New Criteria for Global Exponential Stability of Discrete-Time High-Order Neural Networks with Time-Varying Delays

This article investigates the global exponential stability of discrete-time high-order neural networks with time-varying delays. Firstly, a novel method is proposed to establish delay-dependent sufficient conditions for global exponential stability of the zero equilibrium. The sufficient conditions include a simple matrix inequality. Then two numerical examples are given to illustrate the effectiveness of theoretical results. The proposed method has two advantages: (i) It is directly based on the definition of global exponential stability, and does not need to construct any Lyapunov–Krasovskii functional; (ii) it can result in stability criteria that are easy to verify and less conservative. More valuable, the method presented in this article is suitable for most discrete-time neural networks.