Detection And Description Of Multiple Actuator And Sensor Cyber-Attacks For Cyber-Physical Systems

A Cyber-Physical Systems(CPS),which is an integration of computation,networking and physical processes,plays an increasingly important role in critical infrastructure,government and everyday life.Due to physical constraints,embedded computers and networks may give rise to some additional security vulnerabilities.The vulnerabilities might allow an attacker to penetrate a network,gain access to control software,alter load conditions to destabilize the control system and successful attacks that results in losses of enormous economy benefits or disorder of social life.For these reasons,is necessary to detect when several cyber-attacks aiming to the actuator and sensor communication link occur.The main goal of this research is the detection and description of multiple actuator and sensor cyber-attacks for CPS using the residual information obtained after stochastic cyberattacks to several control and measurement communication links.The detection is based on frequency domain transformation technique and auxiliary detection tools.At the beginning,several stochastic cyber-attacks are designed through Algebraic Riccati Equation,as a reliable method and satisfying Markovian stochastic processes.In a real world,the hackers might attempt to launch multiple sensors and actuator attacks;for this reason,four cases that demonstrated the feasibility and effectivity of this method are developed.The conditions of detectability of the cyber-attacks are determined,and cyber-attacks as non-zero solution of the algebraic equations are considered.The residual information is widely related to stochastic cyber-attacks.A comparison between Kalman filter and algebraic approach is done and the simulation results show that the detection approach is effective and feasible.The presented detection approach is simple and straightforward.