {2019},
  url = {https://openreview.net/forum?id=M7HuafgPO9},
}
Figure