A Multi-Sensor-Based Switching Event-Triggered Mechanism for Synchronization Control of Markovian Jump Neural Networks Under DoS Attacks

This paper investigates the secure synchronization control of Markovian jump neural networks (MJNNs) suffering from denial of service attacks (DoS attacks). The issue is presented for two reasons: 1) multiple sensors are generally used to measure the information of different state variables in practical networked control systems; 2) DoS attacks will disrupt the transmission of data packets in communication channels, thereby causing the system performance degradation. To deal with these problems, a multi-sensor-based switching event-triggered mechanism (SETM) is designed. More specifically, when the data packets are transmitted normally, the SETM will switch to an adaptive memory-based event-triggered mechanism, thereby saving the network resources. When the DoS attack occurs, the SETM will trigger immediately at the end of the DoS attack to improve the system performance. To facilitate the stability analysis, a merging time series is constructed by integrating the triggering instants of successful transmission and the ending instants of DoS attacks together. In light of the merging time series, a switched closed-loop system is established. Then, by utilizing the stability analysis idea of switched systems, a multiple Lyapunov functional is constructed, enabling the exploitation of multi-sampling. On this basis, a synchronization criterion is derived, accompanied by a co-design method for controller and the trigger matrices. Finally, the outcomes of the simulation confirm both the efficacy and the advantage of the suggested approach, especially when dealing with DoS attacks.