A mode-dependent stability criterion for delayed discrete-time stochastic neural networks with Markovian jumping parameters

This paper investigates the problem of stability for a class of discrete-time stochastic neural networks (DSNNs) with mode-dependent delay and Markovian jumping parameters. Throughout this paper, we assume that stochastic disturbances are described by the Brownian motion, jumping parameters are generated from discrete-time discrete-state homogeneous Markov process, and mode-dependent delay d(r(k))d(r(k)) satisfies dm≤d(r(k))≤dMdm≤d(r(k))≤dM. By a novel Lyapunov–Krasovskii functional combining with the delay partitioning technique and the free-weighting matrix method in terms of linear matrix inequalities (LMIs), the new stability criterion proves to be less conservative. Finally, numerical examples are given to illustrate the effectiveness of the proposed method.