Detection of false data cyber-attacks for the assessment of security in smart grid using deep learning

doi:10.1016/j.compeleceng.2021.107211

Computers & Electrical Engineering

Volume 93, July 2021, 107211

https://doi.org/10.1016/j.compeleceng.2021.107211 Get rights and content

Abstract

Smart Grid uses electricity and information flows to set up a highly developed, fully automated, and distributed electricity grid system. To identify the reliability of work and availability, cyber attacks detection in the smart grids play a significant role. This paper highlights the integrity of false data cyber-attacks in the physical layers of smart grids. As the first contribution, the Proposed True Data Integrity provides an attack exposure metric through an Agent-Based Model. Next, the research focuses on the decentralization of Data Integrity Security in the system with an Agent-based approach. Finally, the productivity and efficiency of the developed modeling techniques are experimentally evaluated and compared with the existing state-of-the-art supervised deep-learning models. The obtained results of the studies have shown the improved false data detection accuracy of 98.19% through replay cyber-attacks using the Artificial Feed-forward Network. Based on the research findings, deep neural network can be used to assess cyber data in smart grids to detect malware incidents and attacks.

Introduction

Power grids are formed by the integration of electrical lines and other supporting devices to create a network. It has been used to transform a particular unit of energy for a network. In recent years, to enhance performance, management, planning, and other operative control, the smart grid uses information and communication technologies, and the new framework is called smart grids. These smart grids have to maintain a feature called Advanced Metering Infrastructure (AMI), and it is responsible for gathering and distributing data from the end-user to the service side [1]. NIST (National Institute of Science and Technology) [6] suggested that an advanced power grid comprises various fields like generating, transmitting, distributing, managing, and configuring information across the network [2]. The processes in each area incorporate devices that are meant to be primary and secondary. Electric energy and information are frequently used among smart grid domains, as depicted in Fig. 1. It shows the flow of processes made through the smart grid model. The energy grid devices responsible for power generation, transmission, and accumulation are primary devices. In contrast, end nodes involving intelligent electronic devices, servers, networking, and monitoring devices are secondary devices [3].

The smart power grid is said to be the riskiest and vulnerable cyber-attack. False Data Detection is a kind of cyber-attacks capable of manipulating the condition of the entire system. The current detection system fails to identify the False Data Detection Cyber-Attacks (FDDCA). The research work describes the process of detecting these types of new stealthy cyber-attacks, their effects, and how they are structured on the smart grid. Machine Learning (ML) [4] methods can classify secure measurements with the attacked measures. These have become popular in recent researches. To classify the protected data from the intruder, Artificial Feed-forward Network (AFN) has been proposed [5,7]. This Classification is achieved by considering the distance metric as the cost function in machine learning from high dimensional space. The cost function is used to classify the secured data from malicious data effectively. A complete performance evaluation is designed for a specific set of attacked data, and the results are validated and compared with other ML methods. Using an Artificial feed-forward network, the model can travel in a unidirectional manner, and sometimes it may have hidden layers. This positively supports the complex functions without any difficulty.

The objectives of the research work are mentioned below:

1
The grid's nature is analyzed in various fields: the type of stealthy attack in which the smart grid is vulnerable to an attack. Different methods and algorithms will retrieve data from the smart Data for applying in DL.
2
FDDCA and optimal adaptive load shedding are investigated by decentralizing the system integrity protection scheme with True Data Integrity by Agent-Based Model (TDI-ABM).
3
Deep Learning (DL) methods are applied for data retrieval from the network. False Data Detection (FDD) is to detect the attack in the smart grid.
4
The output will classify the secured data from the intruder data of the result from the neural network. The evaluation and analysis showed that the proposed TDI-ABM works well in identifying malicious cyber-attacks.

The remainder of the paper is organized as follows. Section 2 presents the related works of various methods that are vulnerable to an attack using the smart grid. Section 3 describes the fundamental process of smart grid power system and proposed deep learning model for cyber-attack identification. Section 4 portrays the False Data detection of cyber-attack, and section 5 describes False Data detection. Later, Section 6 presents a detailed study of the Test system used for experimental evaluation. Finally, Section 7 concludes the paper with summary of the work and future work directions.

Section snippets

Related works

In the new age Smart Grids, Supervisory Control and Data Acquisition (SCADA) framework is used to improve interconnection and manipulation necessities [8]. To manage operations and to perform transmission and distribution, the traditional power grid depends upon the SCADA framework as the technology gets effect into cybersecurity in smart cities. But till now, it has various limitations like lack of understanding, importance, and consistency. There are different types of cybersecurity attacks:

Fundamental process of smart grid power system

The reliability of the electric smart grid system in an area is maintained by Independent System Operators (ISOs) [19] and Regional Transmission Organisation (RTOs) [18]. These are proof for distribution, time management, detecting the output from equipment, power redirection during congestion, and intercommunication between nodes expand the framework. The reliability of the system is maintained by implementing independent strategies among organizations with another user in public. Control

False data detection of cyber attack (FDDCA)

FDDCA is similar to spoofing attacks that violate data integrity. Measurement false data is distributed and utilized in a different part of the grid. Hence, any change in the measurement data will damage the sensor. In a secured condition, the measurement vector 'i' is generated by the AFN model's sensor. The injected false value remains undetermined in the False Data Detection (FDD) model. As the attack vector is in the network, it spoils the operational decision and leads to more significant

Method of false data detection (FDD)

An autoencoder is responsible for generating data. Average and false data are classified by the anomaly detection module generated by the encoder. Autoencoder rectifies the error by calculating and analyzing the error. Various features are interrelated to each other, and the autoencoder learns these interrelations. The autoencoder trains the nature of the normal data. Once the autoencoder gets trained, it must be verified. Training starts from normal data, and malicious training is also

Test system

The preparation of the test system is described in this section. MATLAB is used for analysis and system estimation, which is mentioned on the IEEE 14-bus standard.

Conclusion and future work

The presented research on the smart grid demonstrates the need for false data detection to measure the potential and physical damaging effects of FDCA. The proposed model will benefit the smart grid systems under the linear smart power distribution model for grid actions. The utilization of the physical attributes developed by the proposed agent-based model enhanced the reliability of smart grids and proposed model is validated with the Cyber Attack replay. The agent-based approach reinvents

Declaration of Competing Interest

All authors declare no conflict of interest in the subject matter or materials discussed in this manuscript.

Acknowledgements

The authors gratefully acknowledge the Science and Engineering Research Board (SERB), Department of Science and Technology, India for the financial support through Mathematical Research Impact Centric Support (MATRICS) scheme (MTR/2019/000542). The authors also acknowledge SASTRA Deemed University, Thanjavur for extending infrastructural support to carry out this research work.

Author statement

All persons who meet authorship criteria are listed as authors, and all authors certify that they have participated sufficiently in the work to take public responsibility for the content, including participation in the concept, design, analysis, writing, or revision of the manuscript. Only persons who have made substantial contributions to the work reported in the manuscript are listed as authors.

Sudhakar Sengan is presently working as Professor in dept. of CSE at PSN College of Engineering and Technology, Tirunelveli, Tamil Nadu, India. He received Ph.D. degree in ICE in 2014 at Anna University, Chennai. He has 20 years of experience in Teaching/ Research/Industry. His research includes Network Security, Information Security, MANET, Cloud Computing, IoT. He published 100 Articles/Books in Journals and Conferences.

References (25)

Y. Jiang et al.
Blackstart capability planning for power system restoration
Int J Electr Power Energy Syst
(2017)
Peter Eder-Neuhauser et al.
Cyber-attack models for smart grid environments
Sustain Energy Grids Netw
(2017)
Daniel B Araya et al.
An ensemble learning framework for anomaly detection in building energy consumption
Energy Build
(2017)
Z.H. Yu et al.
Blind false data injection attack using PCA approximation method in smart grid
IEEE Trans Smart Grid
(2015)
Y. Mo et al.
Cyber-physical security of a smart grid infrastructure
Proc IEEE
(2012)
O. Stan et al.
A new crypto-classifier service for energy efficiency in smart cities
M. Raciti et al.
Embedded cyber-physical anomaly detection in smart meters
A. Giani et al.
Smart grid data integrity attacks
IEEE Trans. Smart Grid
(2013)
NIST framework and roadmap for smart grid interoperability standards
NIST Spec Publ
(2010)
T.J. Overbye et al.
The smart grid and PMUs: operational challenges and opportunities
IEEE Power Energy Soc Gen Meeting, PES
(2010)

A. Ashok et al.

Cyber-physical attack-resilient wide-area monitoring, protection, and control for the power grid

Proc IEEE

(2017)

ANSI C12.22-2008: Protocol Specification For Interfacing to Data Communication Networks (Draft)

(2008)

Cited by (56)

Multivariate time-series cyberattack detection in the distributed secondary control of AC microgrids with convolutional neural network autoencoder ensemble
2024, Sustainable Energy, Grids and Networks
This paper proposes an unsupervised machine learning-based approach for cyberattack detection in AC microgrids with distributed secondary control architecture. The proposed approach is fully unsupervised and only utilizes the system’s normal datasets for the training of the algorithm. The attack under study is a false data injection (FDI) attack tampering with the operating frequency of inverter-based distributed generators (DGs). The paper utilizes a 1D Convolutional Autoencoder (CAE) for cyberattack detection on a microgrid’s distributed secondary frequency control. An autoencoder is a neural network architecture, where the model is trained to reconstruct its input in an unsupervised manner. CAE can be applied to a time-series dataset to extract features and exploit the known correlation between neighboring temporal features. Due to the correlation between the operating frequency of DGs and their active power ratios, the paper uses the time series of these two variables as inputs to CAE. The effectiveness of the proposed approach has been verified using a simulated microgrid test system in Matlab/Simulink.
Smart grid cyberattack types classification: A fine tree bagging-based ensemble learning approach with feature selection
2024, Sustainable Energy, Grids and Networks
This paper focuses on the detection and classification of the cyberattack types in smart grid substation automation systems. The previous work in the literature focuses only on the detection of the attacks without providing any information regarding the attack’s type, which is a key in identifying the appropriate countermeasures. In this paper, a novel approach that uses a fine tree bagging ensemble learning technique is developed to detect and classify the cyberattack types from normal and power quality disturbances. Furthermore, the relevant features of different cyber-attack types such as message suppression, denial-of-service and data manipulation have been identified. The proposed approach is tested on a publicly available dataset and the results are compared to three other machine learning algorithms, namely decision tree, nearest neighbor, and support vector machine. The results have shown that the proposed approach is very effective in the detection and the classification of the attack types as well as it is insensitive to the selection of the training and the testing datasets unlike other existing approaches in the literature.
Cyberattack detection for electricity theft in smart grids via stacking ensemble GRU optimization algorithm using federated learning framework
2024, International Journal of Electrical Power and Energy Systems
Recently, the detection difficulty of emerging power grid cyberattacks for electricity theft far surpasses that of traditional physical power theft behavior. Reliable cyberattack detection is of great importance for ensuring stable smart grid operation and protecting the interests of power supply companies. Recently, a few studies have demonstrated acceptable performance in detecting individual or minority attacks using machine learning. However, difficulties such as poor data preprocessing, unwarranted feature selection, and deficiency of joint attack detection have limited their reliability. To unravel these issues, this study uses a long short-term memory (LSTM)-based multi-step prediction model to forecast and fill in missing data. Moreover, a series of statistical methods are utilized to explore optimal feature selection by analyzing the significance, correlation, and applicability between normal and attack feature data. Furthermore, a federated-learning-based stacking ensemble gate recurrent unit algorithm (FL-SE-GRU) is proposed for electricity theft detection and cyberattack classification, and its effectiveness is verified by comparison with existing methods. The results of experiments show that the LSTM-based multi-step prediction model exhibits a remarkable data interpolation effect. Meanwhile, FL-SE-GRU achieves 95.0% accuracy, 96.6% precision, 93.8% sensitivity, and 95.1% F1 score in detecting electricity theft, and reaches 96.8% accuracy, 96.1% precision, 97.4% sensitivity, and 96.7% F1 score in classifying 9 kinds of cyberattacks, respectively. This study provides unique perspectives to cope with the increasingly complicated cyberattacks for electricity theft.
Age-appropriate design of smart senior care product APP interface based on deep learning
2024, Heliyon
With the aging of the population, the quality of life and happiness of the elderly are increasingly becoming concerns of society. Smart senior care (SSC) products are an important tool to improve the quality of life of the elderly, and their application has been widely discussed. However, due to differences in cognitive characteristics and habits of the elderly, they may face difficulties when using smart products. Therefore, how to design a suitable SSC product APP interface for the elderly has become an urgent problem to be solved. The rapid development of deep learning (DL) provides a new way to deal with this problem. This paper aims to optimize the APP interface design of SSC products suitable for the elderly by using DL technology. Specifically, this paper designs a model based on a deep-Q-network (DQN) algorithm, which can better adapt to the usage habits and needs of the elderly through training and optimization. At the same time, the proposed model is evaluated comprehensively to verify its performance and superiority. To achieve the above goals, this paper first summarizes the existing SSC technology and the design suitable for the elderly. On this basis, a model based on the DQN algorithm is designed, and four datasets are used for training and testing. In the design of the model, this paper pays special attention to how to make the interface more friendly and easy to operate to meet the specific demands and preferences of the elderly. After a comprehensive evaluation, it is found that the proposed DQN algorithm model has achieved remarkable improvement in performance. This model has an average return of about 8.8 and 3.9 compared to other models, which have an average return of about 0.9 and 0.2, respectively. This shows that the model designed in this paper performs well in terms of accuracy and average return and performs better than other algorithms. The results of this paper reveal that optimizing the age-appropriate design of the SSC product APP interface through DL technology can notably enhance the user experience and satisfaction of the elderly. This not only helps the elderly to make better use of smart products and improve their quality of life but also provides useful inspiration and guidance for research and practice in related fields.
Cyberattack patterns in blockchain-based communication networks for distributed renewable energy systems: A study on big datasets
2024, Data in Brief
Blockchain-based reliable, resilient, and secure communication for Distributed Energy Resources (DERs) is essential in Smart Grid (SG). The Solana blockchain, due to its high stability, scalability, and throughput, along with low latency, is envisioned to enhance the reliability, resilience, and security of DERs in SGs. This paper presents big datasets focusing on SQL Injection, Spoofing, and Man-in-the-Middle (MitM) cyberattacks, which have been collected from Solana blockchain-based Industrial Wireless Sensor Networks (IWSNs) for events monitoring and control in DERs. The datasets provided include both raw (unprocessed) and refined (processed) data, which highlight distinct trends in cyberattacks in DERs. These distinctive patterns demonstrate problems like superfluous mass data generation, transmitting invalid packets, sending deceptive data packets, heavily using network bandwidth, rerouting, causing memory overflow, overheads, and creating high latency. These issues result in ineffective real-time events monitoring and control of DERs in SGs. The thorough nature of these datasets is expected to play a crucial role in identifying and mitigating a wide range of cyberattacks across different smart grid applications.
Digital twins for sustainable design and management of smart city buildings and municipal infrastructure
2024, Sustainable Energy Technologies and Assessments
In the context of sustainable energy, exemplified by smart grids, bidirectional information transmission unfolds as a key feature. The deep integration in this data-driven model poses heightened risks to infrastructure while enhancing control over automation processes. A significant concern is the false data injection attack (FDIA), which intricately targets sustainable energy system state estimation. This attack allows adversaries to bypass detection methods, manipulate outcomes, and jeopardize secure grid operation. This study explores FDIA identification in smart grids, focusing on Recurrent Neural Networks (RNNs) for detection. The objective is to assess the proposed detection methodology through simulation experiments on the IEEE standard test scheme, validating practicality under diverse sustainable energy conditions. This study undertakes a focused exploration into the realm of identifying FDIAs within smart grids, centering on the utilization of Recurrent Neural Nets (RNNs) as a machine learning approach for detection. The objective is to assess the efficacy of the proposed detection methodology through rigorous imitation experiments conducted on the IEEE standard test scheme. Various contrasting scenarios are meticulously designed to validate the practicality and effectiveness of the detection model under diverse conditions. In the intricate landscape of Fault Detection, Isolation, and Diagnosis in smart grids, the imperative is to recognize and comprehend abnormal conditions or faults within the grid's operational framework. A robust strategy poised to detect FDIAs in Smart Grids involves the strategic implementation of an advanced digital twin known as the “Cyber-Physical Digital Twin for Smart Grid Security.” This digital twin is meticulously engineered to emulate the intricate dynamics of the Smart Grid, seamlessly integrating both its physical and cyber elements.

View all citing articles on Scopus

Subramaniyaswamy V is currently working as an Associate Professor in the SASTRA Deemed University, Thanjavur, India. In total, he has 16 years of experience in research and academia. He has published more than 125 papers in reputed international journals and conferences. His research interests include artificial intelligence, recommender systems, machine learning, and Internet of Things.

Indragandhi V is serving as an Associate Professor in the School of Electrical Engineering, VIT, Vellore, India. She published more than 100 research articles in reputed international journals and top conferences. She authored a book which is published by Elsevier, Academic Press. Her research interests include Power Electronics, Soft computing Techniques and Smart Grid.

Priya Velayutham received her Ph.D. degree in Information and Communication Engineering in 2017 at Anna University, Chennai. She is currently working as an Associate Professor in Computer Science and Engineering at Paavai Engineering College, Namakkal, Tamilnadu, India. She has more than 12 years of experience in Teaching and Research. She published many articles in reputed International Journals.

Logesh Ravi is an Associate Professor at Department of Computer Science and Engineering, Vel Tech Rangarajan Dr. Sagunthala R&D Institute of Science and Technology, Chennai, India. Dr. Logesh Ravi has published more than 90 papers in reputed international journals and conferences. His research interests include artificial intelligence, recommender systems, big data, machine learning, information retrieval, fintech and social computing.

^☆: This paper is for special section VSI-gridl. Reviews processed and recommended for publication by Guest Editor Dr. G. Jeon.

View full text

Detection of false data cyber-attacks for the assessment of security in smart grid using deep learning☆

Abstract

Introduction

Section snippets

Related works

Fundamental process of smart grid power system

False data detection of cyber attack (FDDCA)

Method of false data detection (FDD)

Test system

Conclusion and future work

Declaration of Competing Interest

Acknowledgements

Author statement

Int J Electr Power Energy Syst

Sustain Energy Grids Netw

Energy Build

Blind false data injection attack using PCA approximation method in smart grid

IEEE Trans Smart Grid

Cyber-physical security of a smart grid infrastructure

Proc IEEE

A new crypto-classifier service for energy efficiency in smart cities

Embedded cyber-physical anomaly detection in smart meters

Smart grid data integrity attacks

IEEE Trans. Smart Grid

NIST framework and roadmap for smart grid interoperability standards

NIST Spec Publ

The smart grid and PMUs: operational challenges and opportunities

IEEE Power Energy Soc Gen Meeting, PES

Cyber-physical attack-resilient wide-area monitoring, protection, and control for the power grid

Proc IEEE

ANSI C12.22-2008: Protocol Specification For Interfacing to Data Communication Networks (Draft)