Detection and Prevention of Cyber-Attacks in Cyber-Physical Systems based on Nature Inspired Algorithm

Abstract

Cyber-Physical Systems (CPSs) are the complicated formulation of cyber and physical equipment that effectively utilizing the capabilities of sensors, actuators, networks, and communication systems. The system is vulnerable to various cyber-attacks in the smart grid that are active and passive attacks, DoS attacks, data injection attacks, replay attacks, etc., With the exponential growth of communication and electronic devices, it is a large and complex task to manage these devices from vulnerabilities. However, cyber-attacks on CPS can cause severe harm to the resources, network infrastructure, and communication channels with the key objectives of confidentiality, integrity, and authentication. This paper considers the design of detecting cyber-attacks during transmission from control center to plant over the networked control center and preventing cyber-attacks with the help of a genetic algorithm and deep feedforward neural networks. This method ensures the security of networked control center in CPS environment of smart grid. This also mitigates the drawbacks of existing attack detection techniques. The performance is evaluated in benchmarking IEEE 39 bus system for achieving the accuracy, performance metrics, and low false positive rate to improve the attack detection in a smart grid environment in cyberspace.