GenAI-FDIA: Physics-Informed Generative Models for False Data Injection Attacks

📅 2026-05-15
📈 Citations: 0
Influential: 0
📄 PDF
🤖 AI Summary
This study addresses the limitations in false data injection attack (FDIA) detection research—namely, the scarcity of realistic data and the lack of physical consistency in handcrafted attacks—by proposing GenAI-FDIA, a framework that systematically evaluates the capacity of 20 generative models, including Wasserstein GANs, MMD-VAEs, normalizing flows, diffusion models, and their hybrid architectures, to produce physically consistent FDIAs on IEEE benchmark power grids. The work introduces a data-driven bad data detection (BDD) threshold calibration mechanism and makes three core contributions: identifying a novel failure mode wherein affine physical projection in normalized space nullifies attacks, proposing a training-free inference-time coordinator to restore attack stealthiness, and diagnosing and mitigating covariance collapse in hybrid architectures. Experiments demonstrate BDD evasion rates ≥86.6% on the 14-bus system; the coordinator boosts evasion from <2% to 100% on the 30-bus system; and a 50-epoch warm-up schedule markedly improves covariance alignment (κ rising from −0.076 to 0.785).
📝 Abstract
Training and evaluating false data injection attack (FDIA) detectors for power systems is constrained by data scarcity. Operational grid measurements are commercially sensitive, and hand-crafted attacks fail to capture complex distributional structures imposed by network physics. We present \textsc{GenAI-FDIA}, a framework benchmarking a pool of $P{=}20$ architectures for physics-compliant FDIA synthesis, spanning Wasserstein GANs, MMD-VAEs, normalising flows, diffusion models, and cross-family hybrids. These are evaluated across three IEEE testbeds (14-bus DC, 30-bus DC, and 14-bus AC) under a 60/20/20 chronological split using data-driven Bad Data Detection (BDD) threshold calibration. Our empirical results verify that these models generate high-fidelity attacks, with all architectures achieving evasion rates of $ε_{\text{BDD}} \ge 86.6\%$ on the 14-bus network; additionally, limiting an attacker's topological knowledge induces a measurable degradation in stealthiness ($p \le 0.0022$). Crucially, we identify a previously unreported failure mode: applying affine physics projections directly in normalised feature spaces critically displaces the attack vector, collapsing BDD evasion from ${\sim}55\%$ to $<\!2\%$ on the 30-bus testbed. We resolve this via a novel inference-time harmoniser, restoring full stealthiness ($ε_{\text{BDD}}{=}100\%$) across all physics-informed variants without retraining. Finally, we isolate a covariance-collapse phenomenon ($κ\approx {-}0.076$) within advanced hybrid architectures and rectify it through 50-epoch warm-up schedules ($κ\to 0.785$, $Δ\text{MMD}={-}3.1\%$). Ultimately, \textsc{GenAI-FDIA} delivers a robust recovery blueprint applicable to any physics-constrained generative model deployed for power-system security.
Problem

Research questions and friction points this paper is trying to address.

False Data Injection Attack
Data Scarcity
Power Systems
Physics-Informed Generative Models
Bad Data Detection
Innovation

Methods, ideas, or system contributions that make the work stand out.

physics-informed generative models
false data injection attacks
inference-time harmoniser
covariance collapse
BDD evasion
M
Mohammad A. Razzaque
School of Computing, Engineering and Digital Technologies, Teesside University, UK
M
Muta Tah Hira
Smartifier Ltd, Stockton-on-Tees, UK