Development of Cyber Security Mechanism to Detect Cyber Attacks on Cyber Physical Systems by Using Bayesian Belief Networks

Development of Cyber Security Mechanism to Detect Cyber Attacks on Cyber Physical Systems by Using Bayesian Belief Networks

D Suma, D Roopa, Pirangi Hymavathi, M Madhavi Latha, Dr. Mahesh Kotha

Assistant Professor, Department of CSE (AI/DS), Sri Indu College of Engineering and Technology, Hyderabad.

Assistant Professor, CSE Department, Sri Indu College of Engineering and Technology, Hyderabad.

Assistant Professor, CSE Department, Sri Indu College of Engineering and Technology, Hyderabad.

Assistant Professor, Princeton Institute of Engineering and Technology for Women, Hyderabad.

Associate professor, Department of CSE (AI&ML), CMR Technical Campus, Hyderabad.

Abstract: A cyber physical systems (CPS) is a complicated device that integrates sensing, computation, manipulate and networking into bodily methods and items over Internet. It performs a key function in cutting-edge enterprise because it connects bodily and cyber worlds. In order to fulfill ever-converting commercial requirements, its systems and capabilities are continuously improved. Identifying cyberattack vectors on cyber supply chains (CSC) withinside the occasion of cyber assaults are very vital in mitigating cybercrimes successfully on Cyber Physical Systems CPS. A ubiquitous hassle is the truth that cyber assaults can purpose full-size harm to commercial systems, and as a result has received growing interest from researchers and practitioners. However, withinside the cyber protection domain, the invincibility nature of cybercrimes makes it hard and hard to are expecting the chance opportunity and effect of cyber assaults. Although cybercrime phenomenon, risks, and treats include a number of unpredictability’s, uncertainties and fuzziness, cyberattack detection have to be practical, methodical and affordable to be implemented. We discover Bayesian Belief Networks (BBN) as information illustration in synthetic intelligence which will be officially carried out probabilistic inference withinside the cyber protection domain. The goal of this paper is to apply Bayesian Belief Networks to hit upon cyberattacks on CSC withinside the CPS domain. We version cyberattacks the usage of DAG approach to decide the assault propagation. Further, we use a clever grid case observe to illustrate the applicability of assault and the cascading effects. The consequences display that BBN will be tailored to decide uncertainties withinside the occasion of cyberattacks withinside the CSC domain. In this paper, we gift a Bayesian community technique for mastering the causal family members among cyber and bodily variables in addition to their temporal correlations from unlabeled information. We describe the information variations that we finished to cope with the heterogeneous traits of the cyber and bodily information, in order that the included dataset may be used to analyze the Bayesian community shape and parameters. We then gift scalable algorithms to hit upon special anomalies and isolate their respective root-purpose the usage of a Bayesian community. We additionally gift consequences from comparing our algorithms on an unlabeled dataset which include anomalies because of cyber assaults and bodily faults in a industrial constructing device.

Keywords: Cyber Physical System, Cyber Attacks, Cyber Supply Chain Threats, Bayesian Belief Network, Cybercrime.