Dynamic-memory event-based resilient filtering with ℒ1 performance for positive singular semi-Markov jumping systems against multiple cyber-attacks

This paper focuses on the design of resilient filters for positive singular semi-Markov jump systems with dynamic time-triggering mechanism against multiple cyber-attacks. Firstly, to improve the communication efficiency and reduce the communication pressure, a dynamic time-triggering mechanism is introduced. At the same time, a complex unified cyber-attack model including DoS attacks and spoofing attacks is established. In addition, the resilient filter design with L1 performance for positive system is proposed and HMM is used to reflect the asynchronous phenomenon between the original system and the resilient filter. Then, the filtering error system is obtained with the general conditional probability. The criterion of stochastically admissible for the filter error system is obtained by linear matrix inequality. Finally, a practical example is given to verify the validity of the results.