Adaptive Event-Triggered Filter Design For Cyber Physical Systems Under Non-Periodic Denial-of-Service Attacks

In the last few decades,cyber-physical systems(CPSs)have become new-generation engineering systems formed by the networks environment.Due to the involvement of networks,security against cyber-attacks and resource optimization are the primary performance indicator of CPSs,which have excellent research prospects.In this paper,we study the adaptive event triggering scheme(AETS)based H_? filtering problem for CPSs under non-periodic denial-of-service(Do S)attacks.Although significant work has been done on networked filtering,there are still some imperfections that can be found,such as consideration of some traditional network communication problems and overlooked network security issues.On the other hand,different from the conventional time-triggering scheme,an AETS constructed on an adaptive threshold parameter and sends data over the network when a particular event occurs.In this way,the network channel can be prevented from the Do S attacks affects,and it further boosts the network resources utilization by reducing the data occupied by the network channel.However,the integration of AETS and Do S jamming attacks is very significant and challenging for the CPSs filtering problem.Therefore,the solution to the aforementioned problem is effectively proposed in this paper.The main content of this includes the following contributions:1)Firstly,considering the effect of energy-constrained,non-periodic Do S jamming attacks on the network,a novel AETS developed to resist the effect of attacks and optimize the system resources utilization,e.g.,network bandwidth resources,electric energy resource.2)Secondly,under the above-stated AETS and the effect of Do S attacks on network communication,a switched systems model with time-varying delay for filtering-errorsystem is established by using a time-delayed system modeling approach.3)Thirdly,under the Lyapunov-Krasovskii stability theory,some stability criteria in the form of linear matrix inequalities(LMIs)are obtained,which ensure that the filtering-errorsystem is globally exponential stable(GES)with a weighted H_? performance.4)Based on the above stability criteria,also,the co-design of AETS parameters and filter parameters matrices is implemented.The co-design of the desired filter provides sufficient conditions to check the feasibility of the above mentioned LMI sets.Finally,the effectiveness of the proposed strategy is verified by implementing a practical example of an active suspension of quarter-car model in MATLAB.