DiPRoD: Differentially-Private Robust Distributed Detection Method for False Data Injection Attacks in Power Systems with Measurement Redundancy

Shutan Wu; Qi Wang; Ziyi Zhang; Yi Tang; Wei He

doi:10.1016/j.eng.2026.07.030

Engineering ›› :202607030 DOI: 10.1016/j.eng.2026.07.030

research-article

DiPRoD: Differentially-Private Robust Distributed Detection Method for False Data Injection Attacks in Power Systems with Measurement Redundancy

Shutan Wu ^a^,^b
, Qi Wang ^a^,^*
, Ziyi Zhang ^a
, Yi Tang ^a
, Wei He ^b^,^*

Author information +

History +

PDF (1826KB)

Abstract

The increasing deployment of edge devices and the deep integration of cyber-physical infrastructures have made modern power systems highly dependent on measurement data and consequently more vulnerable to false data injection attacks (FDIAs). To overcome the communication, computational, and privacy limitations of centralized FDIA detection, this paper proposes a differentially-private robust distributed (DiPRoD) detection framework. First, FDIA behavior under incomplete system knowledge is explicitly modeled as an adversarial optimization at the regional level, yielding a regional robust detection problem with a composite index that jointly captures attack cost and state estimation deviation. The convexity of this problem and the convergence of an alternating gradient algorithm are established. Second, by exploiting spatiotemporal measurement redundancy, an active measurement transformation mechanism is designed and theoretically shown to enlarge the separability between attacked and normal samples in the composite-index space, thereby improving the detectability of stealthy FDIAs. Third, at the coordination level, a differential-privacy mechanism is incorporated such that each region uploads only Laplace-perturbed local parameters, while the control center performs weighted aggregation to construct a global robust detector. Finally, an online joint adaptation strategy for the global detection threshold and regional privacy budgets is developed to dynamically balance detection performance and privacy protection. Case studies on standard test systems demonstrate that, under prescribed privacy budgets, the proposed DiPRoD method achieves superior detection accuracy and robustness compared with centralized baselines, while maintaining favorable scalability for large-scale power systems.

Keywords

Power systems / False data injection attacks / Distributed detection / Robust optimization / Redundant measurement transformation / Differential privacy

Cite this article

Download citation ▾

Shutan Wu, Qi Wang, Ziyi Zhang, Yi Tang, Wei He. DiPRoD: Differentially-Private Robust Distributed Detection Method for False Data Injection Attacks in Power Systems with Measurement Redundancy. Engineering 202607030 DOI:10.1016/j.eng.2026.07.030

登录浏览全文

4963

注册一个新账户忘记密码

References

Publishing order | Descend order by publishing year | Descend order by cited within

[1]	Suvarna M, Yap K, Yang W, Li J, Ng Y, Wang X . Cyber—physical production systems for data—driven, decentralized, and secure manufacturing—a perspective. Engineering 2021; 7(9):1212-23.

[2]	Hu J, Wang Q, Ye Y, Wu Z, Tang Y . A high temporal—spatial resolution power system state estimation method for online DSA. IEEE Trans Power Syst 2024; 39(1):877-89.

[3]	Liu Z, Liu M, Wang Q, Tang Y . False data injection attacks on data—driven algorithms in smart grids utilizing distributed power supplies. Engineering 2025; 51:62-74.

[4]	Cheng G, Lin Y, Abur A, Gómez—Expósito A, Wu W . A survey of power system state estimation using multiple data sources: PMUs, SCADA, AMI, and beyond. IEEE Trans Smart Grid 2024; 15(1):1129-51.

[5]	Lian Z, Shi P, Chen M . A survey on cyber—attacks for cyber—physical systems: modeling, defense, and design. IEEE Internet Things J 2025; 12(2):1471-83.

[6]	Wu S, Wang Q, Hu J, Ye Y, Tang Y . Dynamic PMU configuration for stealthy multi—snapshot FDIA mitigation: enhancing security in hybrid—measurement—based state estimation in power systems. IEEE Trans Smart Grid 2026; 17(1):650-65.

[7]	Lakshminarayana S, Kammoun A, Debbah M, Poor HV . Data—driven false data injection attacks against power grids: a random matrix approach. IEEE Trans Smart Grid 2021; 12(1):635-46.

[8]	Li B, Ding T, Huang C, Zhao J, Yang Y, Chen Y . Detecting false data injection attacks against power system state estimation with fast go—decomposition approach. IEEE Trans Industr Inform 2019; 15(5):2892-904.

[9]	Qu J, Yang J, Wang Y, Georgievitch PM . Detection of false data injection attack in power system based on hellinger distance. IEEE Trans Industr Inform 2024; 20(2):2119-28.

[10]	Huang K, Xiang Z, Deng W, Yang C, Wang Z . False data injection attacks detection in smart grid: a structural sparse matrix separation method. IEEE Trans Netw Sci Eng 2021; 8(3):2545—58.

[11]	Wei S, Xu J, Wu Z, Hu Q, Yu X . A false data injection attack detection strategy for unbalanced distribution networks state estimation. IEEE Trans Smart Grid 2023; 14(5):3992-4006.

[12]	Musleh AS, Chen G, Dong ZY, Wang C, Chen S . Attack detection in automatic generation control systems using LSTM—based stacked autoencoders. IEEE Trans Industr Inform 2023; 19(1):153-65.

[13]	Xin L, He G, Long Z . Stealthy false data injection attacks detection and classification in cyber—physical systems using deep reinforcement learning. IEEE Trans Autom Sci Eng 2025; 22:141-53.

[14]	Xia W, Li Y, Yu L, He D . Locational detection of false data injection attacks in the edge space via hodge graph neural network for smart grids. IEEE Trans Smart Grid 2024; 15(5):5102—14.

[15]	Li X, Jiao W, Han Q, Lu Z . Detection of FDIA in power grid based on hypergraph and attention mechanism. IEEE Trans Smart Grid 2025; 16(2):1862-71.

[16]	Reda HT, Anwar A, Mahmood A, Chilamkurti N . Data—driven approach for state prediction and detection of false data injection attacks in smart grid. J Mod Power Syst Clean Energy 2023; 11(2):455-67.

[17]	Tahir B, Jolfaei A, Tariq M . Experience—driven attack design and federated—learning—based intrusion detection in Industry 4.0. IEEE Trans Industr Inform 2022; 18(9):6398-405.

[18]	Xu J, Zhang S, Lin T, Jiang Y, Ruan X, Wei S, et al. A distributed secure state estimation framework for unbalanced active distribution systems. IEEE Trans Smart Grid 2025; 16(5):3714—27.

[19]	Li Y, Wei X, Li Y, Dong Z, Shahidehpour M . Detection of false data injection attacks in smart grid: a secure federated deep learning approach. IEEE Trans Smart Grid 2022; 13(6):4862—72.

[20]	Li H, Dou C, Yue D, Hancke GP, Zeng Z, Guo W, et al. End—edge—cloud collaboration—based false data injection attack detection in distribution networks. IEEE Trans Industr Inform 2024; 20(2):1786-97.

[21]	Zhu M, Du D, Li X, Sun Q, Fei M, Wu L . Distributed security state estimation based on homomorphic encryption for privacy—preserving consensus in cloud environment. IEEE Internet Things J 2025; 12(23):50466-79.

[22]	Wu S, Wang Q, Zhang Z, Hu J, Ye Y, Tang Y . Physics—informed state estimation model with controllable uncertainty: enhancing resilience against false data injection attacks in power systems. IEEE Trans Ind Inform. In press.

[23]	Yang H, Wang Z . A false data injection attack approach without knowledge of system parameters considering measurement noise. IEEE Internet Things J 2024; 11(1):1452-64.

[24]	Wu S, Wang Q, Hu J, Ye Y, Tang Y . Attack—resilient state estimation for cyber—physical power systems: a dynamic spatial—temporal redundancy reconfiguration framework for FDIA detection. Appl Energy 2025; 397:126330.

[25]	Yang Q, An D, Min R, Yu W, Yang X, Zhao W . On optimal PMU placement—based defense against data integrity attacks in smart grid. IEEE Trans Inf Forensics Security 2017; 12(7):1735—50.

[26]	Wu S, Wang Q, Hu J, Ye Y, Tang Y, He W . Dynamic spatiotemporal measurement uncertainty—aware detection for robust power system state estimation against stealthy FDIA. IEEE Internet Things J 2026; 13(12):27656—71.

[27]	Wu Z, Zhao X . Parameter—separable Prox—Lagrangian method for convex—concave saddle point problem. IEEE Control Syst Lett 2024; 8:253— 8.

[28]	Muthukrishnan G, Kalyani S . Differential privacy with higher utility by exploiting coordinate—wise disparity: Laplace mechanism can beat Gaussian in high dimensions. IEEE Trans Inf Forensics Security 2025; 20:2836-51.

[29]	Tran HY, Hu J, Yin X, Pota HR . An efficient privacy—enhancing cross—silo federated learning and applications for false data injection attack detection in smart grids. IEEE Trans Inf Forensics Security 2023; 18:2538—52.