You are on page 1of 5

Materials Today: Proceedings xxx (xxxx) xxx

Contents lists available at ScienceDirect

Materials Today: Proceedings


journal homepage: www.elsevier.com/locate/matpr

A review on security analysis of cyber physical systems using Machine


learning
Alshaibi Ahmed Jamal, Al-Ani Mustafa Majid, Anton Konev, Tatiana Kosachenko, Alexander Shelupanov
Department of Complex Information Security of Computer Systems, Faculty of Security, Tomsk State University of Control Systems and Radioelectronics, Tomsk 634050,
Russian Federation

a r t i c l e i n f o a b s t r a c t

Article history: The concept of Cyber Physical System (CPS) is widely used in different industries across the globe. In fact,
Available online xxxx it is the holistic approach towards dealing with cyber space and physical environments that do have
inter-dependencies. In the existing systems, there was a separate approach for security of the two worlds
Keywords: (cyber and physical). However, it does not provide necessary security when security is employed inde-
Cyber security pendently. The research in this paper identifies the need for integrated security for CPS. Besides it throws
Intrusion detection light into different security challenges associated with CPS and the countermeasures that existed based
Cyber threat intelligence
on machine learning and deep learning techniques that come under Artificial Intelligence (AI) and data
Machine learning
science. From the review of literature, it is understood that data science perspective is suitable for pro-
tecting CPS with required adaptive strategy. This paper provides several useful insights related to security
analysis of CPS using machine learning. It paves way for further investigation and realize a comprehen-
sive security framework to protect CPS from internal and external cyber-attacks.
Ó 2021 Elsevier Ltd. All rights reserved.
Selection and peer-review under responsibility of the scientific committee of the International Confer-
ence on Nanoelectronics, Nanophotonics, Nanomaterials, Nanobioscience & Nanotechnology.

1. Main text cyber-physical security. They focused on cyber-physical security


to electric vehicles and developed a set of index-based metrics to
Cyber-Physical Systems (CPS) is a new approach in looking at measure performance degradation of such systems due to cyber-
digital systems. In fact, ‘‘CPS is a new generation of digital system, physical attacks. They proposed and implemented a simulation
mainly focuses on complex interdependencies and integration that provides behaviour of a human driver and investigated the
between cyberspace and physical world”. CPS has an integrated impact of malicious attacks on such vehicles. They explored on
approach towards communication systems, computing resources, detection of replay and Denial of Service (DoS) attacks.
controls and physical environments. It has network of embedded Cyber physical system designs witnessed different kinds of
computers used for monitoring physical processes that will have disciplines to be used including sensor networks, electronics, con-
impact on computations and vice versa [1]. In such environment, trol engineering, software engineering, error handling, physical
cyber-attacks might cause potential risks to underlying systems. dynamics, machine learning and thermos dynamics besides using
With respect to CPS security design conventional approaches the aspect based modelling that separates cross cutting concerns [5].
physical and cyber systems separately and could not address vul- As explored in Wang et al. [6] different machine learning algo-
nerabilities that are associated with networks and embedded con- rithms towards artificial intelligence (AI) such as K-Nearest Neigh-
trollers that are meant for monitoring and controlling physical bour (KNN), Support Vector Machines (SVM) and Deep Neural
processes [2,3]. Therefore, there is need for an integrated security Networks (DNN) are used to prevent DoS attacks in CPS. Code
approach for protecting CPS from cyber-attacks. In the literature, mutation approaches are studied in Chen et al. [7] to learn different
there is substantial evidence on the need for security to such variants of CPS. In Settanni et al. [8], it is argued that CPS can have
systems and the havoc that may cause if the security is ignored. automated Machine Learning (ML) approaches in such a way that
Guo et al. [4] opined that there is increasing prominence to they can have self-adaptive mechanisms to deal with anomalies.
In Babiceanu and Seker [9], it is understood that the data associ-
ated with CPS is the big data and there is need for handling large
E-mail address: ahmed.jamal.alshaibi88@gmail.com (A. Ahmed Jamal) volumes of data in assessing security risks. In some CPS use cases

https://doi.org/10.1016/j.matpr.2021.06.320
2214-7853/Ó 2021 Elsevier Ltd. All rights reserved.
Selection and peer-review under responsibility of the scientific committee of the International Conference on Nanoelectronics, Nanophotonics, Nanomaterials,
Nanobioscience & Nanotechnology.

Please cite this article as: A. Ahmed Jamal, Al-Ani Mustafa Majid, A. Konev et al., A review on security analysis of cyber physical systems using Machine
learning, Materials Today: Proceedings, https://doi.org/10.1016/j.matpr.2021.06.320
A. Ahmed Jamal, Al-Ani Mustafa Majid, A. Konev et al. Materials Today: Proceedings xxx (xxxx) xxx

such as the one described in Machado and Medeiros Frohlich [10] used for cybersecurity research. The datasets include KDD’99 Cup
with Internet of Things (IoT) integration used blockchain technol- [31], NSL-KDD [32], CAIDA [33], ISOT’10 [34], ISCX’12 [35], CTU-
ogy for security. Without integration with a security framework, 13 [36], UNSW-NB15 [37], CIC-IDS2018 and CIC-IDS2017 [38],
CPS developments lead to potential risk [11]. In order to have effec- MAWI [39], CERT [40], DGA [41], Malware [42] and Bot-IoT [43].
tive security to CPS, there needs to be technological characteristics
such as intelligence, integration and cooperation coupled with
4. Deep learning for intrusion detection system
security measures [12]. Research challenges in different domains
of CPS are provided in Allgöwer et al. [13]. Hassan et al. [14]
Deep learning has emerged to have high impact on the indus-
employed differential privacy attacks to protect privacy issues
tries with improved machine learning capabilities to solve prob-
associated with CPS.
lems in the real world. Vinayakumar et al. [1] explored deep
The research carried out in [15–25] also focused on different
learning techniques such as Convolutional Neural Network (CNN)
dimensions of CPS including the usage of machine learning to have
to build an intrusion detection model for preventing cyber-
intelligence on cyberattacks and preventing them. In this paper,
attacks. They made a hybrid intrusion detection framework to sup-
our contributions are as follows. We investigated on CPS and the
port both host based and network based intrusion detection. Their
need for integrated security to such systems in the real world
system was found scalable but it needs computing resources as
besides security challenges and countermeasures using machine
CNN needs high processing power. They used dataset created out
learning approaches. This paper also throws light on research gaps
of a distributed system in order to make experiments with respect
that need further study in the area of security to CPS. Tis paper is
to detection of cyber-attacks. They used different datasets for
structured as follows. Section 2 reviews different security chal-
empirical study including KDDCup 99, NSL-KDD and UNSW-NB15
lenges associated with CPS [58–64]. Section 3 talks about the
to examine different attacks categories such as DoS, U2R, R2L
importance of ML towards data science. Section 4 covers deep
and probe. They also examined attacks such as flooding, grayhole
learning for serving as an Intrusion Detection Systems (IDSs)
and blackhole.
[65–67]. Section 5 covers different ML approaches used in CPS
for leveraging security. Section 6 covers the research gaps found
in the research. Section 7 concludes the paper and gives directions 5. Machine learning for security analysis of cyber physical
for future work. systems

As discussed in the preceding sections, data science with its


2. Emerging security challenges
machine learning and deep learning techniques is becoming an
important contributor to automatic analysis of cyber physical sys-
Embedded controllers in physical environments and digital sys-
tems to enable continuous monitoring for enhancing security mea-
tem (cyberspace) are part of CPS. However, this integrated system
sures [44]. Kholidy [45] proposed a system with underlying
is subjected to various kinds of attacks. Different security chal-
security framework for autonomous mitigation of cyberattacks
lenges are studied by Choo et al. [26] and the attacks are both
on CPS. It integrates traditional IDS such as SNORT and ML based
internal and external in nature. The potential challenges are asso-
approaches in order to have better security. It has consideration
ciated with different systems such as CPS associated with health-
for existing security approaches and risk assessment models to
care and IoT, embedded systems and green cryptography, smart
realize the new Autonomous Response Controller (ARC) frame-
building security, privacy issues in CPS, trust and security in CPS,
work. They used a probabilistic risk assessment model that is that
Emerging security schemes for embedded security, sensitive data
assesses potential risk and make appropriate decisions. Lv et al.
handling in CPS and other open challenges due to technology
[46] explored AI for improving security of CPS. They considered a
diversity involved in IoT integrated CPS. Time synchronization
‘‘CPS-based indoor environment measurement and control system
attacks in CPS were investigated in Wang et al. [27]. As studied
for intelligent buildings” as case study. They built a multi-agent
in Ding et al. [28] different attacks are identified in CPS environ-
system that has different components such as detection, control,
ments. They are known as DoS, replay and deception attacks. Mah-
execution and communication. For classification and regression
moud et al. [29] identified different attacks associated with CPS.
purposes, they used Back-Propagation Neural Network (BPNN).
They include jamming attack, DoS attack, replay attack and decep-
The system is evaluated with different performance metrics such
tion attack.
as Mean Absolute Error (MAE), NRMSE and PSNR and found its
superior performance with least false positives.
3. Emerging security challenges Tantawy et al. [2] proposed an integrated approach for CPS
security with model-based approach. In a real world industry, they
Sarker et al. [30] state that cybersecurity is undergoing techno- used Continuous Stirred Tank Reactor (CSTR) and its integrated
logical shift of late and data science is playing crucial role in driv- with cyberspace for their study on security. The CPS test bed used
ing the change. As data is growing pertaining to cybersecurity for the experiments contains corporate network and control net-
incidents, data science came into picture to figure out the attacks work besides data exchange for machine learning based monitor-
based on analysis of data. As there are incidents growing, the level ing of cyber-attacks. The system has different components such
of training data is increasing thus contributing to the automation as plant, control logic, monitoring system, corporate analysis, and
of the monitoring of intrusions in cyber space with required intel- safety logic. A hybrid automaton is formulated in order to identify
ligence. Towards this end, machine learning plays vital role as it cyberattacks. A tree based approach is used in order to hold data
garners actionable knowledge from different cybersecurity data and use it effectively for identification of attacks. Salahdine and
sources. As studied in Sarker et al. [30], it is understood that there Kaabouch [47] explored cybersecurity threats, vulnerabilities, and
has been increase in the spread of IoT use cases from time to time. countermeasures with respect to physical layer of Cognitive Radio
In distributed environments, IoT use cases with thousands of con- Networks (CRN). They focused on different physical layer attacks
nected and resource constrained devices, throw security chal- associated with dynamic spectrum access, belief manipulation,
lenges. In such context, there is need for an automated and eavesdropping and malicious traffic injection. They have also dis-
intelligent means of monitoring cyber space in order to have timely cussed about detection methods such as compressive sensing,
action when situations arise. They also studied different datasets learning based approaches, IDS model, feature based approach,
2
A. Ahmed Jamal, Al-Ani Mustafa Majid, A. Konev et al. Materials Today: Proceedings xxx (xxxx) xxx

Table 1
Methods and papers on machine learning based IDSs.

Reference Technique Advantages Limitations Datasets


Kholidy [45] Autonomous Response Controller (ARC) Able to mitigate attacks such as Limited data is considered. Industrial cyber-attack
framework. Aurora. dataset
Lv et al. [46] Multi-Agent System (MAS) with Back- High effectiveness and robustness Needs further research to Data from construction
Propagation Neural Network (BPNN) of security to CPS. improve the system. industry.
Tantawy et al. Hybrid automaton based approach. Security and cost-effectiveness Further integration of model Data of Continuous Stirred
[2] based approach to ML is desired. Tank Reactor (CSTR) testbed.
Salahdine and Survey on countermeasures for physical Insights on different kinds of There is no empirical approach. –
Kaabouch layer attacks on CRNs attacks.
[47]
Sargolzaei et al. Neural network based fault detection Improved reliability, safety and Research is limited to fault Vehicular CPS dataset.
[48] system robustness of CPS. detection.
Goh et al. [49] Unsupervised learning for anomaly Low false positives. Feature selection is to be SWaT dataset
detection (LSTM-RNN) optimized.
Wang et al. [27] FDML classifier using ANN Detection of cyber-attacks and Detection convergence analysis is NYISO dataset
time synchronization (TS) attacks. yet to be explored.
Wu et al. [50] Anomaly detection, KNN and random Detection of attacks on 3D printing A novel security framework is yet Data of cyber manufacturing
forest. and CNC milling desired. system.
Adepu et al. Electrical Power and Intelligent Control Enables security research with EPIC Limited to domain specific Electric Power CPS dataset
[51] (EPIC) testbed. attacks.

data assisted approach, localization based method, belief propaga- tect Cyber Manufacturing systems from different kinds of attacks.
tion method and spectrum sensing techniques. The PU emulation As discussed in Chong et al. [55], there is need for characterization
attacks are categorized into cryptography based approaches, fin- of privacy and security attacks on CPS. The work in Mahmoud et al.
gerprint based techniques, game theory based solutions and hybrid [29] suggested that there is need for further research to use an
approves. adaptive approach that considers monitoring multiple cyber-
Sargolzaei et al. [48] considered a vehicular CPS case study and attacks at the same time even in presence of resource constraints.
explored the usage of machine learning for fault detection. It It also needs to consider coordinated cyber-attacks. In case of IoT
focused on False Data Injection (FDI) attack that causes accidents use cases, securing control systems that are part of CPS is yet
in vehicular networks. They proposed a methodology that detects another direction that needs further focus. The research in Jiang
FDI attacks and takes corrective steps in order to produce correct et al. [56] came up with many future possibilities such as data-
signal so as to help drive or controller to know safe distance driven fault tolerant controllers, fault localization, deep learning
between his/her and preceding vehicle. It makes use of a neural based solution to prevent false data injection attacks and attacks
network based approach for fault detection. Goh et al. [49] used such as low sparsity unobservable attack. Delicato et al. [57] inves-
unsupervised learning technique using deep learning by employ- tigated on the CPS and security challenges and opined to have
ing LSTM-RNN to model the system in order to detect abnormal smart CPS with AI that can ensure protection from cyber-attacks.
behaviour. There are two variants of time synchronization attacks
that make use of GPS to influence a CPS. They are known as Time
Synchronization (TS) attack and Stealth Time Synchronization 7. Conclusion and future work
(STS) attack. Wang et al. [27] explored ML techniques to prevent
such attacks. They proposed an Artificial Neural Network (ANN) This paper reviews the state of the art approaches covering dif-
based detection model that leveraged performance in detecting ferent kinds of CPS and different means of security analysis using
the attacks. As well as the authors, the type of the Technique, the ML approaches and deep learning as well. There are different
study Advantages and Limitations, and the dataset of each identi- attacks found in the literature that need to be dealt with in a
fied work in Table 1. CPS. They include replay attack, DoS attack, jamming attacks, time
synchronization attack, stealth time synchronization attack, false
data injection attack and so on.
6. Research gaps There are different countermeasures found in the literature.
Machine learning techniques like KNN, SVM, CNN, LSTM-RNN,
A typical CPS has many layers in its architecture as explored in BPNN, ANN etc. were used in the research. However, it is under-
Liu et al. [52]. Security is essentially a cross-layer perspective [53] stood from the research that ML techniques are indispensable
as different layers are involved in CPS. However, the integration of due to the bulk of data that needs to be processed in order to detect
the cyberspace and the physical environment has vulnerabilities different kinds of attacks. This paper is the first step in understand-
that cause different kinds of attacks. As studied in [27,48,50,54] ing CPS and its security aspects. The insights of the research in this
machine learning techniques are very useful for detecting attacks paper paves way for the work in future. A specific direction
automatically in CPS provided a purposefully built security frame- towards it is to have a comprehensive AI based security framework
work. Tantawy et al. [2] opined that there is need for more usage of that protects CPS from cyber-attacks.
ML techniques integrated with the CPS design in an integrated
approach for automating security that otherwise not possible
when security is employed to cyber space and physical environ- Funding
ment separately. This is an invaluable insight that warrants further
research. Settanni et al. [8] suggested the usage of self-adaptation The article was prepared as part of implantation of the «Leading
for enabling CPS to automate the anomaly detection process. This research center (LRC) «Trusted Sensor Systems», financial support
is an interesting insight that leverages system for autonomous provided by Ministry of Digital Development, Communications
CPS with self-adaptation. The work in Wu et al. [50] advocated and Mass Media of the Russian Federation and Russian Venture
the importance of advanced machine learning techniques to pro- Company (RVC JSC) (Agreement 009/20 dated 04/10/2020).
3
A. Ahmed Jamal, Al-Ani Mustafa Majid, A. Konev et al. Materials Today: Proceedings xxx (xxxx) xxx

Declaration of Competing Interest [23] R. Gifty, R. Bharathi, P. Krishnakumar, Privacy and security of big data in cyber
physical systems using Weibull distribution-based intrusion detection, Neural
Comput. Appl. 31 (S1) (2019) 23–34.
The authors declare that they have no known competing finan- [24] R. Alguliyev, Y. Imamverdiyev, L. Sukhostat, Cyber-physical systems and their
cial interests or personal relationships that could have appeared security issues, Comput. Ind. 100 (2018) 212–223.
[25] T. Dreossi, A. Donzé, S.A. Seshia, Compositional falsification of cyber-physical
to influence the work reported in this paper.
systems with machine learning components, J. Automated Reasoning 63 (4)
(2019) 1031–1053.
[26] K.-K. Choo, M.M. Kermani, R. Azarderakhsh, M. Govindarasu, Emerging
Acknowledgements embedded and cyber physical system security challenges and innovations,
IEEE Trans. Dependable Secure Comput. 14 (3) (2017) 235–236.
[27] J. Wang, W. Tu, L.C.K. Hui, S.M. Yiu, E.K. Wang, Detecting Time Synchronization
The authors would like to thank Tomsk State University of Con- Attacks in Cyber-Physical Systems with Machine Learning Techniques, 2017
trol Systems And Radioelectronics (www.tusur.ru/en) Tomsk - Rus- IEEE 37th International Conference on Distributed Computing Systems (ICDCS)
, 2017.
sian Federation for its support in the present work.
[28] D. Ding, Q.-L. Han, Y. Xiang, X. Ge, X.-M. Zhang, A survey on security control
and attack detection for industrial cyber-physical systems, Neurocomputing.
275 (2018) 1674–1683.
References [29] M.S. Mahmoud, M.M. Hamdan, U.A. Baroudi, Modeling and control of Cyber-
Physical Systems subject to cyber attacks: a survey of recent advances and
challenges, Neurocomputing. 338 (2019) 101–115.
[1] R. Vinayakumar, M. Alazab, K.P. Soman, P. Poornachandran, A. Al-Nemrat, S.
[30] I.H. Sarker, A.S.M. Kayes, S. Badsha, H. Alqahtani, P. Watters, A. Ng,
Venkatraman. Deep Learning Approach for Intelligent Intrusion Detection
Cybersecurity data science: an overview from machine learning perspective,
System [Internet]. Charles Darwin University’s Research Webportal. IEEE,
J. Big Data. 7 (1) (2020), https://doi.org/10.1186/s40537-020-00318-5.
Institute of Electrical and Electronics Engineers; 2019 [cited 2021Jan17].
[31] Blog SIGKDD. [Internet]. SIGKDD : KDD Cup 1999 : Computer network
Available from: https://researchers.cdu.edu.au/en/publications/deep-learning-
intrusion detection. [cited 2019Oct 20]. Available from: https://www.kdd.org/
approach-for-intelligent-intrusion-detection-system.
kdd-cup/view/kdd-cup-1999.
[2] A. Tantawy, S. Abdelwahed, A. Erradi, K. Shaban, Model-based risk assessment
[32] M. Tavallaee, E. Bagheri, W. Lu, A.A. Ghorbani, A detailed analysis of the KDD
for cyber physical systems security, Comput. Sec. 96 (2020) 101864, https://
CUP 99 data set, 2009 IEEE Symposium on Computational Intelligence for
doi.org/10.1016/j.cose.2020.101864.
Security and Defense Applications, 2009.
[3] G. Xinping, Y. Bo, C. Cailian, D. Wenbin, W. Yiyin, A comprehensive overview of
[33] CAIDA: Center for Applied Internet Data Analysis. Dataset for ‘‘DDoS Attack
cyber-physical systems: from perspective of feedback system, IEEE/CAA J.
2007” - Request Form [Internet]. CAIDA. [cited 2019 Oct 20]. Available from:
Autom. Sin. 3 (1) (2016) 1–14.
https://www.caida.org/data/passive/ddos_dataset_request.xml.
[4] L. Guo, J. Ye, Cyber-physical security of electric vehicles with four motor drives,
[34] Faculty of Engineering [Internet]. Botnet and Ransomware Detection Datasets
IEEE Trans. Power Electron. 36 (4) (2021) 4463–4477.
- University of Victoria. [cited 2019 Oct 19]. Available from: https://www.uvic.
[5] I. Akkaya, P. Derler, S. Emoto, E.A. Lee, Systems engineering for industrial
ca/engineering/ece/isot/datasets/botnet-ransomware/index.php.
cyber-physical systems using aspects, Proc. IEEE 104 (5) (2016) 997–1012.
[35] Search UNB [Internet]. University of New Brunswick est.1785. [cited 2019 Oct
[6] T. Wang, Y. Liang, Y.i. Yang, G. Xu, H. Peng, A. Liu, W. Jia, An intelligent edge-
20]. Available from: https://www.unb.ca/cic/datasets/index.html.
computing-based method to counter coupling problems in cyber-physical
[36] The CTU-13 Dataset. A Labeled Dataset with Botnet, Normal and Background
systems, IEEE Network 34 (3) (2020) 16–22.
traffic. [Internet]. Stratosphere IPS. [cited 2019Oct 20]. Available from: https://
[7] Y. Chen, C.M. Poskitt, J. Sun, Learning from Mutants: Using Code Mutation to
www.stratosphereips.org/datasets-ctu13/.
Learn and Monitor Invariants of a Cyber-Physical System, 2018 IEEE
[37] N. Moustafa, J. Slay, UNSW-NB15: a comprehensive data set for network
Symposium on Security and Privacy (SP), 2018.
intrusion detection systems (UNSW-NB15 network data set), 2015 Military
[8] G. Settanni, F. Skopik, A. Karaj, M. Wurzenberger, R. Fiedler, Protecting cyber
Communications and Information Systems Conference (MilCIS), 2015.
physical production systems using anomaly detection to enable self-
[38] Search UNB [Internet]. University of New Brunswick est.1785. [cited 2020].
adaptation, 2018 IEEE Industrial Cyber-Physical Systems (ICPS), 2018.
Available from: https://www.unb.ca/cic/datasets/ids-2018.html.
[9] R.F. Babiceanu, R. Seker, Big Data and virtualization for manufacturing cyber-
[39] X. Jing, Z. Yan, X. Jiang, W. Pedrycz, Network traffic fusion and analysis against
physical systems: A survey of the current status and future outlook, Comput.
DDoS flooding attacks with a novel reversible sketch, Inf. Fusion. 51 (2019)
Ind. 81 (2016) 128–137.
100–113.
[10] C. Machado, A.A. Medeiros Frohlich, IoT Data Integrity Verification for Cyber-
[40] B. Lindauer, J. Glasser, M. Rosen, K. Wallnau, Generating test data for insider
Physical Systems Using Blockchain, 2018 IEEE 21st International Symposium
threat detectors, J. Wireless Mobile Netw. Ubiquitous Comput. Dependable
on Real-Time Distributed Computing (ISORC), 2018.
Appl. 5 (2) (2014) 80–94.
[11] L. Wang, M. Törngren, M. Onori, Current status and advancement of cyber-
[41] J. Glasser, B. Lindauer, Bridging the Gap: A Pragmatic Approach to Generating
physical systems in manufacturing, J. Manuf. Syst. 37 (2015) 517–527.
Insider Threat Data, 2013 IEEE Security and Privacy Workshops, 2013.
[12] A. Napoleone, M. Macchi, A. Pozzetti, A review on the characteristics of cyber-
[42] Y. Zhou, X. Jiang, Dissecting android malware: characterization and evolution,
physical systems for the future smart factories, J. Manuf. Syst. 54 (2020) 305–
IEEE Sympos. Sec. Privacy 2012 (2012) 95–109.
335.
[43] N. Koroniotis, N. Moustafa, E. Sitnikova, B. Turnbull, Towards the development
[13] F. Allgöwer, J. Borges de Sousa, J. Kapinski, P. Mosterman, J. Oehlerking, P.
of realistic botnet dataset in the Internet of Things for network forensic
Panciatici, et al., Position paper on the challenges posed by modern
analytics: Bot-IoT dataset, Fut. Generat. Comput. Syst. 100 (2019) 779–796.
applications to cyber-physical systems theory, Nonlinear Anal. Hybrid Syst.
[44] K. Shaukat, S. Luo, V. Varadharajan, I.A. Hameed, M. Xu, A Survey on machine
34 (2019) 147–165.
learning techniques for cyber security in the last decade, IEEE Access 8 (2020)
[14] M.U. Hassan, M.H. Rehmani, J. Chen, Differential privacy techniques for cyber
222310–222354.
physical systems: a survey, IEEE Commun. Surv. Tutorials 22 (1) (2020) 746–
[45] H.A. Kholidy, Autonomous mitigation of cyber risks in the cyber-physical
789.
systems, Fut. Generat. Comput. Syst. 115 (2021) 171–187.
[15] P. Hehenberger, B. Vogel-Heuser, D. Bradley, B. Eynard, T. Tomiyama, S.
[46] Z. Lv, D. Chen, R. Lou, A. Alazab, Artificial intelligence for securing industrial-
Achiche, Design, modelling, simulation and integration of cyber physical
based cyber–physical systems, Fut. Generat. Comput. Syst. 117 (2021) 291–
systems: methods and applications, Comput. Ind. 82 (2016) 273–289.
298.
[16] J. Lee, C. Jin, B. Bagheri, Cyber physical systems for predictive production
[47] F. Salahdine, N. Kaabouch, Security threats, detection, and countermeasures
systems, Prod. Eng. Res. Dev. 11 (2) (2017) 155–165.
for physical layer in cognitive radio networks: a survey, Phys. Commun. 39
[17] F. Farivar, M.S. Haghighi, A. Jolfaei, M. Alazab, Artificial intelligence for
(2020) 101001, https://doi.org/10.1016/j.phycom.2020.101001.
detection, estimation, and compensation of malicious attacks in nonlinear
[48] A. Sargolzaei, C.D. Crane, A. Abbaspour, S. Noei, A Machine Learning Approach
cyber-physical systems and industrial IoT, IEEE Trans. Ind. Inf. 16 (4) (2020)
for Fault Detection in Vehicular Cyber-Physical Systems, 2016 15th IEEE
2716–2725.
International Conference on Machine Learning and Applications (ICMLA),
[18] H. Xu, W. Yu, D. Griffith, N. Golmie, A survey on industrial internet of things: a
2016.
cyber-physical systems perspective, IEEE Access 6 (2018) 78238–78259.
[49] J. Goh, S. Adepu, M. Tan, Z.S. Lee, Anomaly Detection in Cyber Physical Systems
[19] J.-P. Yaacoub, O. Salman, H.N. Noura, N. Kaaniche, A. Chehab, M. Malli, Cyber-
Using Recurrent Neural Networks, 2017 IEEE 18th International Symposium
physical systems security: Limitations, issues and future trends, Microprocess.
on High Assurance Systems Engineering (HASE), 2017.
Microsyst. 77 (2020) 103201, https://doi.org/10.1016/j.micpro.2020.103201.
[50] M. Wu, Z. Song, Y.B. Moon, Detecting cyber-physical attacks in Cyber
[20] C. Konstantinou, M. Maniatakos, F. Saqib, S. Hu, J. Plusquellic, Y. Jin, Cyber-
Manufacturing systems with machine learning methods, J. Intell. Manuf. 30
physical systems: A security perspective, 2015 20th IEEE European Test
(3) (2019) 1111–1123.
Symposium (ETS), 2015.
[51] S. Adepu, N.K. Kandasamy, A.E.P.I.C. Mathur, An electric power testbed for
[21] O. Kocabas, T. Soyata, M.K. Aktas, Emerging security mechanisms for medical
research and training in cyber physical systems security, Comput. Sec. (2019)
cyber physical systems, IEEE/ACM Trans. Comput. Biol. Bioinf. 13 (3) (2016)
37–52.
401–416.
[52] Y. Liu, Y. Peng, B. Wang, S. Yao, Z. Liu, Review on cyber-physical systems, IEEE/
[22] F. Hu, Y.u. Lu, A.V. Vasilakos, Q.i. Hao, R. Ma, Y. Patil, T. Zhang, J. Lu, X. Li, N.N.
CAA J. Autom. Sin. 4 (1) (2017) 27–40.
Xiong, Robust cyber-physical systems: concept, models, and implementation,
Fut. Generat. Comput. Syst. 56 (2016) 449–475.

4
A. Ahmed Jamal, Al-Ani Mustafa Majid, A. Konev et al. Materials Today: Proceedings xxx (xxxx) xxx

[53] J. Wurm, Y. Jin, Y. Liu, S. Hu, K. Heffner, F. Rahman, M. Tehranipoor, [61] B. Alhayani, S.T. Abbas, H.J. Mohammed, et al., Intelligent secured two-way
Introduction to cyber-physical system security: a cross-layer perspective, image transmission using corvus corone module over WSN, Wireless Pers.
IEEE Trans. Multi-Scale Comput. Syst. 3 (3) (2017) 215–227. Commun. (2021), https://doi.org/10.1007/s11277-021-08484-2.
[54] Z. Wang, H. Song, D.W. Watkins, K.G. Ong, P. Xue, Q. Yang, X. Shi, Cyber- [62] H.S. Hasan, B. Alhayani, et al. Novel unilateral dental expander appliance
physical systems for water sustainability: challenges and opportunities, IEEE (udex): a compound innovative materials, Comput. Mater. Continua 68(3)
Commun. Mag. 53 (5) (2015) 216–222. 2021 3499–3511 10.32604/cmc.2021.015968.
[55] M.S. Chong, H. Sandberg, A.M.H. Teixeira. A Tutorial Introduction to Security [63] A.S. Kwekha-Rashid, H.N. Abduljabbar, B. Alhayani, Coronavirus disease
and Privacy for Cyber-Physical Systems. 2019 18th European Control (COVID-19) cases analysis using machine-learning applications, Appl.
Conference (ECC). 2019; Nanosci. (2021), https://doi.org/10.1007/s13204-021-01868-7.
[56] Y. Jiang, S. Yin, O. Kaynak, Data-driven monitoring and safety control of [64] B. Alhayani, H. Ilhan, Hyper spectral image classification using dimensionality
industrial cyber-physical systems: basics and beyond, IEEE Access 6 (2018) reduction techniques, Int. J. Innov. Res. Electr. Electron. Instrum. Control Eng. 5
47374–47384. (4) (2017) 71–74.
[57] F.C. Delicato, A. Al-Anbuky, K.-K. Wang, Smart cyber-physical systems: toward [65] B. Al Hayani, H. Ilhan, Image transmission over decode and forward based
pervasive intelligence systems, Fut. Generat. Comput. Syst. 107 (2020) 1134– cooperative wireless multimedia sensor networks for Rayleigh fading channels
1139. in medical Internet of Things (MIoT) for remote health-care and health
[58] B. Alhayani, A.A. Abdallah. Manufacturing intelligent Corvus corone module communication monitoring, J. Med. Imaging Heal. Informatics 10 (1) (2020)
for a secured two way image transmission under WSN, Eng. Comput. Vol. 160–168.
ahead-of-print No. ahead-of-print. (2020), 10.1108/EC-02-2020-0107. [66] B. Alhayani, H. Jasim Mohammed, I.Z. Chaloob, J. Saleh Ahmed, Effectiveness of
[59] B.S.A. Alhayani, H. llhan. Visual sensor intelligent module based image artificial intelligence techniques against cyber security risks apply of IT
transmission in industrial manufacturing for monitoring and manipulation industry, Mater. Today Proc. (2021), https://doi.org/10.1016/
problems, J. Intell. Manuf. 32 2021 597–610 10.1007/s10845-020-01590-1. j.matpr.2021.02.531.
[60] Bilal Al-Hayani, Haci Ilhan, Efficient cooperative image transmission in one- [67] B. Alhayani, S.T. Abbas, D.Z. Khutar, H.J. Mohammed, Best ways computation
way multi-hop sensor network, Int. J. Electr. Eng. Educ. 57 (4) (2020) 321–339. intelligent of face cyber attacks, Mater. Today Proc. (2021), https://doi.org/
10.1016/j.matpr.2021.02.557.

You might also like