False Data Injection Attack Detection Method against PMU Measurements

doi:10.3969/j.issn.1671-1122.2022.05.009

Abstract

Abstract:

A novel unsupervised online learning detection method was proposed for false data injection attack detection of PMU measurements, which was called corrected robust random cut forest (CRRCF). Firstly, RRCF was an unsupervised online-learning algorithm, which could quickly adapt to the PMU measurement after the topology change and generate abnormal scores to reflect the abnormal degree of samples. Secondly, according to the abnormal scores of RRCF, CRRCF used Gaussian Q function and sliding window to calculatethe abnormal probability. Thirdly, the abnormal probability modified the judgment of abnormal degree from RRCF and adapted to changes of attack number and attack magnitude. The simulation results show that compared with the static learning method, the online learning method can solve the problem of concept drift caused by topology changes, while compared with other online learning methods, CRRCF can always maintain higher detection accuracy and F1 score when the attack number and the attack magnitude change.

Key words: false data injection attacks, PMU, online learning, anomaly detection

CLC Number:

TP309

ZHOU Jingyi, LI Hongjiao. False Data Injection Attack Detection Method against PMU Measurements[J]. Netinfo Security, 2022, 22(5): 75-83.

Figures/Tables 13

References 25

[1]	LIU Andi, LI Yan, XIE Wei, et al. Line Parameter Estimation Method of Distribution Network Based on Multi Source Data and Multi Time Section[J]. Automation of Electric Power Systems, 2021, 45(2): 46-54.
	刘安迪, 李妍, 谢伟, 等. 基于多源数据多时间断面的配电网线路参数估计方法[J]. 电力系统自动化, 2021, 45(2):46-54.
[2]	YUAN Zhiyong, XIAO Zekun, YU Li, et al. Overview of Smart Grid Big Data Research[J]. Guangdong Electric Power, 2021, 34(1): 1-12.
	袁智勇, 肖泽坤, 于力, 等. 智能电网大数据研究综述[J]. 广东电力, 2021, 34(1):1-12.
[3]	LIU Shengyuan, LIN Zhenzhi, LI Jincheng, et al. Overview and Prospect ofPower System Situation Awareness Technology[J]. Automation of Electric Power Systems, 2020, 44(3): 229-239.
	刘晟源, 林振智, 李金城, 等. 电力系统态势感知技术研究综述与展望[J]. 电力系统自动化, 2020, 44(3):229-239.
[4]	YANG Peng. “Cloud Management” Works at the Same Time and China Mobile's 5g Smart Grid will Take “Three Steps”[J]. Communications World, 2020(23): 18-19.
[5]	LIANG Gaoqi, WELLER S R, ZHAO Junhua, et al. The 2015 Ukraine Blackout: Implications for False Data Injection Attacks[J]. IEEE Transactions on Power Systems, 2017, 32(4): 3317-3318. doi: 10.1109/TPWRS.2016.2631891 URL
[6]	BINDRA A. Securing the Power Grid: Protecting Smart Grids and Connected Power Systems from Cyberattacks[J]. IEEE Power Electronics Magazine, 2017, 4(3): 20-27.
[7]	LIU Yao, PENG Ning, REITER M K. False Data Injection Attacks against State Estimation in Electric Power Grids[J]. ACM Transactions on Information and System Security, 2011, 14(1): 1-33.
[8]	CHU Zhigang, ZHANG Jiazi, KOSUT O, et al. Unobservable False Data Injection Attacks against PMUs: Feasible Conditions and Multiplicative Attacks[C]// IEEE. 2018 IEEE International Conference on Communications, Control, and Computing Technologies for Smart Grids. New York: IEEE, 2018: 1-6.
[9]	YU J J Q, HOU Yunhe, LI V O K. Online False Data Injection Attack Detection with Wavelet Transform and Deep Neural Networks[J]. IEEE Transactions on Industrial Informatics, 2018, 14(7): 3271-3280. doi: 10.1109/TII.2018.2825243 URL
[10]	BASUMALLIK S, MA Rui, EFTEKHARNEJAD S. Packet-Data Anomaly Detection in PMU-Based State Estimator Using Convolutional Neural Network[J]. International Journal of Electrical Power & Energy Systems, 2019, 107: 690-702. doi: 10.1016/j.ijepes.2018.11.013 URL
[11]	KHARE G, MOHAPATR A A, SINGH S N. A Real-Time Approach for Detection and Correction of False Data in PMU Measurements[J]. Electric Power Systems Research, 2021, 191(2): 1-11.
[12]	AHMED S, LEE Y D, SEUNG H H, et al. Unsupervised Machine Learning-Based Detection of Covert Data Integrity Assault in Smart Grid Networks Utilizing Isolation Forest[J]. IEEE Transactions on Information Forensics and Security, 2019, 14(10): 1-13.
[13]	MOHAMMADPOURFARD M, SAMI A, WENG Yang. Identification of False Data Injection Attacks with Considering the Impact of Wind Generation and Topology Reconfigurations[J]. IEEE Transactions on Sustainable Energy, 2018, 9(3): 1349-1364. doi: 10.1109/TSTE.2017.2782090 URL
[14]	YAN Jun, TANG Bo, HE Haibo. Detection of False Data Attacks in Smart Grid with Supervised Learning[C]// IEEE. 2016 International Joint Conference on Neural Networks (IJCNN). New York: IEEE. 2016: 1395-1402.
[15]	GUHA S, MISHRA N, ROY G, et al. Robust Random Cut Forest Based Anomaly Detection on Streams[C]// JMLR. The 33rd International Conference on Machine Learning. New York: JMLR, 2016: 2712-2721.
[16]	BOUANANI F E, MOUCHTAK Y, KARAGIANNIDIS G K. New Tight Bounds for the Gaussian Q-Function and Applications[J]. IEEE Access, 2020, 8: 145037-145055. doi: 10.1109/ACCESS.2020.3015344 URL
[17]	AHMAD S, LAVIN A, PURDY S, et al. Unsupervised Real-Time Anomaly Detection for Streaming Data[J]. Neurocomputing, 2017, 262(1): 134-147. doi: 10.1016/j.neucom.2017.04.070 URL
[18]	TENG S Y, Máša V, TOU M, et al. Waste-to-Energy Forecasting and Real-Time Optimization: An Anomaly-Aware Approach[J]. Renewable Energy, 2022, 181: 142-155. doi: 10.1016/j.renene.2021.09.026 URL
[19]	ANGIULLI F, PIZZUTI C. Fast Outlier Detection in High Dimensional Spaces[C]// ACM. 6th European Conference on Principles of Data Mining and Knowledge Discovery, New York: ACM, 2002: 15-27.
[20]	LIU F T, MING T K, ZHOU Z H. Isolation-Based Anomaly Detection[J]. ACM Transactions on Knowledge Discovery from Data, 2012, 6(1): 1-39.
[21]	PEVNY T. Loda: Lightweight Online Detector of Anomalies[J]. Machine Learning, 2016, 102(2): 275-304. doi: 10.1007/s10994-015-5521-0 URL
[22]	SATHE S, AGGARWAL C C. Subspace Outlier Detection in Linear Time with Randomized Hashing[C]// IEEE. 2016 IEEE 16th International Conference on Data Mining (ICDM). New York: IEEE, 2016: 459-468.
[23]	LIU Yuxiao, WANG Yi, ZHANG Ning, et al. A Data-Driven Approach to Linearize Power Flow Equations Considering Measurement Noise[J]. IEEE Transactions on Smart Grid, 2020, 11(3): 2576-2587. doi: 10.1109/TSG.2019.2957799 URL
[24]	KONAKALLA S A R, CALLAFON R A D. Spectrum-Based Optimal Filtering for Short-Term Phasor Data Prediction[J]. IEEE Transactions on Industry Applications, 2020, 56(2): 2069-2077. doi: 10.1109/TIA.2020.2966186 URL
[25]	BROWN M, BISWAL M, BRAHMA S, et al. Characterizing and Quantifying Noise in PMU Data[C]// IEEE. 2016 IEEE Power and Energy Society General Meeting(PESGM). New York: IEEE, 2016: 1-5.

序号	步骤
1	如果T(X)=φ,返回仅含节点x(t)的树$\text{ }\!\!~\!\!\text{ }T\left( \text{ }\!\!\{\!\!\text{ }\mathbf{x}(t)\} \right)$
2	否则,选择一个随机数$r'$进行切割
3	如果该切割将树T(X)和x(t)分开,则作为$T\left( X+\text{ }\!\!\{\!\!\text{ }\mathbf{x}(t)\} \right)$的第一次切割,否则删除该切割。选择与T(X )中相同的切割,划分子集X₁和X₂
4	设$x(t)\in {{X}_{2}}$,子集X₂重复上述步骤,子集X₁使用和T(X)中相同的子树

序号	步骤
1	在树 T(X )中查询样本点x(p)对应的叶子节点 R
2	删除 R 的父节点,并用 R 的同级节点替换该父节点

学习模式	方法	精度
学习模式	方法	拓扑未更改	拓扑更改
在线学习	CRRCF	0.983	0.984
在线学习	RRCF	0.955	0.961
在线学习	RSHash	0.963	0.968
静态学习	AvgKNN	0.995	0.548
静态学习	LODA	0.983	0.543
静态学习	IForest	0.967	0.533
静态学习	β-VAE	0.996	0.548

学习模式	方法	F1分数
学习模式	方法	拓扑未更改	拓扑更改
在线学习	CRRCF	0.908	0.914
在线学习	RRCF	0.810	0.831
在线学习	RSHash	0.776	0.809
静态学习	AvgKNN	0.974	0.301
静态学习	LODA	0.913	0.295
静态学习	IForest	0.835	0.281
静态学习	β-VAE	0.98	0.304

学习模式	方法	精度
学习模式	方法	拓扑未更改	拓扑更改
在线学习	CRRCF	0.983	0.984
在线学习	RRCF	0.955	0.962
在线学习	RSHash	0.963	0.955
静态学习	AvgKNN	0.995	0.548
静态学习	LODA	0.983	0.542
静态学习	IForest	0.967	0.532
静态学习	β-VAE	0.996	0.548