Stochastic false data injection attacks against distributed state estimation in interconnected large-scale systems

doi:10.1016/j.jfranklin.2026.108434

Stochastic false data injection attacks against distributed state estimation in interconnected large-scale systems

2026-01-15

0

PRE

AI

DOI：10.1016/j.jfranklin.2026.108434

Original

Original request for help

Save

Abstract

En 中文

In this paper, a novel distributed estimator has been designed for the interconnected large-scale systems (LSSs). In LSSs, the interconnection property of the subsystems makes it difficult to design a distributed estimator. Therefore, an unknown-input observer (UIO)-like observer is introduced to handle the unknown system information. Combined with the Kalman filter, the optimal estimation under Gaussian noises has also been guaranteed. Moreover, the vulnerability of the proposed estimation algorithm under malicious attacks has been investigated, including the conditions of the existence of stealthy false data injection (FDI) attacks. Note that the existing researches mainly focus on the attacks targeted at the wireless channels between the cyber layer and the physical layer in general cyber-physical system (CPS) model. However, as one kind of CPSs, a special feature of the LSS is that the distributed estimators need exchange data through wireless channels in the cyber layer. Motivated by the consideration above, the problem of designing optimal FDI attacks in cyber layer has also been addressed. Finally, simulation results are provided based on the three-area interconnected power system to illustrate the feasibility and advantages of the theoretical results.

Stochastic false data injection attacks against distributed state estimation in interconnected large-scale systems

Abstract

Journal

Organization