Security-based control design for synchronization of switched reaction diffusion neural networks with hybrid attacks

This study delves into exploring dissipative synchronization for a class of switched neural networks with external disturbances featuring reaction–diffusion terms under the master–slave scheme. Precisely, the addressed network model comprises a hybrid attack model which entails both deception and denial-of-service attacks. Moreover, security-based control is designed to achieve the intended results, wherein in the realm of control design, the likelihood of cyber attacks is dictated by two separate and independent stochastic Bernoulli distributed factors. Meanwhile, the dissipative theory is employed to effectively curb the external disturbances within the network model. Subsequently, by leveraging the Lyapunov stability theory and linear matrix inequality approach, adequate conditions are acquired for ensuring the mean square exponential synchronization and strict $({\Gamma }_1,{\Gamma }_2,{\Gamma }_3)$-$\theta$ dissipativity of the examined system. Furthermore, the relation for deriving the control gain matrices is set forth in accordance with the acquired criteria. At the end, a numerical example accompanied by simulation results is supplied to vividly demonstrate the efficacy and significance of the acquired theoretical insights.